Welcome to the March 29, 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."
|
|
New Wave of 'Hacktivism' Adds Twist to Cybersecurity Woes
Reuters Joseph Menn March 25, 2021
Activist hackers looking to make political statements constitute emerging threats to U.S. cybersecurity. The U.S. government charged non-binary Swiss hacker Tillie Kottmann with conspiracy for their claimed exposure of artificial intelligence-powered corporate video surveillance by the startup Verkada. Hacktivists also exposed January 6 Capitol riot videos from the right-wing social network Parler, which Gabriella Coleman at Canada's McGill University said indicated support for antiracist or antifascism politics. Emma Best of the Distributed Denial of Secrets website said indictments like Kottmann's "show just how scared the government is, and just how many corporations consider embarrassment a greater threat than insecurity."
|
Deep Learning Tool Predicts Hosts Based on Early SARS-CoV-2 Samples
News-Medical Life Sciences Angela Betsaida B. Laguipo March 24, 2021
Researchers at China's Peking University designed a deep learning tool that can predict hosts for new viruses, including SARS-CoV-2. DeepHoF (Deep learning-based Host Finder) forecasts host probability scores for five host types: germ, plant, invertebrate, non-human vertebrate, and humans. DeepHoF was based on BiPath Convolutional Neural Network, and can automatically infer genomic features from inputted viral sequences. The Peking researchers conducted a deep analysis of the host likelihood profile calculated by DeepHoF, using the earliest samples of SARS-CoV-2 isolates to determine that minks, dogs, bats, and cats were potential hosts—with minks particularly significant. The new model also can help predict potential hosts of viruses that may induce another outbreak or pandemic.
|
Researchers Developed Backpack System to Guide Vision-Impaired Wearers
Engadget Kris Holt March 24, 2021
An artificial intelligence (AI)-powered backpack developed by University of Georgia (UGA) researchers can help vision-impaired wearers navigate their environment. The tool employs a Luxonis OAK-D spatial 4K camera equipped with an on-chip edge AI processor and Intel's Movidius image processing technology. The camera is bundled inside a vest or fanny pack, while Intel's OpenVINO toolkit is used for inferencing; the system can operate for up to eight hours via a pocket-sized battery, while the backpack also holds a lightweight computer with a global positioning system unit. The UGA researchers said the system can detect obstacles and relay the wearer's whereabouts through audio prompts, as well as reading traffic signs and identifying changes in elevation.
|
Robot Learns to Tie Knots Using Only Two Fingers on Each Hand
New Scientist Matthew Sparkes March 26, 2021
Tetsuya Ogata and colleagues at Japan's Waseda University have taught an artificial intelligence-powered robot to tie knots around a box using just two fingers on each hand. The team first directed a two-armed robot via remote control to manually knot a piece of rope dozens of times, then combined data recorded by the arms with information from an overhead camera and proximity sensors on the fingers; half the rope was colored red and the other half blue to aid identification. The Waseda researchers used the combined data to train a neural network to replicate the task, and the robot was 95% successful in tying a bowknot with the colored rope, and 90% successful with a white rope for which it was not trained.
|
'Smart Clothes' Can Measure Your Movements
MIT Computer Science and Artificial Intelligence Laboratory Adam Conner-Simons March 24, 2021
Researchers at the Massachusetts Institute of Technology's Computer Science and Artificial Intelligence Lab (CSAIL) have developed clothing that incorporates special fibers to detect the wearer's movements. The clothes feature "tactile electronics" that can pinpoint whether the wearer is sitting, walking, or performing particular poses. Using these tactile textiles, the researchers have developed prototypes that range from socks and gloves to a full vest, which could be used for athletic training and rehabilitation or to determine whether patients in assisted-care facilities have fallen or are unconscious, among other things. CSAIL's Wan Shou said, "Clothing with high-resolution tactile sensing opens up a lot of exciting new application areas for researchers to explore in the years to come."
|
Fire-Simulating Tool Could Improve In-Flight Fire Safety
U.S. National Institute of Standards and Technology March 24, 2021
Scientists at the U.S. National Institute of Standards and Technology (NIST) and Federal Aviation Administration (FAA) tested a fire-simulating computer model against a real-world scenario for a grounded commercial airliner. The Fire Dynamics Simulator (FDS) models the flow of heat and smoke generated by fires, and the NIST-FAA team built a digital version of the airliner space as an environment for FDS simulations. Real-world experiments and FDS simulations generally correlated in terms of measured temperatures and hot spots. The results indicated the upgraded FDS can capture several properties of an actual overhead compartment fire, and suggested it could be further developed into a reliable tool for fire protection engineers designing aircraft systems.
