Breakthroughs and trends in the world of technology
Additive Manufacturing
Virtual Viewing
In 2015, Dutch designer Joris Laarman announced an audacious project: constructing a stainless-steel pedestrian bridge over the venerable Oudezijds Achterburgwal canal in Amsterdam’s infamous red-light district using an in situ robotic printer. This past July, the world’s first 3-D-printed steel bridge finally opened to the public in a ceremony led by Queen Máxima of the Netherlands. The sinuous, flowing span, which stretches nearly 40 feet across the canal, has won multiple awards. But why did it take so long? One reason, according to the website of MX3D, the robotic 3-D metal printing company Laarman cofounded, was that the design wasn’t finished until 2017. Printing began in late 2018 and concluded six months later. The structure then underwent numerous load and safety tests at London’s Imperial College and the University of Twente, a Dutch tech school. The canal walls needed refurbishing, causing further delay. At some point, MX3D ditched plans to build on-site; instead, the completed structure was delivered by boat and hoisted into place by a crane. Embedded with sensors, the 6-ton bridge also serves as a living lab. Imperial engineers, for example, are monitoring how well the structure is holding up under constant pedestrian and bicycle traffic. Powered by Autodesk 3-D software, fabrication involved four industrial robots that welded layers of molten steel together using standard welding wire and gas at temperatures of 2,732º F. Hot stuff—and a perfect fit with its steamy surroundings. – Thomas K. Grose
© Thijs Wolzak
Artificial Intelligence
Critical Appraisals
Museums and art galleries rely on visitors to generate revenue—both of which plummeted during the pandemic. Italy’s cultural institutions, which lost around $225 million last year, have turned to high tech to better understand which paintings and sculptures are most popular and thereby boost ticket sales, according to Bloomberg News. ShareArt, a system developed by ENEA, a government agency that develops new technologies, uses cameras to record the number of observers, how long they gaze, and their distance to a piece. By better defining the “attraction value” of various works, curators may decide to change exhibit layouts to make some pieces more visible and accessible. Developed in 2016, the technology is only now being piloted at several venues as museums and galleries reopen. Among the findings so far: The average observation time is just 4 to 5 seconds—it’s the rare artwork that entrances for more than 15 seconds. The ShareArt cameras also will be able to conduct cognitive psychology analyses of people’s facial expressions, once mask mandates end. As the iconic Mona Lisa attests, however, a smile is open to interpretation. – T. G.
© ENEA
STEM Education
X Games
Size, scale, and powers of 10 are fundamental concepts in science, engineering, and math. But many students have a hard time conceptualizing just how big a billion is or how tiny atoms are, and that may intimidate them—or deter them from pursuing STEM. Karen Chen, an assistant professor of industrial and systems engineering at North Carolina State University, aims to change that dynamic. Armed with a $1.3 million National Science Foundation award, she is working with colleagues in the colleges of education and design to develop a virtual reality program designed to give students a better feel for extreme magnitudes. “It’s critical for students to develop accurate conceptions of scales that exist well beyond their everyday experience of the world,” Chen says. Her immersive project, Scale Worlds, will let students enter one of 31 virtual environments and see how infinitesimal or amazingly large objects appear in relation to their bodies. For instance, it will show the differences between molecules and cells. Students will be able to make realistic size comparisons in ways that aren’t possible in real life. If the approach works, Chen’s transdisciplinary research could be a very big deal. – T. G.
© Getty Images
Agricultural Engineering
Food Fight
As global warming disrupts agriculture worldwide, many engineering researchers are looking for ways to help communities avoid “food shock,” such as the crunch caused by a dry spell that hit farms in the Great Plains and the West between 2012 and 2015. A new statistical risk model developed by civil and environmental researchers at Penn State and Northern Arizona University highlights the need for urban areas to diversify their supply chains. Their key finding: There’s a distinct data-driven relationship between supply-chain diversity and the potential for sudden shortages. “If a city’s food sources are mainly from regions nearby, then its supply chain is not going to be very diverse,” one researcher explains. The model also indicates a problem if a city’s beef, say, all comes from one place, regardless of distance. Meanwhile, MIT engineering researchers have invented a process to help seeds cope with water shortages during their crucial germination phase. If it works, the technique could help farmers in already distressed arid regions. Inspired by the natural coats that protect chia and basil seeds in very dry soil, the team engineered a two-layer gel-like coating. The outer layer sucks up any available moisture and envelops the seed, while the inner layer uses a mixture of preserved microorganisms and nutrients to help the seeds grow. Early soil tests were promising, and the seeds are now being field-tested in Morocco. – T. G.
