x research

Fewer people are using Elon Musk’s X as the platform struggles to attract and keep users, according to analysts

The number of people using X daily is falling, more than a year after tech billionaire Elon Musk bought the app formerly known as Twitter. 

Data from two research firms and figures published by Musk and X suggest a deteriorating situation for X by some metrics. Musk has marketed it as the world’s “town square,” but in number of users it continues to lag far behind social media rivals that focus on video, such as Instagram and TikTok. 

In February, X had 27 million daily active users of its mobile app in the U.S., down 18% from a year earlier, according to Sensor Tower, a market intelligence firm based in San Francisco. The U.S. user base has been flat or down every month since November 2022, the first full month of Musk’s owning the app, and in total it’s down 23% since then, Sensor Tower said. 

The numbers were nearly as bad worldwide, as daily active users on the mobile app fell to 174 million in February, down 15% from a year earlier, the firm said. The worldwide user base has been flat or down every month during Musk’s tenure began except one, when it grew slightly in October and then resumed falling, according to Sensor Tower. 

Other social media apps experienced modest increases in their worldwide user bases during the same period, according to the research, with Snapchat growing 8.8%, Instagram 5.3%, Facebook 1.5% and TikTok 0.5%. Those apps all experienced declines over that period in the U.S., but none was as steep as the decline on X. 

X had “the most material decline in active users compared to its peers,” Abe Yousef, a senior insights analyst at Sensor Tower, wrote in a research report. 

“This decline in X mobile app active users may have been driven by user frustration over flagrant content, general platform technical issues, and the growing threat of short-form video platforms,” he wrote. 

Under Musk’s ownership, X has relaxed content moderation rules that previously limited hateful content , such as white supremacist imagery, and Musk has welcomed back to the platform some users whom the old Twitter management had banned. In December, he reinstated the accounts of conspiracy theorist Alex Jones and his Infowars website and then held a public audio-only event with Jones. 

X said in a post Monday that the worldwide number is higher than what Sensor Tower data shows, with 250 million people using X every day globally. That would still be a decrease from when Musk bought the app. Musk said in 2022 that, at about the time he completed the purchase in late October, Twitter had about 258 million “monetizable daily active users,” the company’s metric at the time.  

X didn’t respond to a request for additional information. It didn’t say how it defines who counts as an active user — a metric that by its name may appear straightforward but that tech platforms define differently. 

On a monthly basis, X has 550 million people using it, according to the company . That figure represents growth of 1.5% since July, when Musk said X had 542 million monthly users. 

Sensor Tower defines a daily active user as someone who “registered a session of at least two seconds in length, once in that day.” It says its data comes from a panel of consumers who provide access to their information in exchange for the use of other apps, including apps that track screen time. 

Advertisers have also left X, Sensor Tower said, with 75 out of the top 100 U.S. advertisers on X from October 2022 having ceased ad spending on it. The exodus spiked toward the end of last year, after Musk publicly embraced an antisemitic conspiracy theory and told advertisers at a conference in New York, “Go f--- yourself.” 

In recent days, Musk has urged his 177 million followers on X to get more people onto the platform. On Sunday, he posted instructions for how to share posts with friends, a basic function of social media. 

“Please send links from this platform to your friends who are still being misled by the legacy media!” he wrote in a separate post Sunday. 

Musk has also shifted the platform’s business model from being almost entirely ad-supported to one that also has four subscription tiers, from free to a Premium+ service that starts at $16 a month. 

Sensor Tower said that, according to its research, X’s revenue from in-app purchases last month was about $9.5 million, including for X subscriptions and payments to creators. 

“This still remains just a fraction of revenue that the company was previously generating from advertising in its last year as a public entity,” Yousef wrote. Twitter in July 2022 reported $1.18 billion in revenue for the previous three months. 

X has been helped by the lack of a clear text-based social media alternative. Threads, a competitor launched by Instagram and its parent company, Meta, had 1.6 million daily active U.S. mobile users in February, according to Sensor Tower, and 14 million worldwide. 

Threads has a potential major advantage over other upstart apps because it is closely integrated with Instagram — users of Instagram can see Threads posts in their feeds and create accounts relatively easily. That has translated into a whopping disparity in downloads, according to a second research firm, Apptopia, based in Boston, which said Threads beat X in downloads worldwide by an 8-to-1 ratio in February. 

Downloads were even more lopsided in the U.S. in February, with Threads getting about 16 downloads for every one download of X, Apptopia said. 

“For microblogging platforms, X had dominant market share of app downloads right up until Threads launched,” Tom Grant, vice president of research at Apptopia, wrote in an email. “That turned market share completely on its head.” 

There were 2.9 million downloads of X in the U.S. in February, up 14% from a year earlier but still below the 3.7 million in October 2022, the month Musk bought the company, according to Apptopia. Threads had 46.2 million downloads last month, Apptopia said. 

Threads has ranked highly in some app store rankings lately, topping Apple’s chart for free apps Sunday and staying in the Top 4 for most of this week. X ranked No. 34 on the Apple app store Sunday and No. 30 on Friday. On Friday, Threads ranked No. 7 in Google’s Play store and X ranked No. 43.

But so far, the downloads haven’t translated into sustained growth for Threads, according to Sensor Tower. Another X competitor, Bluesky, was even smaller, with 195,000 U.S. daily active mobile users in February, according to the research firm. 

In its own data summary published Monday, X said that “1.7 million people join X every day.” That number is roughly triple the number of daily X downloads worldwide, according to Apptopia, and it suggests that X is growing at a rate of nearly 10% per month — far faster than any other source indicates.

David Ingram covers tech for NBC News.

Google X and the Science of Radical Creativity

How the secretive Silicon Valley lab is trying to resurrect the lost art of invention

I. The Question

A snake-robot designer , a balloon scientist, a liquid-crystals technologist, an extradimensional physicist, a psychology geek, an electronic-materials wrangler, and a journalist walk into a room. The journalist turns to the assembled crowd and asks: Should we build houses on the ocean?

The setting is X, the so-called moonshot factory at Alphabet, the parent company of Google. And the scene is not the beginning of some elaborate joke. The people in this room have a particular talent: They dream up far-out answers to crucial problems. The dearth of housing in crowded and productive coastal cities is a crucial problem. Oceanic residences are, well, far-out. At the group’s invitation, I was proposing my own moonshot idea, despite deep fear that the group would mock it.

Like a think-tank panel with the instincts of an improv troupe, the group sprang into an interrogative frenzy. “What are the specific economic benefits of increasing housing supply?” the liquid-crystals guy asked. “Isn’t the real problem that transportation infrastructure is so expensive?” the balloon scientist said. “How sure are we that living in densely built cities makes us happier?” the extradimensional physicist wondered. Over the course of an hour, the conversation turned to the ergonomics of Tokyo’s high-speed trains and then to Americans’ cultural preference for suburbs. Members of the team discussed commonsense solutions to urban density, such as more money for transit, and eccentric ideas, such as acoustic technology to make apartments soundproof and self-driving housing units that could park on top of one another in a city center. At one point, teleportation enjoyed a brief hearing.

X is perhaps the only enterprise on the planet where regular investigation into the absurd is not just permitted but encouraged, and even required. X has quietly looked into space elevators and cold fusion. It has tried, and abandoned, projects to design hoverboards with magnetic levitation and to make affordable fuel from seawater. It has tried—and succeeded, in varying measures—to build self-driving cars, make drones that deliver aerodynamic packages, and design contact lenses that measure glucose levels in a diabetic person’s tears.

These ideas might sound too random to contain a unifying principle. But they do. Each X idea adheres to a simple three-part formula. First, it must address a huge problem; second, it must propose a radical solution; third, it must employ a relatively feasible technology. In other words, any idea can be a moonshot—unless it’s frivolous, small-bore, or impossible.

The purpose of X is not to solve Google’s problems; thousands of people are already doing that. Nor is its mission philanthropic. Instead X exists, ultimately, to create world-changing companies that could eventually become the next Google. The enterprise considers more than 100 ideas each year, in areas ranging from clean energy to artificial intelligence. But only a tiny percentage become “projects,” with full-time staff working on them. It’s too soon to know whether many (or any) of these shots will reach the moon: X was formed in 2010, and its projects take years; critics note a shortage of revenue to date. But several projects—most notably Waymo, its self-driving-car company , recently valued at $70 billion by one Wall Street firm—look like they may.

X is extremely secretive. The company won’t share its budget or staff numbers with investors, and it’s typically off-limits to journalists as well. But this summer, the organization let me spend several days talking with more than a dozen of its scientists, engineers, and thinkers. I asked to propose my own absurd idea in order to better understand the creative philosophy that undergirds its approach. That is how I wound up in a room debating a physicist and a roboticist about apartments floating off the coast of San Francisco.

I’d expected the team at X to sketch some floating houses on a whiteboard, or discuss ways to connect an ocean suburb to a city center, or just inform me that the idea was terrible. I was wrong. The table never once mentioned the words floating or ocean . My pitch merely inspired an inquiry into the purpose of housing and the shortfalls of U.S. infrastructure. It was my first lesson in radical creativity. Moonshots don’t begin with brainstorming clever answers. They start with the hard work of finding the right questions.

C reativity is an old practice but a new science. It was only in 1950 that J. P. Guilford, a renowned psychologist at the University of Southern California, introduced the discipline of creativity research in a major speech to the American Psychological Association. “I discuss the subject of creativity with considerable hesitation,” he began, “for it represents an area in which psychologists generally, whether they be angels or not, have feared to tread.” It was an auspicious time to investigate the subject of human ingenuity, particularly on the West Coast. In the next decade, the apricot farmland south of San Francisco took its first big steps toward becoming Silicon Valley.

Yet in the past 60 years, something strange has happened. As the academic study of creativity has bloomed, several key indicators of the country’s creative power have turned downward, some steeply. Entrepreneurship may have grown as a status symbol, but America’s start-up rate has been falling for decades. The label innovation may have spread like ragweed to cover every minuscule tweak of a soda can or a toothpaste flavor, but the rate of productivity growth has been mostly declining since the 1970s. Even Silicon Valley itself, an economic powerhouse, has come under fierce criticism for devoting its considerable talents to trivial problems , like making juice or hailing a freelancer to pick up your laundry.

Breakthrough technology results from two distinct activities that generally require different environments— invention and innovation . Invention is typically the work of scientists and researchers in laboratories, like the transistor, developed at Bell Laboratories in the 1940s. Innovation is an invention put to commercial use, like the transistor radio, sold by Texas Instruments in the 1950s. Seldom do the two activities occur successfully under the same roof. They tend to thrive in opposite conditions; while competition and consumer choice encourage innovation, invention has historically prospered in labs that are insulated from the pressure to generate profit.

The United States’ worst deficit today is not of incremental innovation but of breakthrough invention. Research-and-development spending has declined by two-thirds as a share of the federal budget since the 1960s. The great corporate research labs of the mid-20th century, such as Bell Labs and Xerox Palo Alto Research Center ( parc ), have shrunk and reined in their ambitions. America’s withdrawal from moonshots started with the decline in federal investment in basic science. Allowing well-funded and diverse teams to try to solve big problems is what gave us the nuclear age, the transistor, the computer, and the internet. Today, the U.S. is neglecting to plant the seeds of this kind of ambitious research, while complaining about the harvest.

No one at X would claim that it is on the verge of unleashing the next platform technology, like electricity or the internet—an invention that could lift an entire economy. Nor is the company’s specialty the kind of basic science that typically thrives at research universities. But what X is attempting is nonetheless audacious. It is investing in both invention and innovation. Its founders hope to demystify and routinize the entire process of making a technological breakthrough—to nurture each moonshot, from question to idea to discovery to product—and, in so doing, to write an operator’s manual for radical creativity.

II. The Inkling

Inside X’s Palo Alto headquarters, artifacts of projects and prototypes hang on the walls, as they might in a museum—an exhibition of alternative futures. A self-driving car is parked in the lobby. Drones shaped like Jedi starfighters are suspended from the rafters. Inside a three-story atrium, a large screen renders visitors as autonomous vehicles would see them—pointillist ghosts moving through a rainbow-colored grid. It looks like Seurat tried to paint an Atari game.

Just beyond the drones, I find Astro Teller. He is the leader of X, whose job title, captain of moonshots, is of a piece with his piratical, if perhaps self-conscious, charisma. He has a long black ponytail and silver goatee, and is wearing a long-sleeved T‑shirt, dark jeans, and large black Rollerblades. Fresh off an afternoon skate?, I ask. “Actually, I wear these around the office about 98 percent of the time,” he says. I glance at an X publicist to see whether he’s serious. Her expression says: Of course he is .

Teller, 47, descends from a formidable line of thinkers. His grandfathers were Edward Teller, the father of the hydrogen bomb, and Gérard Debreu, a mathematician who won a Nobel Prize in Economics. With a doctorate in artificial intelligence from Carnegie Mellon, Teller is an entrepreneur, a two-time novelist, and the author of a nonfiction book, Sacred Cows , on marriage and divorce—co-written with his second wife. His nickname, Astro, though painfully on the nose for the leader of a moonshot factory, was bestowed upon him in high school, by friends who said his flattop haircut resembled Astroturf. (His given name is Eric.)

In 2010, Teller joined a nascent division within Google that would use the company’s ample profits to explore bold new ideas, which Teller called “moonshots.” The name X was chosen as a purposeful placeholder—as in, We’ll solve for that later . The one clear directive was what X would not do. While almost every corporate research lab tries to improve the core product of the mother ship, X was conceived as a sort of anti–corporate research lab; its job was to solve big challenges anywhere except in Google’s core business.

When Teller took the helm of X (which is now a company, like Google, within Alphabet), he devised the three-part formula for an ideal moonshot project: an important question, a radical solution, and a feasible path to get there. The proposals could come from anywhere, including X employees, Google executives, and outside academics. But grand notions are cheap and abundant—especially in Silicon Valley, where world-saving claims are a debased currency—and actual breakthroughs are rare. So the first thing Teller needed to build was a way to kill all but the most promising ideas. He assembled a team of diverse experts, a kind of Justice League of nerds, to process hundreds of proposals quickly and promote only those with the right balance of audacity and achievability. He called it the Rapid Evaluation team.

In the landscape of ideas, Rapid Eval members aren’t vertical drillers but rather oil scouts, skillful in surveying the terrain for signs of pay dirt. You might say it’s Rapid Eval’s job to apply a kind of future-perfect analysis to every potential project: If this idea succeeds, what will have been the challenges? If it fails, what will have been the reasons?

The art of predicting which ideas will become hits is a popular subject of study among organizational psychologists. In academic jargon, it is sometimes known as “creative forecasting.” But what sorts of teams are best at forecasting the most-successful creations? Justin Berg, a professor at the Stanford Graduate School of Business, set out to answer this question in a 2016 study focused on, of all things, circus performances.

Berg found that there are two kinds of circus professionals: creators who imagine new acts, and managers who evaluate them. He collected more than 150 circus-performance videos and asked more than 300 circus creators and managers to watch them and predict the performers’ success with an audience. Then he compared their reactions with those of more than 13,000 ordinary viewers.

Creators, Berg found, were too enamored of their own concepts. But managers were too dismissive of truly novel acts. The most effective evaluation team, Berg concluded, was a group of creators. “A solitary creator might fall in love with weird stuff that isn’t broadly popular,” he told me, “but a panel of judges will reject anything too new. The ideal mix is a panel of creators who are also judges, like the teams at X.” The best evaluators are like player-coaches—they create, then manage, and then return to creating. “They’re hybrids,” Berg said.

R ich DeVaul is a hybrid. He is the leader of the Rapid Eval team but he has also, like many members, devoted himself to major projects at X. He has looked into the feasibility of space elevators that could transport cargo to satellites without a rocket ship and modeled airships that might transport goods and people in parts of the world without efficient roads, all without ever touching the ground. “At one point, I got really interested in cold fusion,” he said. “Because why not?”

One of DeVaul’s most consuming obsessions has been to connect the roughly 4 billion people around the world who don’t have access to high-speed internet. He considers the internet the steam engine or electrical grid of the 21st century—the platform technology for a long wave of economic development. DeVaul first proposed building a cheap, solar-powered tablet computer. But the Rapid Eval team suggested that he was aiming at the wrong target. The world’s biggest need wasn’t hardware but access. Cables and towers were too expensive to build in mountains and jungles, and earthbound towers don’t send signals widely enough to make sense for poor, sparsely populated areas. The cost of satellites made those, too, prohibitive for poor areas. DeVaul needed something inexpensive that could live in the airspace between existing towers and satellites. His answer: balloons. Really big balloons.

The idea struck more than a few people as ridiculous. “I thought I was going to be able to prove it impossible really quickly,” said Cliff L. Biffle, a computer scientist and Rapid Eval manager who has been at X for six years. “But I totally failed. It was really annoying.” Here was an idea, the team concluded, that could actually work: a network of balloons, equipped with computers powered by solar energy, floating 13 miles above the Earth, distributing internet to the world. The cause was huge; the solution was radical; the technology was feasible. They gave it a name: Project Loon.

At first, Loon team members thought the hardest problem would be sustaining an internet connection between the ground and a balloon. DeVaul and Biffle bought several helium balloons, attached little Wi‑Fi devices to them, and let them go at Dinosaur Point, in the Central Valley. As the balloons sluiced through the jet stream, DeVaul and his colleagues chased them down in a Subaru Forester rigged with directional antennae to catch the signal. They drove like madmen along the San Luis Reservoir as the balloons soared into the stratosphere. To their astonishment, the internet connection held. DeVaul was ecstatic, his steampunk vision of broadband-by-balloon seemingly within grasp. “I thought, The rest is just ballooning! ” he said. “ That’s not rocket science .”

He was right, in a way. Ballooning of the sort his team imagined isn’t rocket science. It’s harder.

Let’s start with the balloons. Each one, flattened, is the size of a tennis court, made of stitched-together pieces of polyethylene. At the bottom of the balloon hangs a small, lightweight computer with the same technology you would find at the top of a cell tower, with transceivers to beam internet signals and get information from ground stations. The computer system is powered by solar panels. The balloon is designed to float 70,000 feet above the Earth for months in one stretch. The next time you are at cruising altitude in an airplane, imagine seeing a balloon as far above you as the Earth is far below.

