Welcome to the September 10, 2021 edition of ACM TechNews, providing timely information for IT professionals three times a week.
ACM TechNews mobile apps are available for Android phones and tablets (click here) and for iPhones (click here) and iPads (click here).
To view "Headlines At A Glance," hit the link labeled "Click here to view this online" found at the top of the page in the html version.
The online version now has a button at the top labeled "Show Headlines."
|
|
|
Portable 'Lab-on-a-Chip' Diagnostic Platform Can Rapidly Test Dozens of People for COVID-19
UT News
September 9, 2021
A prototype portable diagnostic platform built by engineers at the University of Texas at Austin (UTA) and spinoff Omega Optics could help medical staff rapidly detect asymptomatic COVID-19, as well as other difficult-to-spot diseases. The researchers said the "lab-on-a-chip" platform can detect minute virus concentrations in a sample, and test dozens or even hundreds of patients faster and at less cost than current techniques. The system is inexpensive to manufacture, and its reliance on silicon chips enables easy scalability. Said UTA's Ray Chen, “Instead of having to bring people to hospitals, which isn’t always doable in remote areas, we can bring the hospital to the people.”
|
Experts Call for Rigorous Audit to Protect California Recall
Associated Press
Christina A. Cassidy; Kate Brumback
September 3, 2021
Election security experts, including computer scientists and cybersecurity researchers, have sent a letter to California's Secretary of State calling for an audit of the state’s Sept. 14 gubernatorial recall election following the public distribution of Dominion Voting Systems' election management system. The letter stated that "the release of the Dominion software into the wild has increased the risk to the security of California elections to the point that emergency action is warranted." The experts want California counties using the Dominion system to perform a "risk-limiting audit," which employs a statistical approach to ensure reported results and actual votes cast match up. University of Michigan's J. Alex Halderman said even voters now have sufficient physical access to voting systems to implant malware. Said Halderman, "It's just really multiplied the number of people who are in a position to do harm to our elections by a very large factor."
|
VR Puts Drug Researchers Inside Molecules
The Wall Street Journal
Sara Castellanos
September 7, 2021
Drug-discovery researchers are using virtual reality (VR) to examine the structures of molecules in order to determine how fluctuations in the distance, configuration, and chemical properties of such structures could clarify a drug's functions. Scientists can probe virtual environments filled with color-coded drug molecules and proteins, manipulating them with hand controllers to more easily identify promising molecules faster. Startup Nanome, which developed three-dimensional modeling software for this purpose, also is developing an artificial intelligence assistant that can guide users through the VR experience.
*May Require Paid Registration
|
Only Humans, Not AI Machines, Get a U.S. Patent, Judge Rules
Bloomberg
Susan Decker
September 3, 2021
U.S. District Judge Leonie Brinkema in Alexandria, VA, has ruled that only an actual human can be listed as an inventor on patents under U.S. law, while a computer using artificial intelligence (AI) cannot. Brinkema ruled that under federal law, an "individual," defined as a natural person, is required to take an oath that they are the inventor on a patent application. The case involved the Artificial Inventor Project at the U.K.'s University of Surrey, which has undertaken a global effort to have a computer named as an inventor. Courts in South Africa and Australia have ruled in the project's favor. University of Surrey's Ryan Abbott said the U.S. ruling will be appealed, along with those in the U.K. and Europe. Said Abbott, "We believe listing an AI as an inventor is consistent with both the language and purpose of the Patent Act."
|
Quantum Computing Breakthrough: Entanglement of Three Spin Qubits Achieved in Silicon
SciTechDaily
September 6, 2021
Researchers at Japan's RIKEN Center for Emergent Matter Science have entangled three silicon-based spin quantum bits (qubits) in a fully controllable array, and quantified the device with high fidelity. RIKEN's Seigo Tarucha said, "Two-qubit operation is good enough to perform fundamental logical calculations, but a three-qubit system is the minimum unit for scaling up and implementing error correction." The array features a triple quantum dot on a silicon/silicon–germanium heterostructure, controlled via aluminum gates. Each quantum dot can host one electron, whose spin-up and spin-down states encode a qubit; an on-chip magnet produces a magnetic-field gradient that partitions the qubits' resonance frequencies.
|
Universal System Decodes Any Type of Data Sent Across a Network
MIT News
Adam Zewe
September 9, 2021
Scientists at the Massachusetts Institute of Technology (MIT), Boston University, and Ireland's Maynooth University have programmed a universal algorithm to enable a silicon chip to decode any data transmitted across a network, irrespective of structure. The Guessing Random Additive Noise Decoding (GRAND) algorithm removes the need for multiple computationally complex decoders. GRAND guesses the noise or energy affecting the encoded data en route, and uses this pattern to infer the original information. The algorithm produces noise sequences in the order they are likely to occur, eliminates them from the received data, and confirms that the resulting codeword is in a codebook. The researchers say the GRAND chip could decode any moderate redundancy code up to 128 bits long, with approximately 1-microsecond latency.
