Welcome to the December 23, 2022, edition of ACM TechNews, providing timely information for IT professionals three times a week.
Please note: In observance of the Christmas holiday, TechNews will not be published on Monday, December 26. Publication will resume Wednesday, December 28.
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."
|
|
Hackers Used Software Flaw to Take Down County Computer System
The New York Times Sarah Maslin Nir; Nate Schweber December 21, 2022
Suffolk County, NY, officials disclosed that hackers planned this fall's crippling ransomware attack on the county more than a year ago. Forensic analysis indicated the professional hacker gang BlackCat exploited a vulnerability in an obscure but common piece of software to infiltrate Suffolk's computer system on Dec. 19, 2021. Last year, the U.S. Cybersecurity & Infrastructure Security Agency published an advisory about the software flaw, recommending vulnerable organizations update their systems. Several Suffolk departments created a patch, but the county lacks a centralized interdepartmental cybersecurity protocol, while information technology teams are siloed. Suffolk executive Lisa Black said the office of the county clerk failed to patch the bug, enabling the hackers to orchestrate their infiltration and attack.
*May Require Paid Registration
|
Words Prove Their Worth as Teaching Tools for Robots
Princeton University School of Engineering and Applied Science Molly Sharlach December 21, 2022
Princeton University researchers taught a simulated robot arm to use tools more quickly via human-language descriptions through the Accelerated Learning of Tool Manipulation with LAnguage model. Having access to descriptions of a tool's form and function to the training process improved the robot's ability to manipulate tools that were not in its original training set. The researchers queried OpenAI's deep learning GPT-3 language model for tool descriptions, then chose a training set of 27 tools and tasked the robotic arm to push, lift, sweep a cylinder along a table, or hammer a peg into a hole with nine test tools. In most cases, they found the descriptions improved the robot’s ability to use the new tools.
|
Computer Architecture Is Being Reimagined at Technion
The Jerusalem Post (Israel) Judy Siegel-Itzkovich December 21, 2022
Shahar Kvatinsky and colleagues at Israel's Technion-Israel Institute of Technology, working with Israeli chipmaker Tower Semiconductor, have built a neural network directly into a processor's hardware as a proof of concept, and trained it to recognize handwritten letters. The researchers engineered the chip to store and process information, with programming integrated within the processor. The chip, which learns through deep-belief algorithms, was able to distinguish between individual examples of each letter, and recognized them 97% of the time despite extremely low energy consumption.
|
This Is Your Brain. This Is Your Brain on Code
MIT News Steve Nadis December 21, 2022
Massachusetts Institute of Technology researchers determined that the brain's demand and language systems encode specific code properties that correlate with machine-learned representations. The researchers performed functional magnetic resonance imaging (fMRI) scans of programmers' brains as they processed computer program elements and saw a difference in encoding properties related to dynamic analysis, in which information is encoded much better in the brain's multiple demand network compared to the language processing center. A second experimental series incorporating neural networks confirmed the brain signals observed when participants examined pieces of code closely resembled the networks' own activation patterns. Mariya Toneva at Germany's Max Planck Institute for Software System said such findings "raise the possibility of using computational models of code to better understand what happens in our brains as we read programs."
|
Police in Australia Co-Opted COVID-19 Apps to Fight Crime
Associated Press Rod McGuirk December 20, 2022
Law enforcement officials in Western Australia ordered the state Health Department to provide information from the SafeWA COVID-19 contract tracing app as part of an investigation into a murder at a speedway in Perth. The QR code check-in data included the names, phone numbers, and arrival times of 2,439 fans who attended the December 2020 race. Provided under a government order requiring contact tracing information in the event of a COVID-19 outbreak, the data was supposed to be accessible only to contact tracing personnel. The matter has raised concerns over privacy, particularly since Australia's 1988 Privacy Act was implemented before widespread use of the Internet and smartphones.
|
Software Helps Interpret Complex Data
Helmholtz-Zentrum Berlin (Germany) December 20, 2022
Software developed by researchers at Germany's Helmholtz-Zentrum Berlin (HZB) can interpret complex data using self-learning neural networks (NNs), one that compresses the data and another that reconstructs a low-noise version of the data. HZB's Gregor Hartmann explained, "In the process, the two NNs are trained so that the compressed form can be interpreted by humans.” Hartmann said the special class of NNs, known as disentangled variational autoencoder networks, can extract the data's underlying core principle without prior knowledge. The software was used to determine the photon energy of the FLASH free electron laser at the Deutsches Elektronen-Synchrotron research center from single-shot photoelectron spectra. Said Hartmann, "We succeeded in extracting this information from noisy electron time-of-flight data, and much better than with conventional analysis methods."