The balloons have to survive in what is essentially an alien environment. At night, the temperature plunges to 80 degrees below zero Celsius, colder than your average evening on Mars. By day, the sun could fry a typical computer, and the air is too thin for a fan to cool the motherboard. So Loon engineers store the computer system in a specially constructed box—the original was a Styrofoam beer cooler—coated with reflective white paint.

The computer system, guided by an earthbound data center, can give the balloon directions (“Go northeast to Lima!”), but the stratosphere is not an orderly street grid in which traffic flows in predictable directions. It takes its name from the many strata, or layers, of air temperatures and wind currents. It’s difficult to predict which way the stratosphere’s winds will blow. To navigate above a particular town—say, Lima—the balloon cannot just pick any altitude and cruise. It must dive and ascend thousands of feet, sampling the gusts of various altitudes, until it finds one that is pointing in just the right direction. So Loon uses a team of balloons to provide constant coverage to a larger area. As one floats off, another moves in to take its place.

Four years after Loon’s first real test, in New Zealand, the project is in talks with telecommunications companies around the world, especially where cell towers are hard to build, like the dense jungles and mountains of Peru. Today a network of broadband-beaming balloons floats above rural areas outside of Lima, delivering the internet through the provider Telefónica.

Improving internet access in Latin America, Africa, and Asia to levels now seen in developed countries would generate more than $2 trillion in additional GDP, according to a recent study by Deloitte. Loon is still far from its global vision, but capturing even a sliver of one percentage point of that growth would make it a multibillion-dollar business.

III. The Fail

Astro Teller likes to recount an allegorical tale of a firm that has to get a monkey to stand on top of a 10-foot pedestal and recite passages from Shakespeare. Where would you begin? he asks. To show off early progress to bosses and investors, many people would start with the pedestal. That’s the worst possible choice, Teller says. “You can always build the pedestal. All of the risk and the learning comes from the extremely hard work of first training the monkey.” An X saying is “#MonkeyFirst”—yes, with the hashtag—and it means “do the hardest thing first.”

But most people don’t want to do the hardest thing first. Most people want to go to work and get high fives and backslaps. Despite the conference-keynote pabulum about failure (“Fail fast! Fail often!”), the truth is that, financially and psychologically, failure sucks. In most companies, projects that don’t work out are stigmatized, and their staffs are fired. That’s as true in many parts of Silicon Valley as it is anywhere else. X may initially seem like a paradise of curiosity and carefree tinkering, a world apart from the drudgery required at a public company facing the drumbeat of earnings reports. But it’s also a place immersed in failure. Most green-lit Rapid Eval projects are unsuccessful, even after weeks, months, or years of one little failure after another.

At X, Teller and his deputies have had to build a unique emotional climate, where people are excited to take big risks despite the inevitability of, as Teller delicately puts it, “falling flat on their face.” X employees like to bring up the concept of “psychological safety.” I initially winced when I heard the term, which sounded like New Age fluff. But it turns out to be an important element of X’s culture, the engineering of which has been nearly as deliberate as that of, say, Loon’s balloons.

Kathy Hannun told me of her initial anxiety, as the youngest employee at X, when she joined in the spring of 2012. On her first day, she was pulled into a meeting with Teller and other X executives where, by her account, she stammered and flubbed several comments for fear of appearing out of her depth. But everyone, at times, is out of his or her depth at X. After the meeting, Teller told her not to worry about making stupid comments or asking ignorant questions. He would not turn on her, he said.

Hannun now serves as the CEO of Dandelion, an X spin-off that uses geothermal technology to provide homes in New York State with a renewable source of heating, cooling, and hot water. “I did my fair share of unwise and inexperienced things over the years, but Astro was true to his word,” she told me. The culture, she said, walked a line between patience and high expectations, with each quality tempering the other.

X encourages its most successful employees to talk about the winding and potholed road to breakthrough invention. This spring, André Prager, a German mechanical engineer, delivered a 25-minute presentation on this topic at a company meeting, joined by members of X’s drone team, called Project Wing. He spoke about his work on the project, which was founded on the idea that drones could be significant players in the burgeoning delivery economy. The idea had its drawbacks: Dogs may attack a drone that lands, and elevated platforms are expensive, so Wing’s engineers needed a no-landing/no-infrastructure solution. After sifting through hundreds of ideas, they settled on an automatic winching system that lowered and raised a specialized spherical hook—one that can’t catch on clothing or tree branches or anything else—to which a package could be attached.

In their address, Prager and his team spent less time on their breakthroughs than on the many failed cardboard models they discarded along the way. The lesson they and Teller wanted to communicate is that simplicity, a goal of every product, is in fact extremely complicated to design. “The best designs—a bicycle, a paper clip—you look and think, Well of course, it always had to look like that ,” Prager told me. “But the less design you see, the more work was needed to get there.” X tries to celebrate the long journey of high-risk experimentation, whether it leads to the simplicity of a fine invention or the mess of failure.

Because the latter possibility is high, the company has also created financial rewards for team members who shut down projects that are likely to fail. For several years, Hannun led another group, named Foghorn, which developed technology to turn seawater into affordable fuel. The team appeared to be on track, until the price of oil collapsed in 2015 and its members forecast that their fuel couldn’t compete with regular gasoline soon enough to justify keeping the project alive. In 2016, they submitted a detailed report explaining that, despite advancing the science, their technology would not be economically viable in the near future. They argued for the project to be shut down. For this, the entire team received a bonus.

Some might consider these so-called failure bonuses to be a bad incentive. But Teller says it’s just smart business. The worst scenario for X is for many doomed projects to languish for years in purgatory, sucking up staff and resources. It is cheaper to reward employees who can say, “We tried our best, and this just didn’t work out.”

Recently, X has gone further in accommodating and celebrating failure. In the summer of 2016, the head of diversity and inclusion, a Puerto Rican–born woman named Gina Rudan, spoke with several X employees whose projects were stuck or shut down and found that they were carrying heavy emotional baggage. She approached X’s leadership with an idea based on Mexico’s Día de los Muertos , or Day of the Dead. She suggested that the company hold an annual celebration to share stories of pain from defunct projects. Last November, X employees gathered in the main hall to hear testimonials, not only about failed experiments but also about failed relationships, family deaths, and personal tragedies. They placed old prototypes and family mementos on a small altar. It was, several X employees told me, a resoundingly successful and deeply emotional event.

N o failure at X has been more public than Google Glass, the infamous head-mounted wearable computer that resembled a pair of spectacles. Glass was meant to be the world’s next great hardware evolution after the smartphone. Even more quixotically, its hands-free technology was billed as a way to emancipate people from their screens, making technology a seamless feature of the natural world. (To critics, it was a ploy to eventually push Google ads as close to people’s corneas as possible.) After a dazzling launch in 2013 that included a 12-page spread in Vogue , consumers roundly dissed the product as buggy, creepy, and pointless. The last of its dwindling advocates were branded “glassholes.”

I found that X employees were eager to talk about the lessons they drew from Glass’s failure. Two lessons, in particular, kept coming up in our conversations. First, they said, Glass flopped not because it was a bad consumer product but because it wasn’t a consumer product at all. The engineering team at X had wanted to send Glass prototypes to a few thousand tech nerds to get feedback. But as buzz about Glass grew, Google, led by its gung-ho co-founder Sergey Brin, pushed for a larger publicity tour—including a ted Talk and a fashion show with Diane von Furstenberg. Photographers captured Glass on the faces of some of the world’s biggest celebrities, including Beyoncé and Prince Charles, and Google seemed to embrace the publicity. At least implicitly, Google promised a product. It mailed a prototype. (Four years later, Glass has reemerged as a tool for factory workers, the same group that showed the most enthusiasm for the initial design.)

But Teller and others also saw Glass’s failure as representative of a larger structural flaw within X. It had no systemic way of turning science projects into businesses, or at least it hadn’t put enough thought into that part of the process. So X created a new stage, called Foundry, to serve as a kind of incubator for scientific breakthroughs as its team develops a business model. The division is led by Obi Felten, a Google veteran whose title says it all: head of getting moonshots ready for contact with the real world.

“When I came here,” Felten told me, “X was this amazing place full of deep, deep, deep geeks, most of whom had never taken a product out into the world.” In Foundry, the geeks team up with former entrepreneurs, business strategists from firms like McKinsey, designers, and user-experience researchers.

One of the latest breakthroughs to enter Foundry is an energy project code-named Malta, which is an answer to one of the planet’s most existential questions: Can wind and solar energy replace coal? The advent of renewable-energy sources is encouraging, since three-quarters of global carbon emissions come from fossil fuels. But there is no clean, cost-effective, grid-scale technology for storing wind or solar energy for those times when the air is calm or the sky is dark. Malta has found a way to do it using molten salt. In Malta’s system, power from a wind farm would be converted into extremely hot and extremely cold thermal energy. The warmth would be stored in molten salt, while the cold energy (known internally as “coolth”) would live in a chilly liquid. A heat engine would then recombine the warmth and coolth as needed, converting them into electric energy that would be sent back out to the grid. X believes that salt-based thermal storage could be considerably cheaper than any other grid-scale storage technology in the world.

The current team leader is Raj B. Apte, an ebullient entrepreneur and engineer who made his way to X through parc . He compares the project’s recent transition to Foundry to “when you go from a university lab to a start-up with an A-class venture capitalist.” Now that Apte and his team have established that the technology is viable, they need an industry partner to build the first power plant. “When I started Malta, we very quickly decided that somewhere around this point would be the best time to fire me,” Apte told me, laughing. “I’m a display engineer who knows about hetero-doped polysilicon diodes, not a mechanical engineer with a background in power plants.” Apte won’t leave X, though. Instead he will be converted into a member of the Rapid Eval team, where X will store his creative energies until they are deployed to another project.

Thinking about the creation of Foundry, it occurred to me that X is less a moonshot factory than a moonshot studio. Like MGM in the 1940s, it employs a wide array of talent, generates a bunch of ideas, kills the weak ones, nurtures the survivors for years, and brings the most-promising products to audiences—and then keeps as much of the talent around as possible for the next feature.

IV. The Invention

Technology is feral . It takes teamwork to wrangle it and patience to master it, and yet even in the best of circumstances, it runs away. That’s why getting invention right is hard, and getting commercial innovation right is hard, and doing both together—as X hopes to—is practically impossible. That is certainly the lesson from the two ancestors of X: Bell Laboratories and Xerox parc . Bell Labs was the preeminent science organization in the world during the middle of the 20th century. From 1940 to 1970, it gave birth to the solar cell, the laser, and some 9 percent of the nation’s new communications patents. But it never merchandised the vast majority of its inventions. As the research arm of AT&T’s government-sanctioned monopoly, it was legally barred from entering markets outside of telephony.

In the 1970s, just as the golden age at Bell Labs was ending, its intellectual heir was rising in the West. At Xerox parc , now known as just parc , another sundry band of scientists and engineers laid the foundation for personal computing. Just about everything one associates with a modern computer—the mouse, the cursor, applications opening in windows—was pioneered decades ago at parc . But Xerox failed to appreciate the tens of trillions of dollars locked within its breakthroughs. In what is now Silicon Valley lore, it was a 20‑something entrepreneur named Steve Jobs who in 1979 glimpsed parc ’s computer-mouse prototype and realized that, with a bit of tinkering, he could make it an integral part of the desktop computer.

Innovators are typically the heroes of the story of technological progress. After all, their names and logos are the ones in our homes and in our pockets. Inventors are the anonymous geeks whose names lurk in the footnotes (except, perhaps, for rare crossover polymaths such as Thomas Edison and Elon Musk). Given our modern obsession with billion-dollar start-ups and mega-rich entrepreneurs, we have perhaps forgotten the essential role of inventors and scientific invention.

The decline in U.S. productivity growth since the 1970s puzzles economists; potential explanations range from an aging workforce to the rise of new monopolies. But John Fernald, an economist at the Federal Reserve, says we can’t rule out a drought of breakthrough inventions. He points out that the notable exception to the post-1970 decline in productivity occurred from 1995 to 2004, when businesses throughout the economy finally figured out information technology and the internet. “It’s possible that productivity took off, and then slowed down, because we picked all the low-hanging fruit from the information-technology wave,” Fernald told me.

The U.S. economy continues to reap the benefits of IT breakthroughs, some of which are now almost 50 years old. But where will the next brilliant technology shock come from? As total federal R&D spending has declined—from nearly 12 percent of the budget in the 1960s to 4 percent today—some analysts have argued that corporate America has picked up the slack. But public companies don’t really invest in experimental research; their R&D is much more D than R . A 2015 study from Duke University found that since 1980, there has been a “shift away from scientific research by large corporations”—the triumph of short-term innovation over long-term invention.

The decline of scientific research in America has serious implications. In 2015, MIT published a devastating report on the landmark scientific achievements of the previous year, including the first spacecraft landing on a comet, the discovery of the Higgs boson particle, and the creation of the world’s fastest supercomputer. None of these was an American-led accomplishment. The first two were the products of a 10-year European-led consortium. The supercomputer was built in China.

As the MIT researchers pointed out, many of the commercial breakthroughs of the past few years have depended on inventions that occurred decades ago, and most of those were the results of government investment. From 2012 to 2016, the U.S. was the world’s leading oil producer. This was largely thanks to hydraulic fracturing experiments, or fracking, which emerged from federally funded research into drilling technology after the 1970s oil crisis. The recent surge in new cancer drugs and therapies can be traced back to the War on Cancer announced in 1971. But the report pointed to more than a dozen research areas where the United States is falling behind, including robotics, batteries, and synthetic biology. “As competitive pressures have increased, basic research has essentially disappeared from U.S. companies,” the authors wrote.

It is in danger of disappearing from the federal government as well. The White House budget this year proposed cutting funding for the National Institutes of Health, the crown jewel of U.S. biomedical research, by $5.8 billion, or 18 percent. It proposed slashing funding for disease research, wiping out federal climate-change science, and eliminating the Energy Department’s celebrated research division, arpa-e .

The Trump administration’s thesis seems to be that the private sector is better positioned to finance disruptive technology. But this view is ahistorical. Almost every ingredient of the internet age came from government-funded scientists or research labs purposefully detached from the vagaries of the free market. The transistor, the fundamental unit of electronics hardware, was invented at Bell Labs, inside a government-sanctioned monopoly. The first model of the internet was developed at the government’s Advanced Research Projects Agency, now called darpa . In the 1970s, several of the agency’s scientists took their vision of computers connected through a worldwide network to Xerox parc .

“There is still a huge misconception today that big leaps in technology come from companies racing to make money, but they do not,” says Jon Gertner, the author of The Idea Factory , a history of Bell Labs. “Companies are really good at combining existing breakthroughs in ways that consumers like. But the breakthroughs come from patient and curious scientists, not the rush to market.” In this regard, X’s methodical approach to invention, while it might invite sneering from judgmental critics and profit-hungry investors, is one of its most admirable qualities. Its pace and its patience are of another era.

V. The Question, Again

Any successful organization working on highly risky projects has five essential features, according to Teresa Amabile, a professor at Harvard Business School and a co-author of The Progress Principle . The first is “failure value,” a recognition that mistakes are opportunities to learn. The second is psychological safety, the concept so many X employees mentioned. The third is multiple diversities—of backgrounds, perspectives, and cognitive styles. The fourth, and perhaps most complicated, is a focus on refining questions, not just on answers; on routinely stepping back to ask whether the problems the organization is trying to solve are the most important ones. These are features that X has self-consciously built into its culture.

The fifth feature is the only one that X does not control: financial and operational autonomy from corporate headquarters. That leads to an inevitable question: How long will Alphabet support X if X fails to build the next Google?

The co-founders of Google, Brin and Larry Page, clearly have a deep fondness for X. Page once said that one of his childhood heroes was Nikola Tesla, the polymath Serbian American whose experiments paved the way for air-conditioning and remote controls. “He was one of the greatest inventors, but it’s a sad, sad story,” Page said in a 2008 interview. “He couldn’t commercialize anything, he could barely fund his own research. You’d want to be more like Edison … You’ve got to actually get [your invention] into the world; you’ve got to produce, make money doing it.”

Nine years later, this story seems like an ominous critique of X, whose dearth of revenue makes it more like Tesla’s laboratory than Edison’s factory. Indeed, the most common critique of X that I heard from entrepreneurs and academics in the Valley is that the company’s prodigious investment has yet to produce a blockbuster.

Several X experiments have been profitably incorporated into Google already. X’s research into artificial intelligence, nicknamed Brain, is now powering some Google products, like its search and translation software. And an imminent blockbuster may be hiding in plain sight: In May, Morgan Stanley analysts told investors that Waymo, the self-driving-car company that incubated at X for seven years, is worth $70 billion, more than the market cap of Ford or GM. The future of self-driving cars—how they will work, and who exactly will own them—is uncertain. But the global car market generates more than $1 trillion in sales each year, and Waymo’s is perhaps the most advanced autonomous-vehicle technology in the world.

What’s more, X may benefit its parent company in ways that have nothing to do with X’s own profits or losses. Despite its cuddly and inspirational appeal, Google is a mature firm whose 2017 revenue will likely surpass $100 billion. Growing Google’s core business requires salespeople and marketers who perform ordinary tasks, such as selling search terms to insurance companies. There is nothing wrong with these jobs, but they highlight a gap—perhaps widening—between Silicon Valley’s world-changing rhetoric and what most people and companies actually do there.

Explore the November 2017 Issue

Check out more from this issue and find your next story to read.

X sends a corporate signal, both internally and externally, that Page and Brin are still nurturing the idealism with which they founded what is now basically an advertising company. Several business scholars have argued that Google’s domination of the market for search advertising is so complete that it should be treated as a monopoly. In June, the European Union slapped Google with a $2.7 billion antitrust fine for promoting its own shopping sites at the expense of competitors. Alphabet might use the projects at X to argue that it is a benevolent giant willing to spend its surplus on inventions that enrich humanity, much like AT&T did with Bell Labs.

