Welcome to the February 26, 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."
|
|
|
Yale Targets Covid-19 with Innovative Computational Design Methods
YaleNews
Jim Shelton
February 24, 2021
Yale University scientists applied novel computation design techniques to formulate a new class of antiviral agents with potential for creating Covid-19 treatments. The researchers used computer simulation to perform in silico screening to identify the existing antiseizure medication perampanel, a drug that already has received U.S. Food and Drug Administration approval, as a starting point, then used free-energy perturbation calculation to help design the new compounds. Testing showed the compounds inhibited the viral protease following synthesis. The Yale researchers think the new compounds may lead to effective therapeutics for the treatment of SARS-CoV-2. Yale's William Jorgensen said, "From the standpoint of efficient molecular design, this is a truly striking and remarkable report."
|
Biden Signs Executive Order to Address Chip Shortage
CNBC
Lauren Feiner
February 24, 2021
On Wednesday, U.S. President Joe Biden addressed a global chip shortage with an executive order that aims to strengthen supply chains. The order calls for a 100-day review of such products as semiconductors and advanced batteries for self-driving vehicles, to be followed by a long-term review of six sectors of the economy. The pandemic strongly impacted the supply chain for semiconductors, and raised concerns about U.S. reliance on overseas supply chains. Companies like Ford and General Motors (GM) expect production declines as a result of the chip shortage, although GM's Paul Jacobson suggested the worst of the shortage may already be over.
|
AI Here, There, Everywhere
The New York Times
Craig S. Smith
February 23, 2021
Researchers anticipate increasingly personalized interactions between humans and artificial intelligence (AI), and are refining the largest and most powerful machine learning models into lightweight software that can operate in devices like kitchen appliances. Privacy remains a sticking point, and scientists are developing techniques to use people's data without actually viewing it, or protecting it with currently unhackable encryption. Some security cameras currently use AI-enabled facial recognition software to identify frequent visitors and spot strangers, but networks of overlapping cameras and sensors can result in ambient intelligence that can constantly monitor people. Stanford University's Fei-Fei Li said such ambient intelligence "will be able to understand the daily activity patterns of seniors living alone, and catch early patterns of medically relevant information," for example.
*May Require Paid Registration
|
Heavy Rain Affects Object Detection by Autonomous Vehicle LiDAR Sensors
University of Warwick (U.K.)
February 25, 2021
Researchers at the University of Warwick in the U.K. have found that the LiDAR sensors on autonomous vehicles (AVs) are less effective in detecting objects at a distance during periods of heavy rain. The researchers used the university's WMG 3xD simulator to test an AV's LiDAR sensors in different intensities of rain on real roads; they found that when the rainfall increased up to 50 mm per hour, object detection by the sensors dropped in conjunction with a longer range in distance. Warwick's Valentina Donzella said, "Ultimately we have confirmed that the detection of objects is hindered to LiDAR sensors the heavier the rain and the further away they are.”
|
Philadelphia Airport Employs Robot for Contact-Free Food Delivery
UPI
Ben Hooper
February 23, 2021
Philadelphia International Airport is using a robot to facilitate contactless food delivery for passengers. The Gita robot features a 40-pound (18.1-kilo) capacity cargo bin for food orders and uses Bluetooth to follow an airport AtYourGate representative to a location within the airport specified by the customer. The pilot program, which will run through April, allows customers to place food orders at airport restaurants via OrderAtPHL.com. PHL Food & Shops' Megan O'Connell said, "Gita allows our guests to customize their experience by choosing how much or how little human interaction they want when having their food delivered."
|
Tool Allows Computer Scientists to Support Life Scientists
Nevada Today
February 25, 2021
University of Nevada, Reno computer scientist Tin Nguyen and colleagues have developed a novel machine learning (ML) tool to support the research of life scientists. ScDHA (single-cell Decomposition using Hierarchical Autoencoder) employs ML to process data generated by single-cell sequencing. Nguyen said his team engineered a new non-negative kernel autoencoder to remove genes insignificant to cell differentiation, then developed a stacked Bayesian autoencoder to convert the data of tens of thousands of dimensions to a space with just 15 dimensions. Finally, the researchers crafted techniques for visualizing the transcriptome landscape, segregating cells into distinct types, inferring cellular developmental trajectory, and classifying the cells. Said Nguyen, "Without tools such as ours, it is impossible for life scientists to mine information from such large and high-dimensional data."
|
Why Do Men Publish More Papers Than Women?
