Welcome to the October 13, 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."
|
|
|
All the Information You Need Could Be in Your Pocket
University of Waterloo Cheriton School of Computer Science (Canada)
October 12, 2021
Researchers at the University of Waterloo's Cheriton School of Computer Science in Canada have developed a tool to display basic information from incoming messages to smartphones and other wireless devices by shining it through one’s apparel. The PocketView technology uses light-emitting diode (LED) displays that can serve either as standalone devices or be linked wirelessly to smartphones through Bluetooth. Cheriton's Antony Albert Raj Irudayaraj said the displays show only minimal information, which is "good enough if you're walking or biking, for example, to show basic navigation instruction." Experiments found thin and light-colored fabrics transmit LED light better, while many darker-colored fabrics and denser, patterned weaves are sufficiently transparent to see the lit LEDs underneath, especially indoors.
|
Google Wants to Use AI to Time Traffic Lights More Efficiently
Reuters
Paresh Dave
October 6, 2021
Google said it intends to test the use of artificial intelligence (AI) to optimize traffic-signal timing in Rio de Janeiro, following a successful implementation in Israel. The Israeli project slashed fuel use and traffic delays at four locations in Haifa and Beer-Sheva by 10% to 20%. Rio's municipal traffic authority said Google's system should be launched within months, and had high hopes it could better control traffic signals. Although simulations showed AI control of traffic signals could smooth traffic flow, the University of Pittsburgh's Aleksandar Stevanovic questioned whether a technology firm with zero traffic engineering expertise could realize such software.
|
Smart Robots Do All the Work at Nissan's 'Intelligent' Plant
Associated Press
Yuri Kageyama
October 8, 2021
Nissan Motor Co. plans to have its "intelligent factory" in Tochigi, Japan, operational before April. Work in the factory, from welding and mounting to painting, will be done mainly by robots, while human workers at the plant will concentrate on more skilled work, like analyzing data collected by the robots and maintaining the equipment. Nissan's Hideyuki Sakamoto said, "Up to now, people had to make production adjustments through experience, but now robots with artificial intelligence, analyzing collected data, are able to do it. The technology has developed to that level." The factory will use the same assembly line to build vehicles powered by electricity, e-Power (combining an electric motor and internal combustion engine), and standard combustion engines.
|
Professional Footballers Threaten Data Firms with GDPR Legal Action
BBC News
Nick Hartley
October 21, 2021
Hundreds of professional footballers (soccer players) have threatened litigation against the data collection industry, demanding remuneration for the trading of their performance data over the past six years, and an annual fee for any future use. The 850-player Global Sports Data and Technology Group is led by former U.K. football team manager Russell Slade, whose legal team said lack of compensation for licensed use of footballers' personal data violates Europe's General Data Protection Regulation. The attorney leading the group's action, Chris Farnell, thinks it could lead to a game-changing rethink of data trading, particularly in terms of "how that data is being used and how it's going to be rewarded."
|
Engineers 3D-Print Personalized, Wireless Wearables That Never Need Charging
University of Arizona News
Emily Dieckman
October 8, 2021
University of Arizona (UA) engineers have developed personalized biosymbiotic devices that can operate without recharging. The three-dimensionally-printed wearables are based on body scans of wearers, and can function continuously via wireless power transfer combined with compact energy storage. The custom-fitted devices can wrap around various body parts, with specialized sensors positioned to measure otherwise unattainable physiological parameters. The devices are sufficiently accurate to detect body temperature changes caused by walking up a single flight of stairs. UA's Philipp Gutruf said, "These devices are designed to require no interaction with the wearer. It's as simple as putting the device on. Then you forget about it, and it does its job."
|
Cryptography Game-Changer for Biomedical Research at Scale
EPFL (Switzerland)
Tanya Petersen
October 11, 2021
An international research team has developed a federated analytics system that allows healthcare providers to perform statistical analyses and develop machine learning models in collaboration without sharing their underlying datasets. Researchers at Switzerland's EPFL (Swiss Federal Institute of Technology Lausanne), Lausanne University Hospital, the Massachusetts Institute of Technology, and Harvard University used the FAMHE federated analytics system to reproduce two published multicentric studies. They found the same scientific results could have been achieved without transferring and centralizing the datasets. Said EPFL's Jean-Pierre Hubaux, "FAMHE uses multiparty homomorphic encryption, which is the ability to make computations on the data in its encrypted form across different sources without centralizing the data and without any party seeing the other parties' data."