|
Virtual Clothes Could Help Solve Fashion's Waste Problem
CNN Tom Page December 22, 2022
The fashion industry is tapping digital technologies to cut waste and modernize. Italian fashion platform Lablaco is using blockchain to digitize fashion houses and brands' collections in the "phygital" market, when customers purchase both a physical fashion item and its digital twin. Lablaco's model keeps phygital items paired after sale so if a physical item is resold, the digital version goes to the new owner's digital wallet. Blockchain's transparency assures the new owner of the apparel's authenticity, while its creator can track its aftersales journey. Lablaco's Eliana Kuo said digital spaces could serve as a testbed for physical clothing by using sales data to inform real-world marketing.
|
Health-Monitoring Wearable Measures Vital Signs
IEEE Spectrum Kathy Pretz December 20, 2022
Researchers at Sri Lanka's University of Jaffna and French telecommunications company DotsHook designed the multipurpose health-monitoring bracelet (MHMB) to measure vital signs including heart rate, skin temperature, blood pressure, and blood oxygen saturation. The battery-powered device, designed to connect to a cloud-based remote patient-monitoring system through Wi-Fi, applies the optical technique photoplethysmography to measure a patient’s heart rate. Medical personnel can access patient data from the system or via an application, while patients can download an app to their phones to view their readings. The bracelet also can interface with other medical devices, making the wearable a potential central node for a patient-monitoring system.
|
Researchers Create Eye Tissue With 3D Bioprinting
U.S. National Institutes of Health December 22, 2022
Kapil Bharti and colleagues at the U.S. National Institutes of Health's National Eye Institute three-dimensionally (3D)-bioprinted eye tissue from patient stem cells. The researchers produced a combination of cells forming the outer blood-retina barrier, tissue that supports the retina's light-sensing photoreceptors. They integrated pericytes, endothelial cells, and fibroblasts within a hydrogel, then printed it on a biodegradable scaffold to mature into a capillary network. The scientists impregnated the flip side of the scaffold with retinal pigment epithelial cells. The printed tissue was found to have fully matured in 42 days. Bharti said, "By printing cells, we're facilitating the exchange of cellular cues that are necessary for normal outer blood-retina barrier anatomy."
|
Even the FBI Says You Should Use an Ad Blocker
TechCrunch Zack Whittaker December 22, 2022
The U.S. Federal Bureau of Investigation has issued an alert advising online users to install and use ad blockers. The advisory warns of hackers buying online ads highly placed in search results to pose as legitimate brands, in order to steal or extort money from targets. Malicious ads also are used to fool victims into installing malware masked as genuine applications. Ad blockers bar the display of any advertising, making it easier for browsers to find and access the websites of authentic brands. Ad-blocking Web browser extensions also prevent the tracking code within ads from loading, and some of the most effective ad blockers are available to consumers for free.
|
Robot Bird Uses Talon-Like Claw to Perch Safely
New Scientist Alex Wilkins December 15, 2022
Researchers at the Swiss Federal Institute of Technology in Lausanne created a robot bird that uses a claw to grasp a perch after a short flight. The claw has talon-like tips and a softer middle area with which to grip branches. An indoor motion-capture system guides the bird to within a few meters of its target, then an onboard camera directs the center of the claw to the perch; when the claw's soft section makes contact, a spring snaps it shut. The robot also uses an indoor global positioning system to navigate to a branch and a close-range camera to make adjustments prior to landing.
|
Harnessing AI Technology for IVF Embryo Selection
Weill Cornell Medicine Newsroom December 19, 2022
The STORK-A artificial intelligence algorithm developed by Weill Cornell Medicine researchers can ascertain non-invasively whether the chromosome count of in vitro fertilized embryos is normal or abnormal. STORK-A uses microscope images of embryos recorded five days post-fertilization, along with clinical staff's scoring of embryo quality, maternal age, and other data. The algorithm correlates certain data properties that may trend toward aneuploidy (having an abnormal number of chromosomes). STORK-A was 69.3% accurate in predicting aneuploidy involving multiple chromosomes, and 77.6% accurate in predicting euploidy (a change in the entire set of chromosomes). Weill Cornell's Nikica Zaninovic said use of the technology "can reduce the number of embryos to be biopsied, reduce the costs, and provide a very good tool for consultation with the patient when they need to make a decision whether to do PGT-A [preimplantation genetic testing for aneuploidy] or not."
|
Researchers Zoom In on Battery Wear and Tear
University of Chicago Pritzker School of Molecular Engineering Sarah C.P. Williams December 22, 2022
Researchers at the University of Chicago's Pritzker School of Molecular Engineering, the University of California, San Diego, and scientific service provider Thermo Fisher Scientific observed lithium-ion battery degradation at the atomic level, using plasma-focused ion beam-scanning electron microscopy to view changes within a new and a well-used lithium-ion battery cathode. They assembled computational models to see how the batteries degraded over time, and learned that variation between battery areas encouraged many structural changes. The results suggested modifying the carbon binder domain structure network could be key to designing longer-lasting batteries.
|
|