All of that said, X’s soft benefits and theoretical valuations can go only so far; at some point, Alphabet must determine whether X’s theories of failure, experimentation, and invention work in practice. After several days marinating in the company’s idealism, I still wondered whether X’s insistence on moonshots might lead it to miss the modest innovations that typically produce the most-valuable products. I asked Astro Teller a mischievous question: Imagine you are participating in a Rapid Eval session in the mid-1990s, and somebody says she wants to rank every internet page by influence. Would he champion the idea? Teller saw right through me: I was referring to PageRank, the software that grew into Google. He said, “I would like to believe that we would at least go down the path” of exploring a technology like PageRank. But “we might have said no.”

I then asked him to imagine that the year was 2003, and an X employee proposed digitizing college yearbooks. I was referring to Facebook, now Google’s fiercest rival for digital-advertising revenue. Teller said he would be even more likely to reject that pitch. “We don’t go down paths where the hard stuff is marketing, or understanding how people get dates.” He paused. “Obviously there are hard things about what Facebook is doing. But digitizing a yearbook was an observation about connecting people, not a technically hard challenge.”

Recommended Reading

Can Google Succeed as Alphabet?

Is Google Making Us Stupid?

The First Holiday Without a Loved One

X has a dual mandate to solve huge problems and to build the next Google, two goals that Teller considers closely aligned. And yet Facebook grew to rival Google, as a platform for advertising and in financial value, by first achieving a quotidian goal. It was not a moonshot but rather the opposite—a small step, followed by another step, and another.

Insisting on quick products and profits is the modern attitude of innovation that X continues to quietly resist. For better and worse, it is imbued with an appreciation for the long gestation period of new technology.

Technology is a tall tree, John Fernald told me. But planting the seeds of invention and harvesting the fruit of commercial innovation are entirely distinct skills, often mastered by different organizations and separated by many years. “I don’t think of X as a planter or a harvester, actually,” Fernald said. “I think of X as building taller ladders. They reach where others cannot.” Several weeks later, I repeated the line to several X employees. “That’s perfect,” they said. “That’s so perfect.” Nobody knows for sure what, if anything, the employees at X are going to find up on those ladders. But they’re reaching. At least someone is.

To revisit this article, visit My Profile, then View saved stories .

• Backchannel
• Newsletters
• WIRED Insider
• WIRED Consulting

Oliver Franklin-Wallis

Inside X, Google's top-secret moonshot factory

Gandalf arrives on rollerblades. It’s morning in the cafeteria at X – formerly Google X – and Astro Teller, X’s Captain of Moonshots, glides over dressed in coarse grey robes and a pointed hat, carrying oatmeal. Jedi stroll past to their desks, gripping coffee. Star Fleet officers queue for breakfast. This, it should be said, is unusual – it’s Halloween. But X is a surreal place. Outside, self-driving cars loop around the block. Sections of stratospheric balloons, designed to broadcast internet to remote places, hang in the lobby. Robots wheel around, sorting the recycling. Teller likens X to Willy Wonka’s chocolate factory; it seems only fitting that there be costumes.

Even standing inside X – a cavernous former mall in Mountain View, California – it’s hard to articulate exactly what X is. Within Alphabet, Google’s parent company, it is grouped alongside Deepmind in "Other Bets", although in that metaphor, X is more like the gambler. Its stated aim is to pursue what it calls “moonshots” – to try to solve humanity’s great problems by inventing radical new technologies. To that end, besides the self-driving cars (now a standalone company, Waymo) and internet balloons (Loon), X has built delivery drones (Wing), contact lenses that measure glucose in the tears of diabetics (Verily) and technology to store electricity using molten salt (Malta). It has pursued, but ultimately abandoned, attempts to create carbon neutral fuel from seawater, and replace ocean freight with cargo blimps. It once earnestly debated laying a giant copper ring around the North Pole to generate electricity from the Earth’s magnetic field.

That might sound fantastical or even absurd, but every day you almost certainly use something developed at X. Google Brain, the deep-learning division that now informs everything from Google Search to Translate, began at X. So did camera software GCam, used in Google Pixel phones; indoor mapping in Google Maps; and Wear OS, Android’s operating system for wearable devices.

But those are beside the point. “Google Brain, the cars, Verily, everything else – those are symptoms. Side effects of trying weird things, things that are unlikely to work,” Teller says. “We are a creativity organisation, not a technology organisation.” The rollerblades, which he wears every day, are tucked neatly under the table. (They save him eight minutes a day between meetings.) X, he explains, is not so much a company as a radical way of thinking, a method of pursuing technological breakthroughs by taking crazy ideas seriously. X’s job is not to invent new Google products, but to produce the inventions that might form the next Google.

X was once seen as a punchline in Silicon Valley (and on Silicon Valley ). Today, its self-driving cars have logged 10 million miles on public roads, and operate an autonomous ride-sharing service in Arizona. Loon’s balloons provide internet access to communities in rural Peru and Kenya. Wing, X’s drone delivery effort, is carrying food and medicines to customers in Australia. Still, as Alphabet continues to be buffeted by employee protests and leadership changes – in December 2019, founders Larry Page and Sergey Brin stepped down, handing the company to Google CEO Sundar Pichai – X is facing renewed scrutiny to prove that its moonshots are more than just an indulgence, or expensive PR stunts. X celebrates its tenth anniversary in 2020. When will its bets pay off?

Alphabet is not the first company to set up a laboratory for chasing moonshot ideas. In 1925, AT&T and Western Electric founded Bell Labs, which assembled scientists and engineers from different disciplines to advance the field of telecommunications. Bell Labs invented the transistor, the first lasers and photovoltaic cells, winning nine Nobel prizes in the process. Ever since, corporate research labs, from Xerox PARC to Lockheed Martin's Skunk Works and DuPont’s Experimental Station, have played a central role in producing breakthrough inventions. Apple, Facebook, Microsoft and Amazon all have corporate research labs. Google has several, including Google AI (formerly Google Research), Robotics at Google, and Advanced Technologies and Projects, which works on things like AR and smart fabrics.

Amanda Hoover

Andy Greenberg

Mark Andrews

But corporate research labs are flawed. Big companies, chasing quarterly results, often ignore transformative ideas even from within their own organisations. Xerox PARC invented the graphical user interface, but we don’t sit at Xerox laptops. As startups grow into corporations, bureaucracy can take over, and their capacity to think creatively wanes. “Over a 20 to 30 year period, companies tend to move from experimentation to process,” Teller explains. “Process is an attempt to get surprise all the way down to zero. Experimentation is the complete, constant bathing in surprise. You can’t have both.”

X does not call itself a corporate research lab (it uses the term “Moonshot Factory”), but when it was founded in 2010, its remit wasn’t entirely clear. X originally grew out of Chauffeur, Google’s self-driving car project, then spearheaded by the Stanford roboticist Sebastian Thrun. Page and Brin admired Thrun for his work on Streetview and turn-by-turn directions in Google Maps, and at X they offered him free reign to pursue similarly offbeat ideas. “Initially, the title was called ‘Director of Other’,” Thrun says. “We wanted to push technologies in many different directions, including self-driving cars.”

For at least a year, X’s existence was a closely guarded secret. Other Google employees were denied keycard access. Even within Google, where bottom-up management is a founding principle and employees are allowed to spend 20 per cent of their time working on their own ideas, X had a free-wheeling, intellectually anarchic style. Engineers from Project Chauffeur worked alongside those from Google Brain, Loon and a handful of other equally audacious projects. “I wanted to get no bureaucracy, no PowerPoints, no financial reporting, no oversight, so that the people in charge could focus entirely on the challenge,” Thrun says. Most of the early project ideas came from Page and Brin themselves, who took a close interest, and eventually moved into X’s building. (Teller once described X as Brin’s “batcave”.)

When Thrun left X in 2012 for Udacity, his online education company, Teller took over. He was, in many ways, the natural choice. His paternal grandfather is Edward Teller, known as the father of the hydrogen bomb, and the co-founder of the US’s Lawrence Livermore National Laboratory. His maternal grandfather was a Nobel prize-winning economist. “I grew up as the dumb one in my family,” says Teller. “My family believed that being smart was the only thing that mattered. I wasn’t going to win on those terms. As a result, it forced me to explore other ways to be successful.” Before joining X, Astro (a nickname; his real name is Eric) founded an AI hedge fund, and sold a wearable sensors company. He has finished two novels and co-written a book of relationship advice. At school, to compensate for mild dyslexia, he would do every problem twice, using different methods. “Then if I came up with the same answer, it was the right answer,” he says. The experience taught him early the value of experimental thinking – “to just try things fast, to approximate at the beginning, and come at the problem from different angles.”

When Teller took over at X, it had little structure. “I would describe it as like the Wild West. We just started projects when we were interested in things. There was almost no process whatsoever,” says Obi Felten, who joined X from Google in 2012. Teller hired Felten, who then worked in Google’s product marketing department, to formalise the moonshot process. While the engineers pushed the frontiers of artificial neural networks and high-altitude ballooning, Felten says she dealt with “literally everything that wasn’t the tech. Legal policy, marketing, PR, partnerships. They’d never had a business plan for any of the projects before.” Her job title was "Head of Getting Moonshots Ready for Contact With the Real World".

Not all X projects survived first contact. One of the first was Google Glass, a wearable computer inside a pair of spectacles. Brin loved the idea, and pushed X hard to turn the early prototypes into a consumer product. When Glass eventually launched in 2013, Google created huge fanfare. Skydivers wearing Glass parachuted on to the roof of its annual developers' conference. Models wore them on the runway at New York Fashion Week. They were featured on The Simpsons and in Vogue .

But in the real world, Glass faced poor reviews, mockery (“Glassholes”) and outrage at potential invasions of privacy. “The real failure that we had with Glass was when we were trying to talk about it as a learning platform, the public started responding to it like a product,” Teller says. “What was worse is we fell into the trap of talking about it that way ourselves. And that was terrible, because it was not a finished product.”

Glass was discontinued as a consumer product in 2015. It still exists, but as an enterprise tool mostly used in manufacturing and other manual industries. “Sometimes it just doesn’t work, the technology’s not ready, and we have to stop doing it, pause it, slow it down,” Teller says. He still believes that a Glass-like device will catch on eventually. (Apple is reportedly working on AR glasses for debut in 2022.) “There's no way to take moonshots and never be too early. By the definition of what we’re doing, we’re erring on the side of being too early, rather than being too late.”

Once a week or so, X’s smartest minds gather in a conference room and set about systematically killing each other’s craziest ideas. To be considered a moonshot, an idea must fulfil three criteria: it must address a significant global problem, involve inventing breakthrough technology, and result in a radical outcome – at least “10X” better than what exists today. Jetpacks and hoverboards might be fun, but don’t serve a common good.

Similarly, distributing vaccinations is a worthy goal, but isn’t a moonshot. “If most or all of the audacity is just in the scale of the thing, that is not what we’re interested in,” says Teller.

Whatever the problem, no solution is too outlandish. “Everything has to be on the table,” says Teller. “So someone said in a brainstorming session: what if guns actually shot some kind of lethal poison, but there was an antidote in all the jails in the country?” Teller grins. “First of all, that’s a gorgeous idea – I mean, it’s a terrible idea. But in terms of the creativity of the idea, that is good, and that person will think of other really weird ideas that are awesome for society.”

After a moonshot is proposed, X’s Rapid Evaluation team – a rotating cast of Xers with expertise from materials science to artificial intelligence – begin what is known as a pre-mortem. “Let's imagine everything fell apart. What are the things that caused that?” says Phil Watson, who heads the Rapid Evaluation department. Ninety per cent of ideas fail at this stage. Some fail for obvious reasons: too expensive, too difficult. Others break the laws of physics. If an idea can’t be killed easily, then it becomes an investigation, and is assigned a small team to interrogate it further. “We start looking into it more systematically,” Watson explains. “What would it take for this to succeed? What are the skill sets we would need to be able to advance it to the next level? What's the thing most likely to make this fail?” Successful investigations become projects, with a name, budget and full-time staff.

One of the foundational tenets at X is "Monkey First". That is, if you were asked to teach a monkey to stand on a pedestal and recite Shakespeare, you should resist starting with the simplest task (building the pedestal) and start with the hardest (teaching a monkey to speak). Teams are encouraged to set both performance milestones and "kill targets" – thresholds that, if missed, will automatically end the project. For example, Project Foghorn, X’s attempt to turn seawater into fuel, succeeded in producing fuel but couldn’t do so cheaply enough. X killed the project, published its findings as a scientific paper, and gave the team a bonus.

One afternoon I follow Kathryn Zealand, of the Rapid Evaluation team, to see an investigation at work. The idea is to build a pair of assistive trousers that might help the elderly and immobile walk independently. “The space of ageing in general hasn’t had enough investment,” says Zealand, who is Australian. “The demographics mean it’s going to be a big thing in the future.”

The trousers, code-named Smarty Pants, are inspired both by recent advancements in soft robotics and Zealand’s own experience with her 92-year-old grandmother, who has Alzheimer’s. At that age, even simple movements like standing become difficult. “If you help them do that one thing, they take 30 per cent more steps over the course of the day. And the longer people stay walking, the few other health issues they’ll have,” Zealand says. She is wearing what looks like 3D-printed armour on one leg, wired with sensors, which collect data on her gait as we walk.

Early on in any investigation, X begins prototyping. X’s Design Kitchen houses virtually anything required to conduct experiments in numerous fields: a wet lab, milling machines, laser scanners, 3D printers. “We say, OK, what is the quickest experiment we can do that will get us to a yes/no?” Zealand says.

We arrive in a large, airy atrium. Zealand has enrolled her mother, who happens to be visiting, as a test subject. “She struggles with stairs,” Zealand explains. The trousers prototype is rough: actuators on each knee joint, connected to fabric panels around the wearer’s legs. The outer seams are laced, corset-style, giving them a Victorian steampunk vibe. The motors are controlled by a Raspberry Pi in a pearlescent bumbag. Zealand’s team – including a deep learning specialist, a clothing designer and a world expert in biomechanical exoskeletons – fit her mother into the trousers, then monitor her as she climbs several flights of stairs. “It’s amazing,” she says, stepping down, delighted. “Normally I’d be out of breath by the time I got up there.”

Zealand asks me if I want to try them. After a brief fitting, I tread gingerly on the first step, and immediately feel myself being pulled upwards, as if by an additional set of muscles. Climbing is noticeably easier. By using the sensor data and machine learning, Zealand explains, the trousers are learning to "see" the stairs, knowing exactly when to apply force. She hopes that, eventually, soft robotics and material advances might enable a lightweight product a fraction of the weight, with a flexible frame, that could aid a range of mobility issues. “That’s probably 10 years out,” she says. Still, it’s early days. Fewer than half of X’s investigations become Projects. By the time this story is published it will probably have been killed.

The ability to work on such long-term problems is X’s great advantage: the patience of research, without the financial pressures of a startup. “There are some technologies that, because of safety, you have to have several ‘nines’ of reliability before you can even get started,” says Teller. “There’s a really big difference between a 1 per cent error rate and 0.001 per cent error rate.” A software glitch in a mobile app is unlikely to be fatal, but one in a self-driving car might be.

The thought lingers that afternoon, as a driverless Waymo pulls up outside the X campus. Since its start at X in 2009, Waymo has now logged more than 10 million autonomous miles on public roads. For the last year, it has operated Waymo as a small scale ride-hailing service in Phoenix, Arizona, and is currently working with Jaguar on its next generation of vehicles. Morgan Stanley recently valued the company at $105 billion (£80 billion). “Waymo’s goal is to build the world’s most experienced driver,” says Andrew Chatham, a Waymo software engineer. “It is not to build a car. Other people are quite good at that.”

We pull away. A safety driver, Rick, is sitting in the front seat, but the wheel turns itself. Inside the car, a white Chrysler Pacifica, headrest displays show a live view of what the roof-mounted sensors ‘see’: pedestrians in yellow wireframes, the purple outlines of other vehicles. The ride is surprisingly quiet, and, other than some slight hesitations – even now the cars struggle to predict drivers’ intentions at junctions – entirely uneventful.

Even so, mass adoption of self-driving cars is still a long way off. “If I look at where we were in the first six months [back in 2009], we had some really cool videos. And here we are more than a decade later, and those videos… it was too easy to get there,” Chatham says. “Actually getting to the level of deployment is a whole other ball game.”

During the early years, X employees could happily work on technologies that might be decades away, knowing that all the while advertising revenue was flooding into Google. Thrun recalls asking former Google CEO and Alphabet executive chairman Eric Schmidt for $30 million to fund a project. Schmidt gave him $150 million. “Eric said to me, ‘If I give you $30million, you're going to come back next month and ask for another $30million.’”

Then, one morning in 2015, Brin and Page announced that Google was restructuring, to become Alphabet. The news came as a shock inside the company. There were widespread reports of budgets being tightened. But at X, being spun out clarified the team’s essential purpose: “[It] became even more clear that the goal of X is to produce new Alphabet companies,” Felten says.

When projects reach a certain scale, they “graduate” from X to become standalone companies. Most, like Waymo, join Alphabet’s Other Bets. A few have been acquired by Google, or spun out independently, such as the renewable energy startups Dandelion and Malta. Upon graduation, project leaders become executives, and employees are given a stake in the company. “When projects leave here, they're not done,” says Teller. “There's plenty of learning still to be done.”

The transition is not always easy. Since Alphabetisation, the original leaders of several X projects, including Waymo, Loon and Wing, have either left or been replaced. “If you’re trying to accelerate something to grow ambitious and large, the chances that the original person that invented it can take [it through] every stage is pretty hard,” Wendy Tan White, a vice-president at X who oversees growth-stage projects, says. “They would have to grow very fast themselves.”

Alphabet as an organisation has also faced various controversies in its first five years. In 2018, the company was rocked by allegations of sexual misconduct by senior executives; 20,000 employees, including many at X, staged a global walkout in protest. One of the executives named by the New York Times was Richard DeVaul, then head of Rapid Evaluation at X, and one of the original creators of Loon. (DeVaul resigned, reportedly without an exit package.)

Teller has publicly expressed his regret at the episode, and his admiration for those involved in the walkout. “It made me believe in Google and Alphabet more,” he tells me. “Awesome that employees should say, 'This is our company too and that needs to reflect us.’”

There has also been widespread upset at Google’s involvement in Project Maven, a Pentagon artificial intelligence project, and Project Dragonfly, a reported plan to launch a censored search engine in China. (Both have reportedly been shelved.) These events have reignited debates about the responsibility that Alphabet has, in Google’s once famous phrase, to not be evil.