|
Simulated AI Creatures Demonstrate How Mind, Body Evolve, Succeed Together
TechCrunch
Devin Coldewey
October 6, 2021
Stanford University scientists engineered virtual artificial intelligences performing tasks in simulated environments to mimic the evolution of mind and body. The team dropped simulated animals they called unimals (for universal animals) into a simulation, initially so they could learn to walk. The virtual creatures developed various walks based on their environment's terrain; in further experiments, the unimals competed on more complex tasks. Those that had learned to walk on variable terrain learned the latter tasks faster and performed them better than those adapted to flat terrain. The researchers said this work "opens the door to performing large-scale in silico experiments to yield scientific insights into how learning and evolution cooperatively create sophisticated relationships between environmental complexity, morphological intelligence, and the learnability of control tasks."
|
Researchers Use AR to See Radiation Defects in Nuclear Reactors
Forbes
Jennifer Kite-Powell
October 11, 2021
A machine learning platform developed by researchers at the University of Michigan's Michigan Ion Beam Laboratory uses augmented reality (AR) to detect and quantify radiation-induced defects in parts and testing materials in nuclear reactors. The researchers tested a sample of iron, chromium, and aluminum with a krypton beam, which creates radiation defects when the krypton ions hit the sample. This allows for instantaneous quantification of radiation-induced defects, eliminating the need to download video and manually count every defect in selected frames. University of Michigan's Kevin Field said, "The software displays the results in graphics overlaid on the electron microscope imagery, which labels the defects—giving their size, number, location, and density—and summarizes this information as a measure of structural integrity."
|
Racial Bias Skewed Small-Business Relief Lending
The New York Times
Stacy Cowley
October 11, 2021
Financial technology companies were more likely to grant Paycheck Protection Program (PPP) loans to Black business owners than smaller banks were, researchers at New York University's Stern School of Business found, attributing the difference to automated loan vetting and processing systems, as well as human bias. The findings come amid growing scrutiny of how algorithmic systems can inadvertently perpetuate biases. Stern's Sabrina T. Howell said she was “taken aback by the striking disparity—it was a surprising and unexpected fact.” Howell said her research helped illustrate how technology could also help level the playing field. Howell said, “You can constrain an algorithm to meet fair-lending standards, and you can ensure the data it trains on isn’t biased."
*May Require Paid Registration
|
Novel ML Technique Identifies Structural Similarities, Trends in Materials
SciTechDaily
October 8, 2021
Scientists at Lehigh and Stanford universities demonstrated a novel machine learning (ML) method to recognize structural similarities and trends in materials from an unstructured image dataset. The team developed and taught a neural network to factor in symmetry-aware features, then applied the technique to 25,133 piezoresponse force microscopy images of diverse materials systems. The researchers employed the Uniform Manifold Approximation and Projection non-linear dimensionality reduction process to generate projections and cluster similar classes of material together from the image set.
|
Study Reveals Scale of Data-Sharing from Android Phones
Trinity College Dublin (Ireland)
October 11, 2021
A study by researchers at Ireland's Trinity College Dublin and the U.K.'s University of Edinburgh found that vendor-customized Android variants transmit significant amounts of data from mobile phones to the operating system (OS) developer and third parties with system apps pre-installed on those handsets, like Google, Microsoft, LinkedIn, and Facebook. The analysis found that users are unable to opt out from this data collection. Trinity College's Doug Leith said, "We've been too focused on Web cookies and on badly-behaved apps. I hope our work will act as a wake-up call to the public, politicians and regulators. Meaningful action is urgently needed to give people real control over the data that leaves their phones."
|
Platform Enables Comparative Research on Cancerous Tumors
Technion-Israel Institute of Technology
October 12, 2021
An algorithm developed by a team of researchers led by the Technion-Israel Institute of Technology allows for "variance-based comparison" of multidimensional data collected from the cancerous tumors of different patients. With the tuMap algorithm, numerous different tumors from different patients, and the same patient over time, can be compared on a uniform scale. The algorithm outperforms traditional tools due to the resolution it offers, which can be leveraged to predict various clinical indices with high accuracy, among other clinical applications. The researchers tested the algorithm on leukemia tumors, but they believe it will also be relevant for other cancer types.
|
The Rise of the Robo-Voices
The Wall Street Journal
Ellen Gamerman
October 7, 2021
Artificial intelligence (AI) is being used to create human-like voices, which could be used to dub foreign movies and TV shows, fix dialogue in post-production without the need for actors, and even resurrect audio from celebrities who have died or are no longer able to speak. In the coming months, dubbed versions of 2019 indie horror movie "Every Time I Die" are slated to be released in South America, using synthetic voices created by AI voice company Deepdub based on five-minute recordings of each actor speaking English. Said Zeena Qureshi of Sonantic, which recreated about two minutes of actor Val Kilmer's voice as a demonstration, "That idea of being able to customize voice content, to change emotions, pitch, direction, delivery, style, accents—that's now possible where it wasn't before."
*May Require Paid Registration
|
|