© Victor Leshyk, ECOSS, Northern Arizona University
Pediatric Medicine
Heard Immunity
Other than the common cold, middle ear infections are the most common childhood illness, afflicting more than 80 percent of U.S. kids. However, antibiotics are ineffective in more than 30 percent of chronic cases and their continued use raises the risk of drug resistance. One reason the bacteria that cause otitis media are so hard to treat is that they form biofilms behind the eardrum that are tough for antibiotics to penetrate. A research team that includes civil, environmental, and electrical and computer engineers at the University of Illinois is testing a different technique: using a microplasma jet array to zap the bugs. The tiny device creates a plasma composed of charged particles and reactive molecules that kills various pathogens—in this case, Pseudomonas aeruginosa, one of the most common of several bacteria that cause ear infections. In tests on a model of an inner ear that used an excised rat eardrum, the microplasma blasted the biofilm behind the eardrum. The optimal treatment time was found to be at least 15 minutes, with no obvious physical harm to the eardrum. Indirect measurements indicate that the microplasma destroys the biofilm by disrupting the membrane of bacterial cells. The encouraging results suggest the treatment could work in humans. The researchers, who are designing a smaller, earbud-shaped array to allow longer treatments, plan to runs tests on other types of bacteria that cause ear infections. – T. G.
© University of Illinois
Data Analytics
Views You Can Use
More than 700 imaging satellites currently collect and convey some 80 terabytes of geospatial information each day on such subjects as fresh water supplies or soil conditions. Storing—and sorting through—the massive amounts of binary data that researchers receive, however, typically requires expensive facilities and expertise. A team of University of California, Berkeley, researchers from various fields, including computer science, environmental science, economics, and public policy, has devised an inexpensive, easy-to-use technology that could give governments and academics in countries with fewer resources a way to access and analyze imaging data. The system, called MOSAIKS (multi-task observation using satellite imagery & kitchen sinks), uses radically simple and efficient machine-learning algorithms to organize and present data from multiple satellites. Any amateur with a laptop could, for example, track how quickly glaciers in Greenland are melting or gauge soil-moisture levels in remote villages. One researcher likens it to summarizing a dense and lengthy article. Ultimately, the Berkeley team hopes the satellite data can be collected into computer-based, constantly updating atlases that provide “pages” of broad, deep data about conditions in many countries or regions. – T. G.
© Getty Images
Biomedical Engineering
Booster Shot
Autoimmune disorders, including such chronic ailments as Crohn’s disease, psoriasis, and rheumatoid arthritis, afflict around 24 million Americans. Their underlying cause is a misfiring of the immune system that causes it to attack parts of the body. One well-studied malfunction is tied to the overproduction of the chemical TNF (tumor necrosis factor). A standard treatment involves injecting monoclonal antibodies that bind to excess TNF, allowing the body to flush itself. But the treatment causes side effects and doesn’t work well for everyone. Moreover, the immune system eventually learns to recognize and destroy the antibodies. A new approach invented by bioengineers at Duke University uses nanostructures to neutralize TNF without side effects and has been shown to work in mice with psoriasis. Funded by the National Institute of Biomedical Imaging and Bioengineering, the therapy uses a chemical process called supramolecular self-assembly to make many copies of complex molecules from uniquely functioning parts. The nanofiber acts like a backbone to which the protein subunits of the molecules that stimulate the body to sequester excess amounts of TNF are tethered. The nanomaterials “induce the body’s immune system to become an anti-inflammatory antibody factory,” explains Joel Collier, the professor of biomedical engineering who led the study. “It’s essentially an anti-inflammatory vaccine” that would require fewer doses and cost less than current treatments. Collier next plans to use mouse models to test the treatment on rheumatoid arthritis. – T. G.
© Chelsea Fries
Robotics
Soft Touch
Soft robots that are inflatable, flexible, and controlled by air or water rather than electricity are safer for human interactions. They’re also more adept at handling delicate or oddly shaped objects. A big hurdle, however, has been controlling the fluids that allow soft robots to move. Each finger of a robotic hand needs its own fluid line, for example, adding to the design’s complexity. A team of mechanical engineers at the University of Maryland has come up with a way to 3-D print soft robotic hands with a single, integrated fluidic circuit that controls all three fingers. Using PolyJet 3-D printing, the Maryland crew can fabricate a pressure-controlled hand in a day from layers of multi-material “inks.” Low pressure moves one finger; medium pressure, two fingers; high pressure, three fingers. As a proof of concept, the researchers had the robotic hand successfully play the first level of Nintendo’s video game Super Mario Bros. It used one finger to make Mario move, two to make him run, and three to have him run and jump. The design files are open-source to broaden the dissemination and adoption of soft ’bots. The invention isn’t all fun and games, however: The researchers now are investigating applications ranging from biomedical devices to surgical tools and prosthetics. – T. G.
© Joshua Hubbard and Kristen Edwards
Nanotechnology
Squeeze Play
For more than 10 years, Tony Jun Huang, a professor of mechanical engineering and materials science at Duke University, has investigated the use of sound waves to manipulate nanoparticles. His “acoustic tweezers” proved no match for particles smaller than the tiniest viruses, however. So Huang’s team used sound waves to create electric fields. The resulting electric-acoustic waves can control particles as small as 2.5 nanometers (the size of DNA) inside a liquid-filled chamber. His acousto-electronic tweezer method starts with a substrate of piezoelectric material that creates electricity when mechanically stressed. Atop the substrate is a small, liquid-filled chamber containing the particles, with four transducers on its sides that bombard the substrate with sound waves, creating electric fields. Those fields couple with the acoustic waves to generate electric-field patterns in the liquid, polarizing the particles so they can then be manipulated. In their experiments, the researchers arrayed nanoparticles into striped and checkerboard patterns and then shaped them into the letters D, U, K, and E. They also were able to transfer the aligned particles onto dry films. The technology’s many potential applications include blood tests for early detection of tumors or neurodegenerative diseases. It also might enable the placement of engineered nanoparticles into specific patterns to create new materials. – T. G.