Although these projects did not come under X, Teller says he thinks deeply about the ethics of his team’s work. His grandfather, after all, worked on the Manhattan Project. “There definitely have been things people brought up [at X] that are like, 'Nope, that's evil'. We're just not doing that,” he says. But other issues are less clear-cut – for example, projects that may cause jobs to be lost to automation. One current X project is to create all-purpose “everyday robots” that might automate menial tasks. “New technologies tend to create concentrated harm and diffuse benefits,” Teller says. “If the benefits of automation are 100 times the downsides, that leaves us with 99 per cent positive. But we owe it to those people who have experienced those concentrated problems – and it is a public policy matter – to make sure we take care of them.”

After its flurry of early successes, X’s moonshots in recent years have struggled to strike the popular imagination – or financial success. Of its energy startups, only Malta and Dandelion have to date built a commercial product. Chronicle, its cybersecurity moonshot to create an immune system for the internet, was recently folded back into Google. And while Wing’s drones may end up transforming the logistics industry, it’s hard to paint delivering burritos as a genuine moonshot.

Recently, X has increased its effort on tackling problems that might threaten humanity, like climate change. “Climate change is by any reasonable standard the single biggest problem that humanity has,” Teller says. There are several climate-focused projects in development, including research into ocean health. One of the most advanced, currently untitled, is focused on agriculture. “It’s one of our basic needs. It’s one of the largest industries in the world. It has the largest carbon footprint of any major industry,” Teller says.

In an X workshop on the second floor, engineers are working on several boxy blue vehicles with stilt-like legs and off-road tyres. They are farming drones, designed to comb a field in groups, taking hyperspectral images of the crops and topsoil. The drones are already in testing on some farms in California, “collecting millions of images of plants, where every strawberry has a unique ID”, explains Benoit Schillings, an ebullient Belgian who oversees several moonshots at X. “Agriculture is a huge, complex optimisation problem. Right now the way it’s done is by simplifying the problem: we’re going to put hybrid corn over 10,000 acres,” Schillings says. By analysing the data and making suggestions, X hopes to improve crop yields and soil health.

The agriculture project is a typical X moonshot: take a huge problem and apply Alphabet’s massive advantage in computing power, intellectual expertise and financial resources to try and solve it – creating a global business in the process. “Going after ‘Let's solve agriculture’ – this is pretty hubristic,” Schilling laughs. “We take the challenges that I think very few other players would have the courage to take.”

But it’s also possible to see moonshots as something more cynical: attempts to dominate industries that don’t yet exist. Agriculture, worldwide, is a trillion-dollar market. It’s hard not to picture Waymo as the operating system inside every car, Wing as the air traffic control for every airborne package. Google itself, after all, started as a kind of moonshot, to systematically map all of human knowledge. For all of X’s rhetoric about changing the world, it ultimately exists to produce new enterprises – and profit – for Alphabet.

Teller, however, seems unphased by the notion that creating wildly profitable companies and solving issues like climate change are in opposition. To X, and Alphabet, creating the next Google and saving the world are essentially the same thing. “Things that lose money tend to get smaller over time, and things that make money tend to get bigger over time,” he says. “I see purpose and profit not in opposition. I see them as intensely synergistic.”

X recently celebrated its tenth birthday. The day I visit, the senior executives have been in meetings, trying to map out what the next decade might look like. “The world is changing,” Tan White says. Fields that X once pioneered, such as self-driving cars, are now cottage industries. The concept of moonshots is now used widely by both startups and governments. Massive venture capital funds, such as SoftBank’s Vision Fund, are enabling startups to take moonshot-sized risks of their own.

X’s true impact may not be clear for another decade, or more. While it has created sizeable returns for Alphabet – Teller has said the value of Google Brain alone paid for several years of X’s entire budget – there’s still no telling if its graduated companies will survive, let alone become the next Google. In 2018, Alphabet’s "Other Bets" lost $3.36 billion (£2.57 billion). “We have to accept that some of these businesses still won’t make it,” Teller says. But then, almost all attempts at invention end in failure. Genuine breakthroughs require both immense capital, creativity and, perhaps most important, patience.

“If we define moonshots as trying to create something really radical… that’s really hard. The really big things, it’s tough to see who’s going to do it,” Nathan Myhrvold, the former director of Microsoft Research and founder of Intellectual Ventures, told me. “But the flipside is if you’ve got those resources and you don’t do it, then we’ll never know if there was some fabulous technology that took that kind of effort.”

Sitting in front of Teller-as-Gandalf, it’s hard not to think of an even earlier moonshot factory: the labs of Thomas Edison, known as the Wizard Of Menlo Park. It may be that the self-driving car, or one of X’s many other moonshots, will end up transforming society in ways we can’t yet foresee. It might save the world, or it might just help Alphabet grow ever richer and more powerful.

“The real test is 15 to 20 years from now, when the dust is settled and we look backwards. Then how are we doing?” Teller says. Until then, there will always be more crazy ideas worth chasing. “The world’s got more than enough problems, sadly.”

Updated 17.02.20, 11:50 GMT: This article originally stated that of X’s energy startups, only Malta had built a commercial product. Dandelion has also built a commercial product. Sebastian Thrun also created Udacity, not Coursera

This article was originally published by WIRED UK

Kathy Gilsinan

Rachel Lance

David Kushner

Jenna Scatena

• Documentation

Academic Research

• [[result.title]]
• No roles were found. Try a different search or check back as new roles are posted.

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Research (Wash D C)
• v.2021; 2021

Recent Development in X-Ray Imaging Technology: Future and Challenges

1 MOE Key Laboratory for Analytical Science of Food Safety and Biology, College of Chemistry, Fuzhou University, Fuzhou 350108, China

2 Frontiers Science Center for Flexible Electronics, Xi'an Institute of Flexible Electronics (IFE) and Xi'an Institute of Biomedical Materials & Engineering, Northwestern Polytechnical University, 127 West Youyi Road, Xi'an 710072, China

Xianning Xu

Xiaofeng chen, zhongzhu hong, xiaowang liu, qiushui chen.

3 Fujian Science & Technology Innovation Laboratory for Optoelectronic Information of China, Fuzhou 350108, China

Huanghao Yang

X-ray imaging is a low-cost, powerful technology that has been extensively used in medical diagnosis and industrial nondestructive inspection. The ability of X-rays to penetrate through the body presents great advances for noninvasive imaging of its internal structure. In particular, the technological importance of X-ray imaging has led to the rapid development of high-performance X-ray detectors and the associated imaging applications. Here, we present an overview of the recent development of X-ray imaging-related technologies since the discovery of X-rays in the 1890s and discuss the fundamental mechanism of diverse X-ray imaging instruments, as well as their advantages and disadvantages on X-ray imaging performance. We also highlight various applications of advanced X-ray imaging in a diversity of fields. We further discuss future research directions and challenges in developing advanced next-generation materials that are crucial to the fabrication of flexible, low-dose, high-resolution X-ray imaging detectors.

1. Introduction

X-rays are a type of ionizing radiation with a wavelength ranging from 0.01 to 10 nm [ 1 , 2 ]. When X-rays travel through a matter, they are transmitted, absorbed, or scattered. The processes of scattering and absorption depend on the attenuation ability of the matter and are governed by Lambert-Beer's Law (eq. ( 1 )):

where I is the intensity of transmitted X-ray photons, I o is the initial intensity of X-ray photons, μ is the linear attenuation coefficient, and d is the thickness of the matter [ 3 – 6 ]. The attenuation ability is dominated by a combination of the photoelectric effect, Compton scattering, and Rayleigh scattering [ 7 ]. Their ratios are determined by both the nature of the matter and the energy of incident X-rays. Typically, in a low-energy X-ray region, X-ray photons are mainly absorbed by the object through the photoelectric effect, while the Compton scattering is dominant in low- Z materials and high-energy photons [ 8 , 9 ].

The excellent penetration ability of X-rays has made X-ray imaging a powerful medical imaging modality [ 10 ]. The advances in X-ray imaging have stimulated the progress in diagnostic radiography technologies, physically describing the skeleton, including fractures, luxation, bone disease, and the location of foreign matters [ 11 , 12 ]. Such imaging information is particularly useful for guiding the surgery [ 13 ]. Apart from the medical applications, X-ray imaging is further extensively used for nondestructive industrial and safety inspection [ 14 ]. Undoubtedly, the development of X-ray imaging for over a century has promoted the advancement of a wide range of disciplines from fundamental researches to practical applications.

An X-ray imaging system typically comprises an X-ray generator and an X-ray imaging detector ( Figure 1(a) , left panel) [ 15 ]. The X-ray generator is made of two electrodes sealed into an evacuated chamber. Once powered on, the cathode made of tungsten filament can produce energic electrons through a thermionic effect when it is heated to 2200°C by the electric current. When an accelerating voltage is applied, X-rays are produced during energy changes of fast-moving electrons when they collide and interact with the anode material under a vacuum. The lost energies are converted into bremsstrahlung and characteristic X-rays. Typically, 80% of the X-ray photons emitted by the diagnostic X-ray generator are bremsstrahlung [ 16 , 17 ]. The output X-ray spectrum is affected by accelerating voltage, filament heating voltage and current, and cathode materials.

(a) Schematic illustration of an X-ray imaging system. The system constitutes an X-ray generator and an X-ray detector with a signal processing system. X-ray beam produced by the X-ray generator passes through the object (e.g., patient's chest) to arrive at the X-ray detector, followed by the signal processing to produce a visible image. (b) The development of X-ray radiography with the evolution of X-ray detectors. The development can be mainly divided into film-screen radiography and digital radiography. The film-screen radiography converts a latent X-ray pattern into a visible image through tedious chemical processing, whereas digital radiography goes through a series of signal conversions to obtain the X-ray image. ADC: analog-to-digital conversion; DAC: digital-to-analog conversion; AI: artificial intelligence; ML: machine learning; DB: big data.

The X-ray imaging system converts the X-ray photons transmitted from the object into a visible image that can be used for evaluating the internal structures. The X-ray detector is placed behind objects to record the transmitted X-rays for producing an X-ray pattern ( Figure 1(a) , right panel). This pattern is subsequently converted into a visible two-dimensional (2D) radiographic image or three-dimensional radiographic image through tomography. Finally, the contrast-based X-ray images are generated based on the attenuation difference of the objects within the matter towards X-rays [ 18 – 20 ].

In this review, we give a detailed overview of the recent development of X-ray imaging technologies, including film-screen radiography and digital radiography, according to the evolution of X-ray detectors in the imaging system. In each section, we start with a description of the structure of the device and the corresponding working principle. The advantages and disadvantages of each X-ray imaging system are further discussed. This review is ended with a perspective on the further development direction of X-ray radiography.

2. Film-Screen Radiography

2.1. substrate materials.

The first X-ray image was taken by a radiographic plate several months after the X-rays discovered by Röntgen, where the finger bones and the ring of his wife were clearly imaged [ 24 ]. The radiography manifested its original application in medical diagnosis and was further used for the identification of jewelry and art collection and nondestructive detection of metallic objects in the industry soon. Although photographic plate-based X-ray detectors made of a glass plate coated with a thick layer of light-sensitive emulsion show great promise in radiography, they are fragile, heavy, expensive, and difficult for operation and storage.

The challenges in radiographic plates promoted the development of substitutive substrate materials with flexibility, portability, transparency, and relative thinness. The photographic film consisting of cellulose nitrate and emulsion was first developed to replace the glass plate. Since the cellulose nitrate was flammable, nonflammable cellulose triacetate materials such as polyester materials were used for X-ray film instead [ 25 ].

2.2. X-Ray Film and Cassette

As illustrated in Figure 2(a) , the X-ray cassette has a flat, lightproof metal box consisting of an intensifying screen and a radiographic film. The top protective layer made of opaque carbon fiber shows nearly no radiation absorption. The back layer of the cassette utilizing a thin layer of lead with an atomic number of 82 is designed to avoid potential backscattered radiation from the transmitted X-rays [ 21 ]. The X-ray film consists of the protective layer, emulsion, adhesive, and polymer substrate. The substrate is coated with a thick layer of photosensitive emulsion on both sides to increase the X-ray absorption for mitigating blurring. Typically, the emulsion layer is made of innumerable silver-halide compounds mixed with gelatin material [ 26 , 27 ]. However, the sensitivity of X-ray imaging is very limited when the emulsion is directly exposed to X-rays, and this is largely attributed to its low X-ray absorption efficiency.

Film-screen radiography system. (a) Scheme of an X-ray film cassette and profiles of intensifying screen and X-ray film. The lightweight film cassette made of metal materials features a carbon fiber protective shield, intensifying screens, X-ray film, and a back panel. (b) The comparison of X-ray film direct exposure to X-rays (left) with that of X-ray film in combination with a scintillators screen. (c) X-ray absorption spectra of gadolinium oxysulphide (GOS, red), calcium tungsten (CaWO 4 , orange), and X-ray emission spectra of 60 kV (sky blue) and 100 kV (dark blue) X-rays. The atomic number of rare-earth elements ranges from 57 to 70 with the K -edge between 39 and 61 keV. (d) Upon X-ray irradiation, the free silver ions aggregate in negatively charged sensitivity centers to acquire an electron, forming the latent image region. (e) The latent image in the film was converted into the visible image through photochemical processing including development, fixation, washing, and drying. (f) The characteristic curve of the film-screen system in response to X-ray exposure, which can be divided into three parts including the toe region (blue), the straight-line region (yellow), and the shoulder region (red). Notably, base plus fog is the background intensity of the unexposed film produced by accident light irradiation. The optical intensity is a function of X-ray exposure plotted on a logarithmic scale. (a) is reprinted with permission from ref. [ 21 ], copyright 2013 Author . (c) is reprinted with permission from ref. [ 22 ], copyright 2008 Elsevier Ltd . (f) is reprinted with permission from ref. [ 23 ], copyright 2019 Elsevier Springer Nature Singapore Pte Ltd .

2.3. Intensifying Screen and Its Composition

Although silver halide crystals can be directly exposed by the X-rays, a high dosage of X-rays with a risk of irradiation damage is required for qualified X-ray imaging. To reduce the radiation dose, a fluorescent intensifying screen made of scintillators was introduced for converting X-rays into ultraviolet to visible (UV-Vis) light to sensitize the radiographic film [ 28 ]. Figure 2(b) indicates that a radiographic film coupling with an intensifying screen substantially decreases the X-ray exposure as compared to exposing the radiographic film directly. Of which, 95% of the silver halide crystals are efficiently reduced via the visible light produced by intensifying screen, and the remaining are reduced by the direct interaction with X-rays.

The scintillators act as the energy mediator for intensifying screens, and thus their performance plays a significant role in determining image quality. Over the decades, high-quality scintillators are developed to reduce X-ray exposure. Calcium tungstate (CaWO 4 ), a class of scintillator emitting blue light under X-ray exposure, is utilized for X-ray energy conversion by Thomas Edison thanks to its strong X-ray stopping power and high X-ray scintillation efficiency. However, the absorption coefficiency of CaWO 4 is not optically matched to the spectra at 60-100 keV X-rays, as demonstrated in Figure 2(c) [ 22 ]. The development of scintillators with improved absorption in low X-ray energy is desired. Rare-earth-activated materials with high atomic numbers in the range from 57 to 70 ( K -edges between 39 and 61 keV) exhibit high X-ray attenuation and scintillation efficiency, such as lanthanum bromide oxide, lanthanum oxysulfide, and GOS. A rare-earth-based scintillating screen has a low radiation dosage of about 2-3 times less than the CaWO 4 screen with superior X-ray image quality.

2.4. Chemical Processing for the Radiographic Film

The chemical process to capture an X-ray image using a radiographic film involves the formation of a latent image and then developing an X-ray image [ 29 , 30 ]. Silver-halide crystals have a cubic phase structure with lattice points occupied by negatively charged bromide (or iodide) ions and positively charged silver ions. The silver halides absorb the photon energy of visible light or X-rays and release electrons to form electron-hole pairs, and the released electrons combine with silver ions in the photosensitive center composed of defects (point defects, dislocations, etc.) in the crystals to produce neutral silver atoms. As a result, silver atoms accumulate to form photosensitive spots, thereby forming a latent image ( Figure 2(d) ).

After the X-ray film exposure, it was chemically processed to obtain a visible image that can be displayed by transillumination on an appropriate view box for further evaluation. As shown in Figure 2(e) , this processing involves development, fixation, washing, and drying. [ 31 ]. During a typical development process, electrons from the developer migrate to sensitized grains and convert the silver ions into black silver particles to form a visible image on the film. After leaving the developer solution, the unexposed silver bromide on the film is dissolved and removed in the fixer solution containing acetic acid and sodium thiosulfate. At the same time, sodium sulfite and aluminum chloride in the fixer solution are used as a preservative and a hardener, respectively. Finally, the processed film is washed to remove the fixer solution through a water bath and dried in a chamber in which the hot air is circulating [ 32 ].

2.5. Characteristic Curve of a Radiographic Film

The performance of an X-ray film is strongly related to the radiation exposure on a logarithm scale. Contrast is the difference in luminance or color, making an object distinguishable. For a specific radiographic film, the contrast depends on the design of the film, the amount of exposure, and the chemical processing conditions. As described in Figure 2(f) , there are three different regions in the characterization curve including the toe region (blue), straight-line region (yellow), and shoulder region (red) from the bottom to the top. The toe and shoulder regions with shallow slopes correspond to underexposure and overexposure, respectively. The overexposed image in the shoulder region implies that the silver ions have been reduced to silver atoms, whereas the image will be underexposed and generally useless in the toe region. The normal exposure region is the nearly straight-line portion where a well-exposed image is produced (with a density between 0.5 and 2.75) [ 23 ].

3. Computed Radiography

3.1. substitution of film-screen radiography by computed radiography.