CU Boulder Today
Lisa Marshall
February 25, 2021
Women in academia typically experience a decline in productivity after having children while men do not, which University of Colorado Boulder (CU Boulder) researchers attribute to persistent differences in parenting roles. The team polled more than 3,000 faculty members in the computer science (CS), history, and business departments at 450 U.S. and Canadian universities in 2018, and compiled data on publishing rates and institutions' parental leave policies. The research found that academics who are parents tend to publish more papers than non-parents, and fathers publish more than mothers; women publish 20% fewer papers than they would be expected to otherwise in the years after childbirth. Among CS faculty, mothers produce on average 17.6 fewer papers than fathers in the decade following childbirth. CU Boulder's Aaron Clauset said, "Any policy that can entice fathers to be more involved in parenting is likely to reduce this gender productivity gap."
|
Data Transfer System Connects Silicon Chips with Hair's-Width Cable
MIT News
Daniel Ackerman
February 24, 2021
A data transfer system developed by researchers at the Massachusetts Institute of Technology (MIT) and chipmaker Intel couples high-frequency silicon chips with a polymer cable as thin as a strand of hair. The system permits the transfer of information 10 times faster than a Universal Serial Bus, retaining the advantages of copper and fiber-optic conduits while avoiding their shortcomings. The plastic polymer conduit is lighter and potentially more affordable to produce than traditional copper cables, and offers greater efficiency when operated with sub-terahertz electromagnetic signals. Former MIT researcher Jack Holloway said the new conduit is "compatible directly with silicon chips, without any special manufacturing." Intel's Georgios Dogiamis said the cable could "address the bandwidth challenges as we see this megatrend toward more and more data."
|
AAP Team Develops Sustainable Building Simulation Method
Cornell Chronicle
Tom Fleischman
February 25, 2021
Researchers at the Cornell University College of Architecture, Art, and Planning (AAP) have unveiled the Sustainability Evaluation for Early Design (SEED) Framework, which could potentially save architects and design teams time and money later on. The Framework, implemented in the Rhino/Grasshopper computer-aided design platform, can dynamically model multiple factors, including building energy performance, embodied carbon, construction cost, and use of natural light to illuminate interiors. AAP's Allison Bernett and colleagues tested SEED in a case study of a hypothetical mid-sized office building modeled in three cities; the framework generated thousands of design options based on variables specific to each city, giving designers flexibility before the cost of changing options became excessive.
|
Security Flaw Detected for 2nd Time in Credit Cards
ETH Zurich (Switzerland)
Leo Hermann
February 22, 2021
A method for bypassing security measures to use certain credit and debit cards without a PIN code has been uncovered by researchers at Switzerland's ETH Zurich. Previously, the researchers had demonstrated that bypassing security was possible using Visa cards, while the new research shows security methods may be bypassed with Mastercard and Maestro cards by exploiting the data exchanged between the card and the card terminal. The method initially only worked with Visa cards, but the researchers were able to manipulate the payment process so the card terminal performed a Visa transaction and the card itself performed a Mastercard or Maestro transaction. The researchers informed Mastercard of their findings, after which the company updated the relevant safeguards.
|
ML Aids in Simulating Dynamics of Interacting Atoms
Los Alamos National Laboratory News
February 23, 2021
Los Alamos National Laboratory researchers were able to simulate the motions of atoms in materials like aluminum using a new machine learning approach. The researchers developed an automated "active learning" methodology for iteratively building a training dataset. The method employs the current best machine learning model at each iteration, with new reference data collected through quantum simulations as new physical situations are encountered in order to retrain the model. The data collected includes different types of atomic configurations, including various crystal structures and defect patterns within crystals. Los Alamos' Justin Smith said, "Simulating the dynamics of interacting atoms is a cornerstone of understanding and developing new materials. Machine learning methods are providing computational scientists new tools to accurately and efficiently conduct these atomistic simulations."
|
Prototype of Intelligent Underground Robotic System for Urban Environments Developed
Universidad Carlos III de Madrid (Spain)
February 25, 2021
An international research team led by Spain's Universidad Carlos III de Madrid (UC3M) has developed a robot that can intelligently navigate underground urban environments, which can be deployed to install wires, pipes, and other small-diameter underground scoring below a city’s streets without having to open a ditch. Part of the BADGER (roBot for Autonomous unDerGround trenchless opERations, mapping and navigation) project, the prototype robotic system includes a surface vehicle equipped with geo-radar to scan the ground, and a 3-meter-long drilling robot. After the rover scans a location’s subsoil, software developed by the research team determines entry and exit points for the work. Said UC3M's Carlos Balaguer, "The use of innovative localization, mapping and navigation techniques, along with sensors and geo-radars, allows the systems to be adapted to different fields."
|
Virtual Computer Chip Tests Expose Flaws, Protect Against Hackers
New Scientist
Matthew Sparkes
February 24, 2021
Researchers at the University of Michigan, Virginia Polytechnic Institute and State University, and Google have accelerated computer-chip testing by simulating chips and applying advanced software testing tools for analysis of the simulations. Virtual testing lets engineers utilize fuzzing, a method that monitors for unexpected results or crashes that can be reviewed and corrected. The researchers had to modify software fuzzers to run over time, rather than trigger a single input and wait for the response. This approach enabled a chip that would usually take 100 days to test to be analyzed in one day. The researchers think faster hardware testing could reduce development time and bring more reliable, more secure next-generation chips to market faster.
*May Require Free Registration
|
|