|
AR Collection by Ukrainian Companies Hits Runway
Reuters
Margaryta Chornokondratenko
September 8, 2021
Ukrainian clothing brand FINCH teamed up with Ukrainian technology company FFFACE.ME to develop semi-digital clothing that can be upgraded with augmented reality (AR). FINCH's Kateryna Biakova said, "By using [AR], a designer can create items beyond physical characteristics of fabrics or elements of clothing which do not exist in the real world, but can be seen as dynamic [three-dimensional] 3D objects in virtual reality." Scanning a QR code printed on one of these articles of clothing activates an AR layer through a FFFACE.ME-developed Instagram filter. Each piece of clothing, which costs $70 to $178, can have several digital looks and even comprise an entire digital wardrobe in AR.
|
Next-Gen 3D-Printed Catalysts to Propel Hypersonic Flight
Royal Melbourne Institute of Technology (Australia)
September 8, 2021
Three-dimensionally (3D)-printed catalysts designed by researchers at Australia's Royal Melbourne Institute of Technology (RMIT) could address overheating issues in hypersonic aircraft. The cost-effective, scalable, and versatile catalysts could potentially power hypersonic flight while also cooling the system. The team printed tiny heat exchangers from metal alloys and coated them with zeolites; extreme heat caused some of the metal to move into the zeolite framework, a critical process for enabling efficiency. Said RMIT’s Roxanne Hubesch, “Our 3D-printed catalysts are like miniature chemical reactors, and what makes them so incredibly effective is that mix of metal and synthetic minerals.”
|
One-Legged Robot Hops About as Researchers Try Knocking It Over
New Scientist
Bas den Hond
September 7, 2021
The one-legged TTI Hopper robot can stand, hop, and maintain balance on sloping or unsteady surfaces. Researchers at Japan's Toyota Technological Institute (TTI) constructed the robot using simple direct current motors and inexpensive gears, as well as an algorithm that compensates for the limited functionality of those components. The algorithm slackens joints when shocks must be absorbed, and stiffens them when the robot must right itself or push off while running, keeping the robot upright. The team proved mathematically that the algorithm ensured the robot's stability by calculating the vertical and horizontal forces on the leg from the motors and gravity. Former TTI researcher Barkan Ugurlu said he now is working to adapt the algorithm for use in a bipedal exoskeleton for legless people. “My idea is to use the terrain adaptation ability to get rid of the crutches.”
|
Companies Need More Workers. Why Do They Reject Millions of Résumés?
The Wall Street Journal
Kathryn Dill
September 4, 2021
Software that sorts through job applicants and deletes those it judges unsuitable may be a major reason why many people are not being hired, despite massive numbers of available positions. A Harvard Business School study estimated such software systems exclude more than 10 million workers from hiring discussions, and many company leaders acknowledge the software is filtering out good applicants. The Harvard researchers said 99% of Fortune 500 companies and 75% of 760 polled U.S. employers are using applicant-tracking systems that filter numerous job prospects down to a select group. The algorithms on which such software is based usually rank applicants based on broad criteria, and the more complicated the job description, the more people the software rejects. Some companies believe the technology can be enhanced to better match candidates to jobs, while others are using less-automated methods to find qualified applicants.
*May Require Paid Registration
|
AWS Researcher Merges 2 Quantum Computers to Help Make Cryptography Keys Stronger
ZDNet
Daphne Leprince-Ringuet
September 3, 2021
Amazon Web Services' Mario Berta combined the capabilities of two quantum computers in order to generate truly random numbers to strengthen cryptographic keys. Berta merged quantum processors from Rigetti and IonQ to exploit quantum particles' state of superposition, and the phenomenon that an equal number of quantum bits can yield a string of bits with an equal number of random values. The processors generate two independent bit strings that are processed by a randomness extractor (RE) algorithm to combine multiple weakly random bit sources into one nearly perfect random string.
|
Paint the Town with ML
Research at Osaka University (Japan)
September 5, 2021
Scientists at Japan's Osaka University used machine learning to classify building types and primary facade color within street-level images. The team used an open source version of the Baidu Street View navigation application to generate color and building classification maps for three Chinese cities. The researchers surveyed the locales to establish the ground truth for assessing facade segmentation, dominant color, and building classification. A semantic segmentation algorithm extracted facade outlines from the street view images, with color calibration of photos performed during pre-processing before a multi-label deep learning algorithm classified facade color and building type. The researchers then produced color distribution maps for each city. Said Osaka University’s Tomohiro Fukuda, “Our work helps computers interact with the aesthetic part of architecture and city planning, which has been very difficult to do previously.”
|
|