Although conventional film-screen radiography has contributed extraordinarily to medical diagnosis and industrial inspection since 1895, it suffered from several limitations, including complicated chemical processing, low automatic processing efficiency, high costs of film materials, time and labor consumption, inconvenient images storage and communications, and environmental pollution [ 33 – 35 ]. To this end, digital radiography was developed to replace film-screen radiography. This new technology involves using a digital detector to convert X-ray patterns into digital signals which are subsequently processed and displayed on the screen for observation. It mainly comprises imaging acquisition, laser stimulation, electric signal processing, image display, postprocessing, storage, and communication components [ 36 ]. In addition, when compared with film-screen radiography (FSR, blue dotted line), computed radiography shows an improved linear exposure range (10 4  : 1), suggesting a wide range of radiation exposure ( Figure 3(a) ) [ 37 , 38 ].

Computed radiography and its imaging mechanism. (a) The characteristic exposure curve of film-screen radiography (FSR, blue dotted line) and computed radiography (CR, red line). The film-screen radiography shows a linear exposure range of 10 : 1, and the digital radiography shows a linear exposure of 10 4  : 1. (b) Schematic diagram showing a typical computed radiography reader system and the corresponding image readout process. (c) Schematic diagram of the cross-section of the imaging plate (I). Scanning electron microscope (SEM) image of the structured (III) and unstructured (II) phosphors. (d) X-ray absorption spectra of thallium-doped cesium iodide (CsI: Tl, green), terbium-doped gadolinium oxysulphide (GOS: Tb, orange), and europium-doped barium fluobromide (BaFBr: Eu 2+ , blue) as a function of X-ray photon energy. (e) The physical process of photostimulation using BaFBr: Eu 2+ phosphors. It can be divided into two steps, including radiation storage (light yellow) and photostimulated luminescence (light blue). The X-rays penetrating the object are absorbed by phosphors, creating a lot of electron-hole pairs, which subsequently migrate to emitting centers or are captured by metastable energy traps. Electrons and holes in the metastable energy traps absorb low-energy laser irradiation to overcome the energy barrier, escaping from the traps, followed by recombination at emitting centers to generate photostimulated luminescence. (a, b) are reprinted with permission from ref. [ 38 ], copyright 2007 Elsevier Ltd . (d) is reprinted with permission from ref. [ 41 ], copyright 2007 American College of Radiology .

3.2. Image Readout Process of Computed Radiography

Computed radiography, firstly introduced by Fujifilm in 1983, is a technology on the basis of recording the latent image in a photostimulable phosphor-contained imaging plate through laser-light stimulation [ 39 , 40 ]. A computed radiography system mainly comprises two components, including an imaging plate and a computed radiography reader. They are designed to store the latent image of the X-ray attenuation pattern in the imaging plate and to read out the stored latent image through the reader, respectively. On a separate note, the computed radiography reader (point-scan, laser flying spot) consists of a set of subcomponents, such as the stimulating laser source, reflecting mirror, light collection guide, and photomultiplier tubes (PMT) [ 38 ].

During a computed radiographic imaging process, an X-ray attenuation pattern transmitted from the object is stored in photostimulable phosphors embedded into the imaging plate, leaving a latent image [ 42 ]. Then, a laser raster scanning can be used to read out the stored imaging information through releasing the photostimulated luminescence using photomultiplier tubes. Thereafter, in situ generated luminescence signals were converted to electric signals for generating high-quality images by an analog-to-digital converter ( Figure 3(b) ). The imaging plate can be repeatedly used by removing the residual energy within the phosphors through intense laser light [ 43 ]. However, the residual energy in the imaging plate cannot be completely erased since it is hard to release all the trapped energy in phosphors by a laser scanning. It is essential to extend the erasure time and increase the erasure cycle to eliminate all the residual energy for further use.

3.3. The Composition of the Imaging Plate and the Property of Phosphors

In computed radiography, an imaging plate is used to replace the intensifying screen and photographic film. As shown in Figure 3(c) , the protective layer on both sides prevents the imaging plate from being scratched, ensuring the durability of the imaging plate and allowing laser transmission. The phosphors layer, which can store the latent image, is made of phosphors mixed with a polymer binder. The electroconductive layer prevents the image quality from degrading by static electricity. The support layer in the middle endowed the imaging plate with a certain mechanical strength. The backscatter radiation is blocked by the light shield layer with a lead backing.

Regarding the phosphors within the imaging plate, there are three prerequisites: first, the emission of the phosphors is required to overlap with the maximum quantum efficiency wavelength of the photomultiplier; second, the irradiated phosphors should exhibit a fast response to the laser scanning for fast imaging; third, no significant signal deterioration for at least 8 h is required for practical use. However, there are almost no phosphors that can simultaneously satisfy the above three aspects at the same time. Among them, BaFX: Eu 2+ ( X = Cl, Br, or I) phosphor family has been extensively studied. Although RbBr: Tl + and CsBr: Eu 2+ can also be used as phosphors in imaging plates benefitting from their easy preparation in the form of a needle-like structure array, the quick latent image loses (tens of seconds) limit their further use for computed radiography systems ( Figure 3(c) , (I)).

The commercial materials for the phosphor layer are BaFBr: Eu 2+ and CsBr: Eu 2+ ( Figure 3(c) , (II)). Trace amounts of Eu 2+ activators are doped to replace Ba 2+ ions in the crystal to form the luminescent centers. Such a doping treatment can alter the structure and consequently the physical properties of the photostimulated phosphors [ 44 , 45 ]. As compared with the rare-earth-based materials used in the screen system, BaFBr: Eu 2+ shows efficient X-ray absorption in the range from 35 to 50 keV because of low K -edge absorption of barium, as presented in Figure 3(d) . Beyond this range, either GOS:Tb phosphors or CsI: Tl phosphors display better performance, allowing their widespread application in indirect flat-panel X-ray detectors or optically coupled digital radiography systems [ 41 , 46 ].

3.4. The Mechanism of Photostimulated Luminescence

The possible energy transfer mechanism inside the photostimulated phosphors is illustrated in Figure 3(e) . The in situ generated electron-hole pair concentration within phosphors is proportional to the absorbed radiation energy of the host lattice. In addition, X-ray patterns transmitted from the object could interact with halide ions to displace them into interstitial host sites, thus creating halide ion vacancies and interstitials. Electrons and holes are captured by traps, leading to the formation of latent images. Subsequently, the electrons and holes can spontaneously escape from the traps at ambient conditions, resulting in the gradual deterioration of the storing energies. When the scanning laser light is applied, the carriers trapped in the defects absorb enough energy from stimulation light to overcome the energy barrier, moving freely in the crystal until the occurrence of recombination to release their energy to luminescence centers (e.g., Eu 2+ ) accompanied by emitting the light-stimulated luminescence. At last, the carriers still trapped should be effectively excited to empty residual energy to prevent the generation of a ghost image in the next use [ 47 , 48 ].

4. Flat-Panel Detector-Based Radiography

4.1. the origin of flat-panel-based digital radiography.

With the advancement of photolithography and microelectronic fabrication technology, large-area, flat-panel-based digital radiography was developed in the early 1990s [ 49 ]. Digital radiography technology converts the incident X-ray photons into electrical charges and reads the images using photoelectric conversion arrays, displaying a faster readout time than computed radiography [ 50 ]. Low-dose, real-time X-ray imaging using flat-panel detectors has been widely used for clinical diagnosis, including chest X-rays, dental X-rays, mammography, and lumbar spine X-rays. Digital radiography is also used in industrial inline nondestructive inspection, such as high-resolution analysis of circuit boards for solder joint porosity measurements and defects detection. Moreover, digital radiography has been widely used in X-ray security scanners in train stations and airports for the screening of dangerous goods and prohibited items.

The charge-coupled device-based detector appeared in 1990 was the first large-area flat-panel-based radiography. A charge-coupled device is made of metal-oxide-semiconductor capacitors as a light-sensitive sensor for recording images. In general, a large number of charge-coupled devices are coupled to create a detector array for large-area detection. The incident X-ray photons can be converted into visible luminescence by scintillators (e.g., CsI: Tl, and GOS: Tb). Next, the luminescence is directed to the charge-coupled device array using an optical lens system ( Figure 4(a) ) [ 51 , 52 ]. However, the optical lens system can reduce the number of photons reaching the charge-coupled device arrays, which may result in low quantum efficiency and high image noise, and thus lead to poor image quality. Meanwhile, the optical coupling may also cause geometric distortions and light scattering and consequently a reduced imaging spatial resolution. Besides, high-working temperatures give rise to signal noise within the charge-coupled device itself, deteriorating the image quality. Although the electric cooling charge-coupled device could alleviate this effect, it has an unacceptably high cost. In addition, the size limitation of the charge-coupled device and the optical coupling method make it rigid to fabricate a large-area X-ray detector [ 53 ].

Flat-panel-based digitized radiography and the technical factors influencing imaging quality. (a) Schematic illustration of an optical len-coupled indirect conversion digitized radiography system based on a charge-coupled device. The incident X-rays are converted into UV-Vis light by the scintillators and further into electric signals after being focused by an optical lens and directed to the charge-coupled device array. (b) The schematic illustration of the internal construction of a flat-panel detector (middle panel), which could be classified into indirect conversion flat-panel detector (left panel) and direct conversion flat-panel detector (right panel) based on the X-ray energy conversion modality. For indirect conversion, X-rays transmitting through the scintillator (purple) are converted into UV-Vis light, which is further converted into an electrical charge by the pixelated amorphous silicon photodiodes ( α -Si; violet), whereas X-ray photons are directly converted into electrical charge in a direct conversion detector. TFT: thin-film transistor. (c) The schematic diagrams (left panel) and line spread function (right panel) of the unstructured and structured scintillators. The X-ray-induced visible luminescence in the unstructured scintillators exhibits a severe scattering in all directions to reduce the imaging spatial resolution, resulting in a wide line spread function. The structured scintillators consist of phosphors in a needle-like structure in favor of reducing the lateral scattering of light, contributing to a narrow line spread function. (d) Comparison of the modulation transfer function (MTF) for direct conversion flat-panel detector (red) and indirect conversion flat-panel detector (blue). (e) Relationships between image quality parameters, including detective quantum efficiency (DQE), modulation transfer function (MTF), signal-to-noise ratio (SNR), and Wiener spectra, and physical image measurements, including contrast, resolution, and noise. Panel (a) is reprinted with permission from ref. [ 51 ], copyright 2007 RSNA . Panel (c) is reprinted with permission from ref. [ 56 ], copyright 2011 American Institute of Physics . Panel (e) is reprinted with permission from ref. [ 57 ], copyright 2008 Elsevier Ltd .

4.2. Evolution of Thin-Film Transistor Array-Based Digital Radiography

By contrast, flat-panel detectors with large-area photoelectric arrays allow the integration with an X-ray energy conversion layer and thin-film transistor (TFT) array-based electronic readout layer [ 54 ]. Unlike charge-coupled devices with coupling optical lenses systems, TFT-based flat-panel X-ray detector is capable of achieving low-dose, real-time X-ray imaging through coupling an energy transfer layer and large-area pixelated TFT arrays ( Figure 4(b) , middle panel)), becoming popular for applications in angiography, radiography, and mammography. According to the difference in the pathway of converting X-ray radiation to charge carriers, flat-panel X-ray detectors are categorized into indirect conversion systems and direct conversion systems [ 55 ].

4.2.1. Direct Conversion X-Ray Detector

Direct conversion X-ray flat-panel detector is fabricated by depositing a layer of X-ray-sensitized materials onto pixelated TFT arrays capable of directly converting X-ray photons into electrical charges that allow being transferred to thin-film transistors ( Figure 4(b) , right panel) [ 58 ]. The most commonly used photoconductor material is amorphous selenium ( α -Se) fabricated by evaporation at high temperatures [ 59 ]. Upon X-ray irradiation, the α -Se photoconductor can absorb the X-ray energy and convert it into charge carriers which are proportional to the incident X-ray photons. The hole-electron pairs generated in the photoconductor travel along the field lines parallelly with limited lateral diffusion because of the electric field applied in the α -Se. Holes can be collected by the positive bias electrode, whereas electrons can be collected by collection electrodes. The charges are stored on the storage capacitor and then are subsequently read out by thin-film transistors. Each pixel is effectively separated by the field-shaping in the α -Se layer, contributing to a high-quality X-ray image [ 60 ].

4.2.2. Indirect Conversion X-Ray Detector

Indirect conversion flat-panel X-ray detector is made of a layer of scintillator thin-film on the top for X-ray energy conversion, pixelated amorphous silicon ( α -Si) photodiode arrays adjacent to scintillators, and a TFT array ( Figure 4(b) , left panel) [ 61 ]. When X-ray irradiates the flat-panel X-ray detectors, X-ray photons are converted into visible luminescence by scintillators and subsequently converted into electric charges by the α -Si photodiode arrays. Eventually, the electric charges are recorded by a TFT array [ 62 ].

The most widely used scintillators are CsI: Tl with a thickness of 150-600  μ m and terbium-doped GOS: Tb [ 63 , 64 ]. The scintillators deposited in indirect flat-panel X-ray detectors can be either unstructured or structured thin-film layers. For the unstructured scintillators, such as GOS: Tb powder crystals (turbid phosphors), the emitted light traveling in the materials may spread to the neighboring pixels, resulting in a reduced spatial resolution. This matter could be overcome by utilizing structure scintillators, like CsI: Tl consisting of discrete and parallel “needles” with 5-10  μ m wide [ 65 ]. In this case, the X-ray-excited luminescence only travels along with the fiber-like crystal to the photodiodes, which improves the spatial resolution, making unstructured scintillators superior to that achieved by the structured scintillators, as illustrated in Figure 4(c) [ 56 , 66 , 67 ].

4.3. Primary Physical Parameters of X-Ray Imaging

A high-quality digital radiographic image is important for accurate testing and diagnosis. The X-ray imaging quality can be evaluated by three primary parameters, including spatial resolution, contrast, and noise ( Figure 4(e) ). The physical parameters are generally evaluated by measurements of Wiener spectra (WS), modulation transfer function (MTF), and signal-to-noise ratio (SNR) [ 57 ].

4.3.1. Spatial Resolution

The ability to distinguish adjacent details in an object and its related sharpness can be defined by spatial resolution. For digital systems, the spatial resolution relates to the pixel size in the matrix, which is crucial to achieving a high spatial resolution for digital X-ray imaging [ 68 ]. This parameter could be measured using a narrow slit, a sharp-edged object, and a bar test pattern. In most cases, a line spread function is used for narrow slit imaging. For instance, α -Se-based direct conversion flat-panel X-ray detectors exhibit better imaging spatial resolution than that of indirect conversion flat-panel detectors since the former has nearly no light scattering.

4.3.2. Contrast

The contrast is another key parameter used for evaluating X-ray imaging quality. It refers to the relative brightness of two positions in an X-ray image by measuring the characteristic exposure curve of an X-ray imaging system. For producing a useful image, the contrast is described by a dynamic range of an X-ray detector in response to various X-ray dose exposure. When compared with screen-film radiography, digital radiography exhibits a much wider and linear dynamic range, reducing the risk of overexposure or underexposure. Moreover, the differences between specific tissues (e.g., bones and soft tissue) could be reflected in one image through post-processing without further exposure [ 69 ].

4.3.3. Noise

The noise signals originating from various sources (e.g. collection element, coupling element, capture element, etc.) are characterized by the variations of signals in an X-ray image of a uniform object [ 70 ]. The noise of an X-ray detector is important for determining image quality. The factor of Wiener spectra (WS) is used to measure the noise variation of an X-ray image, indicative of the functional relationship between spatial frequency and the corresponding noise.

4.3.4. Modulation Transfer Function (MTF)

The spatial resolution of an X-ray imaging detector can be measured by the MTF. More specifically, the MTF is used to convert the values of object contrast into contrast intensity levels of an X-ray image [ 71 ]. As mentioned above, due to the limited lateral scattering, the MTF for direct conversion flat-panel X-ray detectors is obviously higher than that measured by the typical indirect conversion flat-panel X-ray detectors, as presented in Figure 4(d) .

4.3.5. Detective Quantum Efficiency (DQE)

DQE is currently used as the standard measurement to evaluate image quality in radiography and assess the efficiency of an X-ray imaging detector in detecting X-ray photons [ 72 ]. Remarkably, the DQE takes into consideration the signal-to-noise ratio (SNR) and the system noise. The DQE indicates the performance of the X-ray imaging detector in terms of X-ray imaging quality and the X-ray radiation dose. The DQE for digital radiography is higher than that for conventional screen-film radiography, indicating that digital radiography can convert a higher proportion of incident radiation into image signals compared to conventional screen-film radiography. In particular, the DQE for CsI: Tl-based indirect flat-panel detector could reach 40-45% at 0.5 lp/mm, while that for computed radiography is generally less than 30% at 0.5 lp/mm.

5. Computed Tomography (CT)

5.1. the development of three-dimensional (3d) radiography.

Regarding projection radiography, a large proportion of the depth information is lost since all structural details from a 3D object are projected on a 2D plane X-ray detector, producing an overlapped radiographic image, which will lead to misinterpretation of the internal structures. Fortunately, a new technique named CT was developed to overcome this limitation in the 1970s [ 73 ]. As shown in Figure 5(a) , series of projection images are acquired from various angles to generate tomographic images. A 3D image is then obtained by reconstructing these tomographic images using computer algorithms [ 74 , 75 ]. Compared with projection radiography, CT can provide comprehensive 3D anatomical reconstructions and has a greater diagnostic capability.

Demonstration of CT. (a) Schematics illustrating the working principle of CT scanning and imaging. The collimated X-ray photons penetrating through the object are recorded using an X-ray detector, which is positioned opposite to the X-ray generator. In a typical CT scanning, the solid-state detectors rotate around the object in synchrony with an X-ray generator to produce a series of 2D projection images. Subsequently, the 2D slice images are obtained after reconstruction with filtered back projection. Eventually, a 3D tomographic image is reconstructed through computer algorithms. (b) The structure of dual-energy CT (i) and multienergy CT (ii). The dual-energy CT or multienergy CT is equipped with two- or multi-X-ray generators, which allow simultaneous acquisition of images under two- or multienergy level X-rays in a single scan. (c) The development history of CT. The evolution of CT has gone through seven generations. (d) Combined diagnosis of CT and an artificial intelligence algorithm deep convolutional neural network (CNN) on four sectioned chest images. CT images are used as input data for the CNN model; subsequently, the output images (right panel) are presented as the heat maps, where red indicates a high risk of COVID-19 infection. (e) The reconstructed 3D porous structure of six coal samples. The pore size distribution, pore volume, porosity, and permeability data could be obtained. (f) The 3D X-ray images showing the volume fraction of organelles (bottom panel) and the nuclear membrane (top panel). L: lysosomes; M: mitochondria; ER: endoplasmic reticulum; V: vesicles; E: external. (d) is reprinted with permission from ref. [ 76 ], copyright 2020 The Author(s), under exclusive license to Springer Nature America, Inc . (f) is reprinted with permission from ref. [ 77 ], copyright 2010 Nature America, Inc .