© Peiran Zhang, Duke University
Assistive Technology
Compound Interest
Blind students typically are dissuaded from taking chemistry courses because of their visual nature and the potential danger of chemicals. Those who do are at a disadvantage in understanding the shape and function of molecules. While handheld models exist, they’re nowhere near as abundant as the number of molecules students must learn. Bryan Shaw, a professor of chemistry and biochemistry at Baylor University, has a potential workaround. In a recent paper, he describes his success in using small bite-size models that would allow blind students to sense and visualize protein structures with their mouths, Smithsonian magazine reports. Shaw’s eureka moment came when he realized that his son, who is blind in one eye, and a fully blind friend often would get a sense of objects by putting them in their mouths. His peanut-size models are made from gummy bear gelatin, some flavored to enhance the sensory experience. Smaller nonedible models are rice-size and made from a nontoxic resin that can be cleaned and reused. In tests with 281 sighted but blindfolded college students, participants were able to recall the models with high rates of accuracy. A second test using 31 elementary students proved equally successful. Shaw hopes his models eventually will be approved for sale, enabling more blind students to satiate their appetite for chemistry. – T. G.
© Brian Shaw
Neurology
Speak Your Mind
In a world first, researchers at the University of California, San Francisco, have created a device that successfully decoded and translated the brain signals of a severely paralyzed man when he attempted to speak words and sentences, which then appeared on a computer screen. Every year, strokes, accidents, and diseases rob thousands of their ability to talk. Current neuroprostheses typically use technologies that allow subjects to type out words letter by letter. For instance, the sole participant in the UCSF clinical trial—a man in his late 30s who suffered a debilitating stroke more than 15 years ago that damaged the connections between his brain, vocal tract, and limbs—uses a pointer attached to a baseball cap to tap letters on a touchscreen. The BRAVO (brain-computer interface restoration of arm and voice) project is based on 10 years of analyzing the speech patterns of epilepsy patients who underwent neural implants of electrode arrays to control their seizures. That data allowed the researchers to map out the cortical activity patterns linked to vocal tract movements. Bioengineers then devised an algorithm that can interpret those patterns as words and sentences. For the clinical trial, the participant underwent surgery to implant an electrode array over the part of his brain that controls speech. Twenty-two hours of his efforts to say 50 different words, such as “water” and “family,” were recorded over several months. He now can respond to questions with short sentences. UCSF’s algorithm can decode up to 18 words per minute with an accuracy rate of 93 percent. The researchers plan to expand the study to more people while working to increase their system’s vocabulary and rate of speech. – T. G.
© Ken Probst, UCSF
Machine Learning
Read the Room
Humans are pretty good at interpreting body language. Computers? Not so much. But a computer-vision algorithm designed by Columbia University engineering researchers is teaching machines to emulate our intuitive sense of human behavior and coordinate their actions to ours. The team trained the algorithm to anticipate reactions from handshakes to hugs using thousands of hours of movies, sports competitions, and TV shows like The Office. If the social cues are ambiguous, the algorithm hedges its bets, merely noting that two people are about to greet one another. The researchers say that the ability to divine behavior can help build trust between humans and machines—a vital component for safe interactions in increasingly automated workplaces, for example. It also could bolster the decision-making processes of autonomous automobiles. Next, the team wants to see how well the system works outside the lab in diverse settings—including, presumably, a paper distribution company with a diabolical salesman named Dwight. – T. G.
© Getty Images
Aerodynamics
Second Wind
Back in 2009, aerospace engineer Rick Cavallaro and a team at San Jose State University cobbled together a wind-powered vehicle named Blackbird to demonstrate it could race 2.8 times faster than the breeze blowing its giant propeller—a feat that seemed to defy the conservation of energy law but that they achieved the next year. Fast-forward to earlier this year, when Derek Muller, a popular YouTube science communicator with a Ph.D. in physics education research, test-drove Blackbird at El Mirage dry lake bed in the Mojave Desert and came away convinced. Not so viewer Alexander Kusenko, a physics professor at the University of California, Los Angeles, who thought Blackbird’s greater speeds resulted from wind gusts and inertia—and bet Muller $10,000 to prove him wrong. According to Business Insider, Muller supplied data from his test run showing that Blackbird accelerated for more than two minutes, undercutting the notion that its velocity came from random gusts. He also enlisted another YouTuber, engineer Xyla Foxlin, who built and tested a scale model of the vehicle on a treadmill with similar results. After three weeks of debate, Kusenko conceded that Blackbird could “slightly” outrun the wind, but only for short distances. Nevertheless, he ponied up. Muller has launched a video contest to divide his windfall among three inspiring science communicators. – T. G.
© YouTube.com/veritasium