|
NJIT, Ben-Gurion University of the Negev Launch Institute for Future Technologies
New Jersey Institute of Technology March 16, 2021
The New Jersey Institute of Technology (NJIT) and Israel's Ben-Gurion University of the Negev (BGU) have partnered to launch the Institute for Future Technologies in New Jersey. The NJIT-BGU collaboration aims to deliver bespoke cyber technologies, civil and environmental engineering education, applied research and development, and innovation and entrepreneurship via technological commercialization. The Institute will operate from NJIT’s main campus in Newark, as well as a satellite location in Jersey City. In announcing the partnership, New Jersey Gov. Phil Murphy said, "NJIT and BGU will combine their expertise and track records in technological research and development to help strengthen the economic opportunity and tech leadership that I have long envisioned for our state."
|
TikTok Does Not Pose Overt Threat to U.S. National Security, Researchers Say
The Wall Street Journal Eva Xiao March 22, 2021
Cybersecurity researchers at the University of Toronto's Citizen Lab in Canada said TikTok's underlying computer code does not pose a national security threat to the U.S. The researchers said a technical analysis of the app, owned by China's ByteDance Ltd., found no evidence of "overtly malicious behavior." Although they determined that TikTok's data collection practices are no more intrusive than Facebook's, the researchers acknowledged there could be security issues they did not uncover. Further, ByteDance could be forced to turn data over to the Chinese government under the country's national security laws. ByteDance said it was committed to working with authorities to resolve their concerns.
*May Require Paid Registration
|
Computer Model Tracks Cellphone Data to Predict Covid Spread
Binghamton University Chris Kocher March 25, 2021
Binghamton University's Arti Ramesh and Anand Seetharam have designed an algorithm to predict the spread of Covid-19 by tracking cellphone data. The scientists matched cellphone data with coronavirus infection rates in Rio de Janeiro's municipal districts to construct a mathematical model that predicts how cases would shift the next week for Rio’s municipal districts; forecasts were based on current Covid rates blended with mobility trends between districts with the most cases. Seetharam said, "This is one of the first studies that has quantified mobility in a manner that it can be used to demonstrate how cases are going to spread. It’s not just the number of cases in a particular region that contributes to future cases in that region."
|
Novel Thermometer Can Accelerate Quantum Computer Development
Chalmers University of Technology (Sweden) March 23, 2021
Researchers at Sweden's Chalmers University of Technology have built a novel thermometer that measures temperature during quantum calculations with high accuracy, providing a valuable benchmarking tool for quantum computing. Chalmers' Simone Gasparinetti said, "Our thermometer is a superconducting circuit, directly connected to the end of the waveguide being measured. It is relatively simple—and probably the world's fastest and most sensitive thermometer for this particular purpose at the millikelvin scale." Critical to operation of the 100-quantum-bit (qubit) quantum computer Chalmers scientists hope to build based on superconducting circuits by 2030 is a processor that ideally operates at 10 millikelvin (-459 degrees Fahrenheit); the thermometer would measure whether that temperature is reached during the device’s operation. Chalmers' Per Delsing said, "If we succeed in lowering the temperature at the end where the waveguide meets the qubit to 10 millikelvin, the risk of errors in our qubits is reduced drastically."
|
UCLA Researchers Develop Noninvasive AI Method to Inspect Live Cells, Gain Critical Data
UCLA Samueli School of Engineering March 24, 2021
At the University of California, Los Angeles (UCLA) Samueli School of Engineering, researchers have developed a noninvasive artificial intelligence (AI)-based technique to analyze live biological cells. Using a deep learning model, the cells are viewed and a snapshot captured under a brightfield microscope. The UCLA team trained the model to deduce and identify antibody-labeled fluorescent cellular images, and to note subtle distinctions in size and shape in order to predict protein levels. This analysis yields data on protein concentrations and location without destroying the sample, and the AI tool can predict as many proteins as the model has been trained to identify. UCLA's Neil Lin said this method could be of use to academic and industrial cell biology laboratories, and "it could be especially important in cell therapies, where the cells themselves are valuable."
|
Optimization for Resource Management Using Multi-Agent Systems
University of Eastern Finland March 25, 2021
Researchers at the University of Eastern Finland (UEF) testing a thesis that multi-agent systems can be used to optimize resource management modeled car drivers using artificially intelligent agents to simulate traffic in road networks. The agents behave as drivers would behave in real-world situations, for example by taking the shortest route to their destination, but possibly opting to reroute if they encounter congestion. The UEF team applied this approach to explore two scenarios: how drivers in Singapore would impact the country's energy consumption, assuming every vehicle is electric, and the effect of bridge removal across the Finnish city of Joensuu's Pielisjoki river on traffic load. In Singapore, the thesis results are being applied to help digitally transform the city's infrastructure.
|
|