The first applicable CT scanner, consisting of an X-ray generator and two collimated sodium iodide crystals-photomultiplier detectors, was invented by Godfrey N. Hounsfield in 1968. Hounsfield was awarded the Nobel Prize for his contribution to CT. From then on, seven generations of the CT have developed, including updating the shape of X-ray source from pencil to cone, increasing the number of imaging slices and detectors, and changing the scanning mode from rotation and translation to helical scanning ( Figure 5(c) ).

The first CT scanner uses a rotate/translate system equipped with an X-ray generator with a pinhole collimator to produce the collimated X-rays (i); the second-generation CT scanner incorporates an X-ray generator, which could produce a narrow, fan-shaped X-ray beam, and increases the X-ray sensor number (ii); the third-generation scanner involves a fan-shaped X-ray beam with an angle ranging between 40 and 60 degrees, which enable scanning the object in a rotated modality (iii); the fourth-generation CT system employs a rotating X-ray tube and a stationary, closed X-ray detector ring to alleviate the ring artifacts produced by the third generation (iv); the fifth-generation CT scanner is composed of no moving parts, and the electron beam is directed around the target ring, allowing for all stationary instrumentation (v); the addition of a slip ring stimulated the development of six-generation CT (vi); the seventh-generation CT scanner consists of a multiple detector array and a cone-shaped X-ray beam (vii).

More importantly, dual- and multienergy CT was allowed to be constructed by equipped dual- and multi-X-ray tubes, permitting to operate at different tube voltages to make dual and multienergy scanning possible ( Figure 5(b) ). The merits of dual- and multienergy CT lie in the fact that data sets at two different photon spectra can be obtained simultaneously upon a single scanning. Furthermore, the dual-energy algorithm can increase the contrast of bone, which is powerful to directly visualize the iodinated vessels without interference. As a result, dual-energy CT is widely used in angiography to create a virtual noncontrast image.

5.2. Application of 3D Radiography

As is presented in Figure 5(d) , medical CT scanner is extensively used to screen the size, types, location, and numbers of pulmonary nodules, which could offer an accurate assessment of the risk for further treatment. Recent studies showed that CT is valuable for the COVID-19 diagnosis [ 76 ]. CT helps to obtain the pathophysiology characters of COVID-19 infected person, such as consolidations of the lungs and bilateral/peripheral ground-glass opacities. These data provide the most intuitive and precise diagnosis information, thereby greatly enhancing diagnostic efficiency.

Apart from the medical application, CT technology is also introduced to industrial nondestructive inspection in early 1980. Industrial CT is a promising nondestructive tool for characterizing the flaws, inclusions, cracks, and insufficient fusion within the body of materials [ 78 – 80 ]. For instance, the 3D pore structure of coal samples can be obtained through CT to reproduce the precise distribution of coal pore and pore structure ( Figure 5(e) ). The heterogeneity inside different coal samples can be directly observed, providing insights into the structure-dependent attributes of coal, including gas transport, thermomechanical, and failure behaviors.

As an added benefit, the tomography technology is able to perform 3D imaging of nano- to microsized biological organisms when coupling with X-ray microscopy [ 81 – 83 ]. This combined technology enables the diffraction limit of the conventional microscope to be overcome because of using shorter wavelength X-ray photons. This feature suggests its power in elucidating the detailed structural information of in vivo or ex vivo biological samples with a cellular resolution. Note that an emerging transmission soft X-ray microscope could generate 3D cell imaging at a nanoscale resolution based on the difference in X-ray absorption between organic matter and water, filling the gap between cryoelectron tomography and fluorescence superresolution microscopy ( Figure 5(f) ) [ 77 ].

6. X-Ray Microscopy

6.1. the development of x-ray microscopy.

Optical microscopy is great of significance to study microstructures [ 84 ]. Fluorescence microscopy provides an approach to image the structures at a microscale resolution by taking advantage of site-specific fluorescence labeling. However, the imaging resolution of fluorescence microscopy is largely limited by the wavelength of UV-Vis light, as confined by Abbe or Rayleigh laws [ 85 , 86 ]. The imaging resolution can be significantly enhanced to a few angstroms using an electron beam as the incident light in transmission electron microscopy [ 87 ]. This technology shows considerable disadvantages in the observation of biological samples, especially considering the tedious sample preparation process including dehydration, formalin fixation, paraffin-embedding, and section. Besides, the poor penetration depth of electrons in biological samples has a limitation to imaging the sample thickness larger than 100 nm.

By taking advantage of the powerful penetration and nearly no scattering properties of X-rays, the emerging X-ray microscopy techniques break the penetration depth limitation of transmission electron microscopy and allow the intact sample to be imaged without specimen sectioning. The wavelength of X-ray locates at a range of 0.01-10 nm, which is suitable to be used as a light source for imaging biological objects at a very high spatial resolution. It is worth noting that the penetration ability of soft X-rays is much greater than that of electrons. Meanwhile, the water is nearly transparent to X-rays compared to organic compounds at X-ray energy located at 284–540 eV (water window), where the K absorption edges of carbon and oxygen are 284 eV and 540 eV, respectively ( Figure 6(b) ). Therefore, the development of X-ray microscopy is very useful for imaging biological specimens with improved spatial resolution under wet and normal pressure conditions ( Figure 6(a) ) [ 88 ]. Besides, the X-ray energies of 10-100 keV cover the spectroscopic features of all elements and offer the opportunity to detect elements and probe chemical bonds of an object [ 89 , 90 ].

X-ray microscopy setups and their optics. (a) X-ray and optical images of a fibroblast. The area outlined with yellow dashed in the low-magnification image (left panel) is shown at a higher magnification image by the means of X-ray microscopy (middle panel) and optical microscopy (right panel). (b) The penetration distances of X-rays and electrons in water and protein as a function of their energy. The lines from left to right represent attenuation lengths (1/ μ ) of carbon (protein) and oxygen (H 2 O) for X-rays and the mean free paths ( λ ) of H 2 O (elastic scattering), protein (elastic scattering), H 2 O (inelastic scattering), and protein (inelastic scattering), respectively. (c) A Fresnel zone plate is made of several transparent and opaque concentric circulars with radially increasing line density. The central absorbing region is responsible for suppressing the strong zero-order diffraction. (d) One-dimensional multilayer Laue lens for hard X-ray focusing. Alternating layers are fabricated by the thin-film deposition technique to implement thin thickness and ultra-high aspect ratio. (e) A Kirkpatrick-Baez mirror for hard X-ray focusing. Multilayer coatings were designed to increase the angles of operation and to perform photon energy selection. (f) Compound refractive lenses for X-ray focusing at a range of 5-40 keV. A linear array of lenses is manufactured by high-density low-atomic number materials. (g) Full-field transmission X-ray microscopy. A full-field image was projected by a microzone plate onto the X-ray detector. (h) Scanning transmission X-ray microscopy. A zone plate is used to focus coherent X-rays on the sample, whereas an X-ray-sensitive detector is used to capture X-ray images. The sample is mounted on a stage having stepping or piezoelectric driven motors to perform the raster scan. (a, b) are reprinted with permission from ref. [ 88 ], copyright 1995 Cambridge University Press . (c)–(f) are reprinted with permission from ref. [ 94 ], copyright 2010 Macmillan Publishers Limited .

6.2. The Optics in X-Ray Microscopy

In the late 1940s, the invention of grazing incidence mirror optics offers a great opportunity to develop X-ray microscopy. However, the technical issues of long exposure time and insufficient spatial resolution become a major challenge for the use of X-ray microscopy. In the 1970s, the development of high-quality zone plates for high-energy X-ray focusing opens the modern era of X-ray microscopy. The high-quality X-ray focusing optics are then extensively used to increase the spatial resolution of X-ray imaging. Nowadays, X-ray optics can be well designed by combination with thin-film deposition, electron beam lithography, and nanofabrication with capabilities to improve diffractive, reflective, and refractive X-rays [ 91 ]. The Fresnel zone plates consist of several concentric rings of transparent zones and alternating opaque, as shown in Figure 6(c) [ 92 ]. The X-rays passing through the transparent sections are diffracted and subsequently generate constructive interference, focusing on a small spot. Hence, the zone plates can be used both as condenser and objective for X-ray focusing. The zone plate-based microscopy is achievable for high-resolution imaging, which is largely determined by the zone plate's outer width (Δ r N ). The smaller outer zone width can lead to a higher spatial resolution. The selection of the type of zone plate is determined by several factors including photon energy, required spatial resolution, and the number of zones. At present, a 12 nm spatial resolution has been successfully performed using the 12 nm zone plate [ 93 ].

The efficiency and resolution of hard-X-ray focusing are also achieved using a multilayer Laue lens with varied d -spacing, a multilayer coating-based 1D zone plate fabricated by magnetron sputtering [ 95 ]. As shown in Figure 6(d) , the multilayer Laue lens are tiled to meet the Bragg condition for the outer, smallest layer spacing, providing efficiency larger than conventional Fresnel zone plates. Multilayer Laue lens of a 16 nm width was used to focus 20 keV photon energy of X-rays. Recently, a focal spot size smaller than 10 nm was achieved by fabricating multilayer Laue lenses with sufficiently high numerical aperture. Besides, for 2D focusing, the two multilayer Laue lenses with different focal lengths are required to be positioned orthogonal to each other.

Reflective optics are further developed to achieve the imaging resolution of several nanometers by exploiting Kirkpatrick-Baez systems ( Figure 6(e) ). Grazing incidence reflective mirrors are capable of focusing hard X-rays and enhancing X-ray reflection efficiency. Kirkpatrick-Baez mirrors are typically made from multilayers of dense metals or hard silicon carbide coated on silicon crystals with near atomic roughness. The efficiency of the Kirkpatrick-Baez mirror is constrained by the shape and surface roughness.

Since the refractive indices of all materials for X-rays are always slightly less than ones in vacuum and air, conventional optical refractive lenses are not available for X-ray focusing. In addition, an appropriate curvature radius and a double concave shape are essential for X-ray focusing lenses, as illustrated in Figure 6(f) . In 1996, compound refractive lenses were designed using parabolic concave lenses [ 96 ]. To reduce X-ray absorption and increase compound refractive lenses' efficiency, compound refractive lenses are typically made of high-density, low- Z materials, such as lithium, boron, silicon, carbon, beryllium, or aluminum. Advantages in simple manufacture, low cost, small size, easy alignment, and tunable focal length make the compound refractive lenses great promise in hard X-ray focusing.

6.3. General X-Ray Microscopy Modes

Current state-of-art X-ray microscopy includes full-view transmission X-ray microscopy and scanning transmission X-ray microscopy [ 97 , 98 ]. The full-view transmission X-ray microscopy is similar in principle to that of optical bright-field microscopy. X-rays travel through the focusing optics to irradiate the sample, and the transmitted X-rays are magnified by a zone plate to provide a magnified projection onto the detector [ 94 ]. As shown in Figure 6(g) , the central stopper coupled with an order sorting aperture is applied to filtrate a certain portion of X-rays which is not in the first-order diffraction. The sample is placed near the spot of the first-order diffraction. The X-rays penetrating out from the samples are magnified using the microzone plates as an objective, projecting onto the X-ray detector. The imaging resolution of the full-view transmission X-ray microscopy depends on the outer zone width (Δ r N ) of the zone plate, imaging geometry, and illumination coherence. Since the quick acquisition of a 2D projection image, a 3D image is reconstructed by many 2D projection images from different angles of the sample.

The scanning transmission X-ray microscopy is another technology suitable for imaging the local scale structure. Figure 6(h) shows the schematic setup of the scanning transmission X-ray microscopy. A coherent part of X-rays from a monochromator passes through the zone plate to produce a diffraction-limited focal spot, and then the transmitted X-rays are detected. As a result, scanning transmission X-ray microscopy image is reconstructed since the sample is scanned in 2D perpendicular to the optical axis. The imaging resolution is determined by several factors, including the quality of focusing lenses, the precision of instrumental setup, and the coherence of X-rays. The imaging resolution can be improved by a higher-order focusing zone plate, while the detection efficiency will be reduced. One key advantage of scanning transmission X-ray microscopy is its easy extension to multisignal, simultaneous detection in combination with X-ray scattering, diffraction, fluorescence, or electron emission yield [ 99 ].

7. Material Opportunity for X-Ray Imaging

The rapid development in materials science offers a great opportunity to revolutionize the future of X-ray imaging technology. Over the past decades, scintillator materials, which can convert high-energy radiation into UV-Vis photons, are critical to high-performance X-ray imaging. In the early stage, CaWO 4 and ZnS powders were widely used for X-rays imaging. After the 1940s, scintillator crystals (e.g. NaI: Tl, CsI: Tl, and Bi 4 Ge 3 O 12 ) were gradually used for fabricating high-performance X-ray detectors such as commercial flat-panel detectors. However, conventional scintillators are synthesized through a solid-state method at high temperatures, resulting in large crystals that are unsuitable for manufacturing large-area, flexible X-ray detectors.

Recently, solution-processed materials have been developed for advancing next-generation X-ray imaging technologies with low cost, high sensitivity, and flexibility. In particular, perovskites, featuring tunable bandgap, high photoluminescence quantum yields, narrow emission, and high charge-carrier mobility, have emerged as promising materials in photovoltaic devices, luminescence displays, and radiation detection [ 100 – 103 ]. The heavy atom-contained perovskites with efficient X-ray absorption show great potential in X-ray imaging applications. Lead-halide perovskite nanocrystals can generate multicolor radioluminescence upon X-ray irradiation [ 104 ]. The solution-processable and easily scalable CsPbBr 3 nanocrystals are synthesized to fabricate large-area flat-panel X-ray detectors ( Figure 7(a) ). In addition, CsPbBr 3 nanosheets synthesized at room temperature can be assembled into a uniform and dense thin film as an X-ray scintillating screen for high-resolution radiography [ 105 ]. Although indirect conversion-based X-ray detectors are most popular in practical applications, they generally suffer from a relatively low spatial resolution due to the optical crosstalk among neighboring pixels.

Perovskites-based X-ray detectors. (a) Schematic illustration of the perovskite nanocrystals-based flat-panel detector by coating a layer of CsPbBr 3 nanoscintillators onto a commercial pixelated α -silicon thin-film-transistor (TFT) panel. (b) Schematic of a flexible perovskite X-ray detector (left) and I-V curves of the flexible device under X-ray irradiation (right). (c) Scheme illustrating the fabrication of Si-integrated MAPbBr 3 single crystals. (d) Diagram of an X-ray detector based on 2D RP perovskite p-i-n thin-film. (e) Photography of a bulk (NH 4 ) 3 Bi 2 I 9 single crystal (left) and X-ray sensitivity measurement of (NH 4 ) 3 Bi 2 I 9 single-crystal device in the direction parallel and perpendicular to the (001) plane (right). (a) is reprinted with permission from ref. [ 104 ], copyright 2018 Springer Nature Limited . (b) is reprinted with permission from ref. [ 20 ], 2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim . (c) is reprinted with permission from ref. [ 60 ], copyright 2017 Macmillan Publishers Limited . (d) is reprinted with permission from ref. [ 107 ], copyright 2020 American Association for the Advancement of Science . (e) is reprinted with permission from ref. [ 110 ], copyright 2019 Springer Nature Limited .

X-ray imaging detectors based on direct conversion present the advantages of a high signal-to-noise ratio and high resolution since the X-ray-generated charges can be directly collected by pixelated arrays [ 106 ]. Liu et al. fabricated a flexible X-ray detector using solution-processable perovskite nanocrystals by an inexpensive inkjet printing method ( Figure 7(b) ) [ 20 ]. Tsai et al. put forward an ultrasensitive X-ray detector by fabricating Ruddlesden-Popper (RP) layered perovskites in a fully depleted p-i-n architecture ( Figure 7(d) ) [ 107 ]. In 2015, Yakunin et al. reported that perovskite crystals of methylammonium lead iodide (MAPbI 3 ) were developed to achieve indirect X-ray detection with strong X-ray absorption and high sensitivity [ 108 ]. Compared to the commercial direct conversion X-ray detectors using amorphous selenide as a photoconductor, perovskites which feature low-cost, defect-tolerance, solution-processibility, and tunable bandgap hold great promise for X-ray imaging, which is presented in Figure 7(c) [ 60 , 109 ]. Despite their great progress, lead-halide perovskites suffer from the issues of poor long-term stability, and the toxicity of lead composition is harmful to the environment and human health. The diversity in substitution strategies offers the structural and functional flexibility to synthesize lead-free perovskites ( Figure 7(e) ) [ 110 ]. As such, 2D and zero-dimensional perovskites are further developed for achieving X-ray detection through tailoring ionic radius, chemical composition, and coordination environment based on the classical structure of ABX 3 perovskites [ 111 ]. Recent studies have shown that many heavy atom-contained double perovskites have merits of efficient X-ray absorption, short decay time, and high stability, ideal for X-ray imaging [ 106 , 112 , 113 ].

Another focus of recent research is developing flexible X-ray detectors that are applicable to 3D X-ray imaging of irregularly shaped objects. Very recently, lanthanide-doped fluoride materials prepared by wet chemical methods were developed for high-resolution, flexible X-ray luminescence extension imaging. These materials prolonged radioluminescence and X-ray memory after the stoppage of the X-ray source, making it possible to fabricate flexible X-ray detectors [ 114 ]. After rational surface coating, the persistent luminescence intensity was enhanced by 6.5-fold, suggesting that the surface passivation can efficiently block the pathway of energy quenching by defects on the surface. The X-ray energy trapping capability and solution processibility allow fabricating the flexible X-ray detectors through embedding the nanoscintillators into the soft substrate, which is promising for portable X-ray devices, point-of-care radiography, and nondestructive testing in special conditions [ 115 ].

Apart from the inorganic scintillators, metal-free organic scintillators display great potential in large-area and flexible X-ray detectors, by taking advantage of flexibility, solution-processability, transparency, and ease to large-area fabrication. To date, the scientific community mainly focuses on developing lanthanide-doped materials, perovskites, and metal organic frames [ 116 ]. Considering that organic scintillators composed of carbon, hydrogen, oxygen, and nitrogen elements show a relatively low X-ray attenuation coefficient, the radioluminescence of organic scintillators can be brightened by introducing heavy atoms (such as chlorine, bromine, and iodine) to turn on the triplet excitons [ 117 ]. Overall, the emerging advanced materials present opportunities for promoting X-ray imaging technology with low-dose, high-resolution, and portability, and the performance of X-ray imaging can be improved in the terms of device physics, materials, and manufacturing methods.

8. Conclusion and Perspectives

X-ray imaging technology has been rapidly developed for various applications since 1895, offering new opportunities to scientific and industrial communities. Considering the fundamental and technical advances of X-ray detectors, we have summarized various X-ray working mechanisms that are crucial for specialized applications. The contrast-based X-ray imaging using a screen-film scintillation screen is a classical technique that greatly advances noninvasive medical imaging. The emergence of computed radiography has led to the technological evolution for digital X-ray imaging with more precise and instant information, while its separated readout mechanism suffers from technical limitations such as a high radiation dose and nondynamic imaging. Since the pioneering study in the 1990s, flat-panel X-ray detectors have been most prominent for achieving real-time digital radiography, which is popularly used in hospitals and industries in place of traditional computed radiography. In further development, CT integrating advanced helical scanning techniques and image reconstruction techniques is capable of providing comprehensive 3D structure information, which is a well-established cardiac, pectoral, and encephalopathic imaging modality with widespread acceptance and application.

Despite great efforts and tremendous achievements made in the past decades, the field of X-ray imaging is still in search of low-dose, high-resolution, large-area, flexible X-ray detectors. A low radiation dose used for X-ray imaging is an important technical consideration that people are always pursuing. One important aspect is to search advanced X-ray energy converting materials, which are critical for achieving efficient X-ray scintillating to increase the sensitivity of X-ray detectors. To date, the most efficient scintillators are limited to bulk CsI: Tl and GOS: Tb phosphors, while suffering from the drawbacks such as harmful scintillation decay, harsh synthesis process, and unsatisfied light yields. Another challenge that X-ray imaging faces are achieving a high spatial resolution for practical radiography due to the optical crosstalk on the transistor and the low sensitivity of the X-ray detectors. The combination of a high-efficiency X-ray converting layer and metasurface technology may be a promising strategy. In addition to the general considerations described above, one of the tremendous interesting directions is to develop large-area and flexible X-ray imaging detectors for potential applications in dental X-ray inspections, imaging of irregular objects, portable X-ray testing, and so on. The recent development outlined in this review is expected to stimulate future investigations for next-generation X-ray imaging technologies.

Acknowledgments

This work was supported by the National Key Research & Development Program of China (2020YFA0709900, 2020YFA0210800), the National Natural Science Foundation of China (21635002, 62134003, 22027805, 21705025, 22077101, 22104016), the Major Project of Science and Technology of Fujian Province (2020HZ06006), the Joint Research Funds of Department of Science & Technology of Shaanxi Province and Northwestern Polytechnical University (2020GXLH-Z-008, 2020GXLH-Z-021), Natural Science Foundation of Ningbo (202003N4065), Key Research and Development Program of Shaanxi (2020ZDLGY13-04), China-Sweden Joint Mobility Project (51811530018), the Special Funded Project of China Postdoctoral Science Foundation (2021T140117), Fundamental Research Funds for the Central Universities, and Fujian Science & Technology Innovation Laboratory for Optoelectronic Information of China (2021ZZ128).

Conflicts of Interest

The authors declare that there is no conflict of interest regarding the publication of this article.

Authors' Contributions

Xiangyu Ou, Xue Chen, and Xianing Xu contributed equally to this work.

A serendipitous discovery and the choreographed dance of fragile X research

The choreography of development is a delicate dance. Beginning in utero, chromosomes, DNA, genes and RNA twirl, tap, and sashay their way in a precise pattern. A misstep or a missing step that changes the routine causes body and brain functions to go awry – as is the case with many intellectual and developmental disabilities (IDD). Fragile X syndrome is the most common known single-gene cause of inherited IDDs, including autism. Scientists know the misstep in this syndrome is in the gene FMR1. FMR1 is responsible for making the protein FMRP, which is necessary for typical brain development.

Lynne Maquat, Ph.D. , founding director of the Center for RNA Biology at the University of Rochester , and professor of Biochemistry & Biophysics , Oncology , and Pediatrics , did not set out to study fragile X. It was through another line of research – her seminal discovery of and decades’ worth of work on nonsense-mediated mRNA decay (NMD) – that fragile X syndrome entered her radar. NMD is a cellular quality-control mechanism that plays a role in both healthy and disease states, and her lab discovered that it is overactive in people with fragile X.

“It was complete serendipity,” Maquat said. “No one ever thought to look at NMD and fragile X. So now we’re trying to figure out what happens at the molecular level when FMRP is absent; we want to understand the network of altered gene expression by identifying mis-regulated messenger RNAs (mRNAs).”

“It is key to finding any kind of hope for a future therapy.” Christoph Pröschel, Ph.D., associate professor of Biomedical Genetics

One of the most prominent surveillance systems in the body that protects against mistakes in gene expression that lead to disease, NMD is a complex pathway that is at the heart of many of the collaborations between Maquat and other University of Rochester scientists. Together, with funding from the National Institutes of Health (NIH) and the FRAXA Research Foundation, they aim to gain a deeper understanding of the sophisticated mechanisms related to NMD that will contribute to developing new drug therapies for genetic disorders such as fragile X syndrome, cystic fibrosis, and hundreds of others.

INTO THE BRAIN

Associate professor of Biomedical Genetics Christoph Pröschel, Ph.D. , has spent much of his career interested in neurogenetic diseases that primarily affect the white matter of the brain, which carries signals throughout the organ. His lab started working with induced pluripotent stem cells (iPSCs) to understand different neural cell types, providing a solid foundation for their IDD research. “My lab and the Maquat lab have a mutual interest in the molecular mechanism of fragile X,” said Proschel. “It is key to finding any kind of hope for a future therapy.”

The Pröschel lab makes and differentiates neural stem cells that mimic fragile X syndrome, allowing his team to test hypotheses and understand how different therapies impact cell biology and function. He and Tatsuaki Kurosaki, Ph.D. , research assistant professor in the Maquat lab, used these neural stem cells to understand the relationship between FMRP and NMD. They discovered that NMD controls the amounts of messenger RNAs deriving from a wide range of genes throughout the brain, including genes that govern motor control and cognitive processes related to attention, learning, and language. They also found that when FMRP is absent from cells, as it is in people with fragile X syndrome, NMD shifts into overdrive.

This work was part of a 2021 study published in Nature Cell Biology led by Maquat that revealed that tamping down NMD with small molecule inhibitors restored a large proportion of neurological functions in these cells.

Most recently, Pröschel co-authored research published in Molecular Cell led by Maquat and co-authored by Hana Cho, Ph.D., and Elizabeth Abshire, Ph.D. , of her lab. The study highlighted a complex molecular dance between NMD and the enzyme AKT, which plays a key role in cell growth and survival. Both AKT and NMD are overactive in fragile X. Using neural stem cells that lack the FMRP protein, they tested a drug called Afuresertib, which inhibits AKT. They discovered that blocking AKT in the fragile X cells decreased its activity and decreased NMD. These cells then acted more like typical, non-disease cells.

There is still a lot the team doesn’t know about how AKT and NMD interact, because they both influence and regulate multiple activities in cells, but this work provides good direction that could inform the development of future treatments for fragile X syndrome.

“This has been one of the real fun chapters of my career – working with this group,” said Pröschel. “Everyone brings such a different perspective to the project.”

FROM SURGERY TO THE LAB

As a neurotologist (subspecialist of Otolaryngology ), Hitomi Sakano, M.D., Ph.D. , spends time in the clinic with patients with hearing issues or hearing loss. In the lab, she aims to understand how the brain adapts to sound information.

Her work with fragile X syndrome began as a resident at the University of Washington when she took interest in FMRP, which is highly expressed in the auditory brainstem nuclei of a typical brain and is the same protein missing in fragile X patients. When Sakano came to the Medical Center in 2018, she brought the fragile X mouse model to study this and joined the Center for RNA Biology.

“I also use the [knockout] mouse model to study hyperacusis – extreme sensitivity to sound,” said Sakano. “We know that fragile X patients have sensory and auditory sensitivity, so this model is a great tool to study both.” Hyperacusis is also very common in the general population (some report up to 15 percent) so understanding the mechanism could potentially impact our broader community.

Sakano hypothesizes that FMRP regulates genes that enable neuroplasticity to maintain

From left: Hitomi Sakano, M.D., Ph.D., and Lynne Maquat, Ph.D.

normal processing of auditory information. If true, there may be therapeutic targets for symptoms like auditory hypersensitivity in fragile X. Funding from the Schmitt Program in Integrative Neuroscience (SPIN) through the Del Monte Institute for Neuroscience Pilot Program and a NIH Research Career Development Award for clinician-scientists are supporting her research, which involves investigating the gene expression abnormalities in the auditory brainstem of the fragile X mouse model that might explain the auditory hypersensitivity in these mice. To date, she has found some interesting RNAs that encode synaptic proteins. These findings open up the possibility of targeting these genes for the treatment of hyperacusis.

UR designated - Intellectual and Developmental Disabilities Research Center

In 2020, the National Institute of Child Health and Human Development (NICHD) designated the University of Rochester as an Intellectual and Developmental Disabilities Research Center. This recognition acknowledged the Medical Center’s national leadership in research for conditions such as autism, Batten disease, Rett syndrome, and most recently fragile X syndrome. The Center’s researchers, including Maquat, Pröschel, Sakano, Telias, and Brima, work to translate scientific insights into new ways to diagnose and treat these conditions.

She co-authored a study with the Maquat and Pröschel labs in Genome Biology . The research used the mouse model whose FMR1 gene is knocked-out. These findings build upon Maquat’s previous research that showed NMD hyperactivation in neuronally induced stem cells from fragile X patients. This hyperactivity negatively impacts many neuronal mRNAs important to brain development. The Genome Biology paper showed NMD goes into overdrive in the brain during early development in a mouse with fragile X. These researchers are now testing various therapeutics to inhibit NMD.

“Being able to collaborate to gain meaningful results to move this science forward is the value of being at an academic medical center like Rochester,” said Sakano. “These steps are what will ultimately lead to treatments and therapies that I use in the clinic someday to help my patients.”

ON THE HORIZON

Forthcoming research aims to broaden the scope of the fragile X work at the Medical Center. One of the world’s largest clinics for fragile X is in Israel, where an estimated 80 percent of women are screened for the inherited disease. Michael Telias, Ph.D. , assistant professor of Ophthalmology , Neuroscience , and Center for Visual Science , began studying fragile X as a graduate student in Israel. He uses human embryonic stem cells that carry the mutation for fragile X to look inside neurons at the molecular and cellular levels to shed light on the human-specific mechanisms affected by this syndrome.

From left: Christoph Pröschel, Ph.D., and Michael Telias, Ph.D.

“Human neurons have shown us that these cells have a problem receiving information and communicating information to the next cell,” said Telias. “We cannot do this work in humans, so using human cells enables us to know what to target in the cell. That is the only way we will be able to develop treatments that work.”

In the Frederick J. and Marion A. Schindler Cognitive Neurophysiology Laboratory , research assistant professor Tufikameni Brima, Ph.D. , is aiming to use electroencephalography (EEG) and event-related potentials (ERP) to better understand how the brains of patients with fragile X respond to various stimuli. This work has the potential to build upon the ongoing molecular research being conducted by Telias and others.

“Ultimately, what we are figuring out is what happens when FMRP is absent. We don’t know the whole story,” Maquat said. “However, FMRP is an RNA-binding protein, and in work soon to be published in Molecular Cell, Kurosaki and I have now defined those messenger RNAs that are normally bound by FMRP and how the absence of FMRP binding results in those mRNAs making too much protein. These results have allowed us to identify which genes are affected and how. Our work will pave the way for better therapeutics for those living with fragile X.”

• Del Monte Institute for Neuroscience
• students and trainees
• featured story

Welcome to Bio-X

• Bio-X History
• Map & Directions
• Dining Options
• Room Scheduling
• General Facilities Issues
• Urgent Facilities Issues
• Non-Emergency Facilities Requests
• Building Access Request
• Shared Equipment
• Diversity, Equity, Inclusion
• Browse Seed Grants
• Visiting Scholars/Visiting Postdocs
• PhD Fellows
• Undergraduate Research
• NeuroVentures
• Travel Awards
• Research Partners
• Browse Videos
• Browse All Research
• Clark Center @ 10x Video
• Bio-X in the News
• USRP Faculty Talks
• Symposium Lectures
• Additional Videos
• Upcoming Events
• Talk Videos
• Stanford Bio-X Frontiers in Interdisciplinary Biosciences: 2019/2020
• Courses & Workshops
• Alumni & Friends
• Partnership Models
• Benefits of Partnership
• Corporate Member Projects
• Corporate Forum Newsletter Archive
• Stanford Bio-X White Paper

Stanford Bio-X Funded Research

Stanford Bio-X supports research and educational opportunities that cross disciplines between the biological or biomedical sciences and fields of engineering, physics and computational science. Encouraging research that bridges disciplines brings together different ways of approaching questions and has resulted in innovations and breakthroughs in our understanding of the human body. Altogether, we have over 1,300 affiliated faculty and more than 5,000 students and postdoctoral fellows, representing more than 70 departments from across Stanford University.

Browse Research

Opportunities for research include:

• the biannual Interdisciplinary Initiatives Seed Grant Program for collaborations among affiliated faculty
• Visiting Scholars/Visiting Postdocs to bring high-level early career researchers to Stanford
• PhD Fellowships open to Stanford graduate students
• the Undergraduate Summer Research Program for Stanford undergraduate students
• Bio-X Ventures to fund large-scale endeavors and resources
• Travel Awards to support PhD candidates and postdoctoral fellows giving oral presentations at scientific conferences

Please also explore the exciting Research Partners with laboratories located at the Clark Center.  

• Seed Grants

• Tel: +1 718 618 4545
• [email protected]

Market Reports

Market research reports include comprehensive and in-depth analysis of market size, segmentation and competitor.

Our research consulting team helps clients to identify and develop strategic growth opportunities supported by data & insight

Competitive Intelligence

Get company profiles with in-depth financial data, SWOT analysis, successful strategies and associated news, innovation and deal.

Market Research Reports

Industries catered.

Life Sciences

Aerospace & Defense

Consumer Goods

Mechanical & Chemical

Food and Beverages

Market research reports industries.

Construction & Manufacturing

ICT & Media

Our Business Complete

Delivering actionable insights.

Industry Analyzed

Country Coverage

Companies Tracking

From the publication, our latest publication.

Global Sports Apparel Market

Rising health awareness and increasing fitness activities among consumers is adding the growth of sports apparel market.

Global Prefilled Syringes Market

Rising prevalence of chronic and lifestyle diseases, growing pharmaceutical industry and benefits of prefilled syringes

Global Confectionery Market

Availability of wide range of confectionery products coupled with expansion of retail market such as hypermarkets, super

LoRa and LoRaWAN Devices Market

Live Cell Imaging Market

Liquid crystalline polymers market, digital ooh market, air duct cleaning chemicals market, what our customer say, customer’s words, head of marketing.

(Environmental Services Company)

Vice President - BDE

(Security And Aerospace Company)

Strategy Manager

(Leading Global Bathroom Space Developer)

Subscribe our newsletter

Deals of the week, popular categories.

Public Address and Voice Alarm Systems M

Project Portfolio Management Market

Penetration Testing Market

Hyperconnectivity Market

Betting Market

Multimodal User Interface (UI) Market

Music Publishing Market

Wi-Fi 6 Market

Electric Coolant Pump Market

Pressure Washer Market

Light Gauge Steel Framing System Market

Vibration Control System Market

Street Cleaning Machine Market

Commercial & Industrial Refrigeration Ma

Plowing and Cultivating Machinery Market

Blended Cement Market

Dibber Market

Potato Planter Market

Please complete the form below, we will then pass on your request to our team for processing.

• Application Season

Your Complete Application Season Tool Set

Hunt Research Tools, Draw Odds, Tag Trends, Application Reminders, and Expert Knowledge—All in One Spot

Not an onX Member? Click Here.

Research and Planning With onX Hunt

Doing the groundwork for Application Season has never been easier thanks to our new onX Hunt Research Tools and our partners Huntin’ Fool and HuntReminder. FREE for onX Hunt Elite Members, this trio of tools brings you a complete tool set to help you find and apply for tags, then dig into the research needed to make your hunt a success.

Own Your Odds This Application Season

Applying for big game tags out West is a complex, time-consuming, and sometimes intimidating process. Whether you grew up hunting the West or have always dreamed of coming out for the hunt of a lifetime, application season can seem like a maze of data, math, and annoyingly-complex projects.

With the help of our partners at Huntin’ Fool and HuntReminder, teamed with our new onX Hunt Research Tools, doing the groundwork for Application Season is easier than ever before. As an onX Hunt Elite Member you get FREE access to Huntin’ Fool’s individualized assistance as well as access to countless hours of in-depth, in-the-field knowledge acquired by the Huntin’ Fool community each season.

onX Hunt Research Tools helps provide a personalized experience catering to your specific draw odds for each state. Once you set up your profile and punch in the hunt you want to apply for in 12 Western states (more to come in the future), we do the work to find the best possible outcome for you to successfully draw a tag. Dive into how the unit appears on the map and what previous harvest rates look like, so you can form a plan to punch that tag.

You can’t draw a tag if you forget to apply. HuntReminder is a simple, effective service designed to help hunters set text message and email reminders to apply for their chosen tags. If you tend to procrastinate on your applications until the final minute or need help keeping track of the different tags you need to apply for, HuntReminder is a critical tool to round out your application season strategy.

You drew the tag… now the next stage begins. Before you go on the hunt, utilize some of the best Elite-only features to prepare for success in the field. From pro deals of many top brands, Imagery on Demand of where your hunt will take place, and live updated imagery so you can see your hunting unit before you go, and so much more, onX Hunt Elite helps set you up for success.

Select a State

From game management units in Alaska to deer zones in Florida, onX has gathered a complete list of species districts, management units and hunting zones in every state. Whether you’re exploring a new hunting area or checking the boundaries of a species-specific district, you can scout, plan and hunt with confidence knowing onX has the most detailed land data available. Click a state to dive into its hunting details.

A neural codec language model for speech synthesis

With the help of discrete neural audio codecs, large language models (LLM) have increasingly been recognized as a promising methodology for zero-shot Text-to-Speech (TTS) synthesis. However, sampling based decoding strategies bring astonishing diversity to generation, but also pose robustness issues such as typos, omissions and repetition. In addition, the high sampling rate of audio also brings huge computational overhead to the inference process of autoregression. To address these issues, we propose VALL-E R, a robust and efficient zero-shot TTS system, building upon the foundation of VALL-E. Specifically, we introduce a phoneme monotonic alignment strategy to strengthen the connection between phonemes and acoustic sequence, ensuring a more precise alignment by constraining the acoustic tokens to match their associated phonemes. Furthermore, we employ a merge codec approach to downsample the discrete codes in shallow quantization layer, thereby accelerating the decoding speed while preserving the high quality of speech output. Benefiting from these strategies, VALL-E R obtains controllablity over phonemes and demonstrates its strong robustness by approaching the WER of ground truth in experimental results. In addition, it requires fewer autoregressive steps during inference, resulting in over 60% time savings in inference time.

This page is for  research demonstration purposes  only.

Model Overview

The overview of VALL-E R, a robust and efficient neural codec language model for zero-shot TTS. It incorporates phoneme information (green) when predict audio codec (blue), which can enhance the connection between phoneme and audio to improve the robustness of decoder-only transformer TTS model. Note that VALL-E R achieves faster inference speeds by adopting compact codec codes, derived from the proposed merge codec method, within its autoregressive model.

Audio Samples

• Follow on Twitter
• Like on Facebook
• Follow on LinkedIn
• Subscribe on Youtube
• Follow on Instagram
• Subscribe to our RSS feed

Share this page:

• Share on Twitter
• Share on Facebook
• Share on LinkedIn
• Share on Reddit

Numbers, Facts and Trends Shaping Your World

Read our research on:

Full Topic List

Regions & Countries

• Publications
• Our Methods
• Short Reads
• Tools & Resources

Read Our Research On:

5 things to keep in mind when you hear about Gen Z, Millennials, Boomers and other generations

Generation Z. Millennials. Baby Boomers. It’s hard not to run into eye-catching headlines about generations these days. And it’s easy to feel like many of these headlines are just clickbait, all fluff and no substance. But is that really the case?

At Pew Research Center, we think it can be useful to talk about generations. But there are some important considerations for readers to keep in mind whenever they come across a news story or research about generations:

Generational categories are not scientifically defined. The boundaries that place one person in Gen Z and another in the Millennial generation are not precise, definitive or universally agreed on. Even the names of generations are not uniformly adopted: Is it Millennials or Generation Y ? Gen Z or iGen ?

People born near the boundaries of these generational groupings can feel particularly uncomfortable being lumped in with those much older or younger than them, and for good reason. The media and researchers – Pew Research Center included – have not always been as clear as we should that generational boundaries are not a hard science.

Generational labels can lead to stereotypes and oversimplification. All Millennials or Baby Boomers are not the same, just as all Southerners, all Catholics or all Black Americans are not the same. Shared experiences and identities should be recognized – and at their best can even be empowering – but this shouldn’t come at the expense of individuality.

Discussions about generation often focus on differences instead of similarities. Conflict tends to get more attention than consensus. So watch out for news stories or research articles that assume or exaggerate intergenerational divides that may actually be quite small. “ OK Boomer ” became a cultural meme, but it probably overstates the divide between younger and older generations. After all, most of us have some combination of parents, grandparents, kids and grandkids we love, making our family lives interconnected.

Conventional views of generations can carry an upper-class bias. Popular history recalls that Baby Boomers in the 1960s and ’70s were deeply opposed to the Vietnam War. This notion is based on attention-grabbing protests on college campuses and at political events. But many high-quality surveys at the time showed that younger Americans – most of whom were not attending college – were more supportive of the war than older generations who had lived through previous conflicts. Readers today should similarly question whether stereotypes of Gen Z might be skewed toward the experiences of the upper middle class.

People change over time. It’s worth pausing when you hear someone say that “kids today” are so different from their predecessors. Young adults have always faced a different environment than their parents, and it’s common for their elders to express some degree of concern or alarm. (“Why is his hair dyed green?”)

Don’t assume that what you see today is what you’ll get tomorrow. People change as they grow older, pursue careers and form families. Gen Zers will no doubt walk differently in the world by 2050, just as today’s Baby Boomers are different from their younger selves. Generational signals can sometimes be lasting, but youth itself is not a permanent state.

So is it all just hype?

If you’ve read this far, your suspicions about generational labels may have hardened. That’s OK. Our recommendation is for readers to bring a healthy dose of skepticism to the generational discussions they see. Readers should also hold media and research organizations that focus on generations – including Pew Research Center – to a high standard.

Despite these cautions, we still believe generational thinking can help us understand how societies change over time. The eras in which we come of age can leave a signature of common experiences and perspectives. Events such as terrorist attacks, wars, recessions and pandemics can shape the opportunities and mindsets of those most affected by them.

Similarly, historical advances like desegregation, effective birth control, the invention of the internet and the arrival of artificial intelligence can fundamentally change how people live their lives, and the youngest generations are often in the vanguard. At the same time, some events can affect people across generations, moving everyone in one direction or another.

It’s wise to think of terms like Gen Z, Millennial, Gen X and Baby Boomer as general reference points instead of scientific facts. At Pew Research Center, we’ll continue to use these and other labels to help our readers navigate a changing world. But we’ll do so sparingly – and only when the data supports the use of the generational lens .

• Age & Generations
• Demographic Research
• Generation X
• Generation Z
• Generations
• Greatest Generation
• Methodological Research
• Millennials
• Silent Generation

Michael Dimock is the president of Pew Research Center

How Teens and Parents Approach Screen Time

Who are you the art and science of measuring identity, u.s. centenarian population is projected to quadruple over the next 30 years, older workers are growing in number and earning higher wages, teens, social media and technology 2023, most popular.

1615 L St. NW, Suite 800 Washington, DC 20036 USA (+1) 202-419-4300 | Main (+1) 202-857-8562 | Fax (+1) 202-419-4372 |  Media Inquiries

Research Topics

• Coronavirus (COVID-19)
• Economy & Work
• Family & Relationships
• Gender & LGBTQ
• Immigration & Migration
• International Affairs
• Internet & Technology
• News Habits & Media
• Non-U.S. Governments
• Other Topics
• Politics & Policy
• Race & Ethnicity
• Email Newsletters

ABOUT PEW RESEARCH CENTER  Pew Research Center is a nonpartisan fact tank that informs the public about the issues, attitudes and trends shaping the world. It conducts public opinion polling, demographic research, media content analysis and other empirical social science research. Pew Research Center does not take policy positions. It is a subsidiary of  The Pew Charitable Trusts .

Copyright 2024 Pew Research Center

Terms & Conditions

Privacy Policy

Cookie Settings

Reprints, Permissions & Use Policy

Texas Tech Now

Highlighting the impact of texas tech’s innovation hub at research park.

April 23, 2024

The Hub plays a significant role in entrepreneur Abigail Rosilier’s company.

On the corner of Fourth Street and Texas Tech Parkway is a 44,000-square-foot pillar of light for West Texas entrepreneurs. Texas Tech University 's Innovation Hub at Research Park signifies that the future is more attainable today than ever before. 

The Hub has 15 entrepreneurial programs to assist startups, including seed funding. These are tools that aid new entrepreneurs of the Hub to outline progress and obtain their goals. 

In 2023, the Innovation Hub startup companies raised $8 million in capital investments and were awarded 68 new trademarks, patents, and/or copyrights. The Hub has 60 industry mentors supporting startups with over 7,000 volunteer mentor hours logged last year.

“We're all about economic development -- we want students at Texas Tech to learn here, grow their company here and stay in West Texas,” said Taysha Williams managing director at the Innovation Hub. 

The Hub has multiple startup initiatives such as its Texas Tech Accelerator Program that supports entrepreneurs launching innovative startups. This is a year-long program that includes $25,000 in funding, mentors, monthly business bootcamps and additional resources. The Hub is a space for collaboration and ideation, providing entrepreneurial minds with all the resources and connections needed to be successful.

The Accelerator program is on its seventh cohort of entrepreneurs, comprising eight startups. One of those is abbyrose , a phone case company started by current Jerry S. Rawls College of Business student Abigail Rosilier. She is set to graduate in May with a bachelor's in business management and several years of business experience.  

Being a college student is difficult enough, but Rosilier has managed to juggle school and a successful business throughout her college career. 

“I struggle with that. I would love to feel like I have it all down, but I don't,” Rosilier said humbly. “It's definitely hard, but I find the more I have on my plate the more scheduled I become. Some advice I would give is make your schedule and hold yourself accountable. You'll find a groove.”

Rosilier always has had an entrepreneurial spirit. She had a few stints with small businesses, beginning in the fifth grade, that she'd start and end often within a few months. 

Rosilier began abbyrose as a 16-year-old in her parents' garage during her junior year of high school. She transitioned to homeschooling due to the demanding training schedule of competing in acrobatic gymnastics. She credits acro with shaping her into the person she is today, beginning the sport at age 5. Turns out acro gymnastics also helped with abbyrose because of the connections it afforded her.

“I had so many connections and friends from different states because I was constantly competing,” Rosilier said. “That was a lot of help when I first started abbyrose because I started selling to people in Maryland and California. It was good to have a ton of friends in different states.”

Rosilier explained she had a phone case business in years prior to starting abbyrose, but like before, ended it within a few months. 

“Theres no ‘aha moment' that was like, let me do cases. It'll do so well,” said Rosilier. “It was kind of just like, I have the stuff. I haven't done a business in a while and I'm homeschooled now for my sport, so might as well try it again.”

When she tried again, she was met with a resounding response as abbyrose has amassed nearly 500 thousand followers and over 15 million likes across social media platforms. Rosilier runs the company with the help of her three siblings and her mother. 

The family-owned and operated business continues to flourish with the support of The Hub. 

“I started with their iLaunch Competition, their version of ‘Shark Tank,' and then the Texas Tech Accelerator, and then I got an office,” said Rosilier. “It all just kind of fell into place – opened so many doors that you didn't even know could be opened.

“We got into the TTU Accelerator Program a year ago. I've never had mentors before; they helped us focus on the important things. We were trying to launch five products at once, and they were like, ‘You need to scale.'” 

Taking one glance at Rosilier's office space at the Hub, her room or even her company's logo you may notice a common theme: pink, and lots of it. Her favorite color covers her well-organized office floor to wall. The office is filled with equipment she uses to promote her products on social media like ring lights, phone stands and bins filled with phone cases. 

“I'm there every day,” said Rosilier. “It's such a good environment there, you'll want to go and stay because there's such good energy. The Hub has so many opportunities constantly. They've helped literally in every way possible.” 

If you have a business idea or would like to explore entrepreneurship, visit the Innovation Hub's website, innovationhub.ttu.edu to learn more about its programs.

You may also like

A funny thing happened on the way to a career, first-gen alumna finds success through texas tech’s waco site, generous donation founds texas tech’s first veterans legal clinic.

• MyU : For Students, Faculty, and Staff

Roberts group publishes synthetic chemistry research in Science

MINNEAPOLIS / ST. PAUL (04/25/2024) – The Roberts group recently published a new paper in  Science that explores enabling the use of a previously inaccessible functional group for N-heteroaromatic compounds.  Science – the flagship journal for the American Association for the Advancement of Science (AAAS) – publishes groundbreaking research across the spectrum of scientific fields. 

N-Heteroaromatic are an important class of molecules which are key to elements of pharmaceutical, agrochemicals and materials. Efficient and innovative methods to make functionalized heteroarenes are needed to make these critical molecules more readily available. One attractive method for the synthesis of N-heteroaromatic compounds would be the use of a N-heteroaryne – an aromatic ring containing a nitrogen atom and a triple bond. N-heteroarynes within 6-membered rings have been used as key intermediates for synthetic chemists, however after 120 years of aryne research the use of 5-membered N-heteroarynes has remained elusive. Notably, a computational model has predicted these 5-membered N-heteroarynes to be “inaccessible”, meaning they cannot be accessed synthetically due to the excessive strain associated with forming a triple bond within a small 5-membered ring.

The Roberts group hypothesized by applying principles of organometallic chemistry, forming 5-membered N-heteroarynes at a metal center would alleviate strain through back-bonding and allow access to this previously inaccessible functional group.  In a report which was published in  Science , the Roberts group achieved the first synthesis of 7-azaindole-2,3-yne complexes using phosphine-ligated nickel complexes. The complexes were characterized by X-ray crystallography and spectroscopy. Additionally, the complexes showed ambiphilic reactivity, meaning they react with both nucleophiles and electrophiles, making them an exceptionally versatile tool for the synthesis of N-heteroaromatic compounds. This exciting research breakthrough will have important applications in expanding the “chemist’s toolbox” for developing new pharmaceuticals, agrochemicals, and materials, and also provide fundamental insights on accessing synthetically useful strained intermediates.

This new work from the Roberts group was enabled by the National Institutes of Health, and by a multitude of fellowships held by the paper’s collaborators. Fifth-year PhD candidate Erin Plasek is supported by the UMN Doctoral Dissertation Fellowship;  fifth-year student Jenna Humke is supported by the National Science Foundation Graduate Research Fellowship Program; both Plasek and Humke are supported by Department of Chemistry Fourth-Year Excellence Fellowships; and third-year graduate student Sallu Kargbo was supported by the Gleysteen Departmental First Year Fellowship. For leadership excellence of her research program, Courtney Roberts has been awarded the 3M Alumni Professorship, the McKnight Land-Grant Professorship, the Amgen Young Investigator Award, and the Thieme Chemistry Journal Award in the past year alone.

“It is incredibly exciting to see this work, which started out as a few lines in my initial job proposals, come to fruition because of the exceptional team of students and postdocs behind it. We are delighted to finally share this new functional group for 5-membered N-heterocycles with the synthetic community,” Roberts writes.

Founded in 2019, the Roberts group uses inorganic and organometallic chemistry and catalysis to solve fundamental problems in synthetic organic chemistry related to pharmaceuticals, agrochemicals and materials. They have published work related to early transition metal catalysis, photochemical reactions, and inducing regioselectivity in metal-mediated aryne reactions. The group now consists of 14 graduate students, two postdoctoral associates, and one undergraduate researcher from a range of organic and inorganic backgrounds, which allows the team to take a multidisciplinary approach to solving research problems. They value diversity, collaboration, inclusivity, and radical candor in everything they do.

Roberts Group Website

Science Vol. 384 Issue 6694

Related news releases

• Hannah Kenagy and Melissa Ramirez join Department of Chemistry
• Taimeng Liang awarded ACS Division of Medicinal Chemistry Predoctoral Fellowship
• 2023 UMN Chemistry patents
• Jan-Niklas Boyn and Kade Head-Marsden join Department of Chemistry
• Frontiera, Massari, Pomerantz, and Tonks promoted to full professor
• Future undergraduate students
• Future transfer students
• Future graduate students
• Future international students
• Diversity and Inclusion Opportunities
• Learn abroad
• Living Learning Communities
• Mentor programs
• Programs for women
• Student groups
• Visit, Apply & Next Steps
• Information for current students
• Departments and majors overview
• Departments
• Undergraduate majors
• Graduate programs
• Integrated Degree Programs
• Additional degree-granting programs
• Online learning
• Academic Advising overview
• Academic Advising FAQ
• Academic Advising Blog
• Appointments and drop-ins
• Academic support
• Commencement
• Four-year plans
• Honors advising
• Policies, procedures, and forms
• Career Services overview
• Resumes and cover letters
• Jobs and internships
• Interviews and job offers
• CSE Career Fair
• Major and career exploration
• Graduate school
• Collegiate Life overview
• Scholarships
• Diversity & Inclusivity Alliance
• Anderson Student Innovation Labs
• Information for alumni
• Get engaged with CSE
• Upcoming events
• CSE Alumni Society Board
• Alumni volunteer interest form
• Golden Medallion Society Reunion
• 50-Year Reunion
• Alumni honors and awards
• Outstanding Achievement
• Alumni Service
• Distinguished Leadership
• Honorary Doctorate Degrees
• Nobel Laureates
• Alumni resources
• Alumni career resources
• Alumni news outlets
• CSE branded clothing
• International alumni resources
• Inventing Tomorrow magazine
• Update your info
• CSE giving overview
• Why give to CSE?
• College priorities
• Give online now
• External relations
• Giving priorities
• Donor stories
• Impact of giving
• Ways to give to CSE
• Matching gifts
• CSE directories
• Invest in your company and the future
• Recruit our students
• Connect with researchers
• K-12 initiatives
• Diversity initiatives
• Research news
• Give to CSE
• CSE priorities
• Corporate relations
• Information for faculty and staff
• Administrative offices overview
• Office of the Dean
• Academic affairs
• Finance and Operations
• Communications
• Human resources
• Undergraduate programs and student services
• CSE Committees
• CSE policies overview
• Academic policies
• Faculty hiring and tenure policies
• Finance policies and information
• Graduate education policies
• Human resources policies
• Research policies
• Research overview
• Research centers and facilities
• Research proposal submission process
• Research safety
• Award-winning CSE faculty
• National academies
• University awards
• Honorary professorships
• Collegiate awards
• Other CSE honors and awards
• Staff awards
• Performance Management Process
• Work. With Flexibility in CSE
• K-12 outreach overview
• Summer camps
• Outreach events
• Enrichment programs
• Field trips and tours
• CSE K-12 Virtual Classroom Resources
• Educator development
• Sponsor an event