Welcome to the October 5, 2022, 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."
|
|
|
Nobel Prize in Physics Awarded to Three Scientists for Work in Quantum Technology
The New York Times
Isabella Kwai; Cora Engelbrecht; Dennis Overbye
October 4, 2022
Researchers Alain Aspect, John F. Clauser, and Anton Zeilinger have been awarded the Nobel Prize in Physics for what the Nobel Committee for Physics called "groundbreaking experiments using entangled quantum states, where two particles behave like a single unit even when they are separated." The committee noted that their work "laid the foundation for a new era of quantum technology." At the Lawrence Berkeley National Laboratory in 1972, Clauser measured the polarizations of photon pairs and observed a correlation, demonstrating that the proton pairs were entangled. Clauser's work was followed by experiments by Aspect at the University of Paris in 1982 and Zeilinger in 1998. The three scientists are credited with establishing the violation of Bell inequalities and pioneering quantum information science.
*May Require Paid Registration
|
White House Unveils AI 'Bill of Rights'
Associated Press
Garance Burke
October 4, 2022
On Oct. 4, the Biden administration released the Blueprint for an AI Bill of Rights that offers guidelines for the U.S. government to protect citizens' digital and civil rights in a world that increasingly depends on artificial intelligence (AI). The blueprint, which does not lay out specific enforcement actions, was developed in consultation with various agencies and includes feedback from technologists, civil society groups, businesses, and industry researchers. The White House Office of Science and Technology Policy's Alondra Nelson said, "This is the Biden-Harris administration really saying that we need to work together, not only just across government, but across all sectors, to really put equity at the center and civil rights at the center of the ways that we make and use and govern technologies. We can and should expect better and demand better from our technologies."
|
Walgreens Turns to Prescription-Filling Robots to Free Up Pharmacists
The Wall Street Journal
Sharon Terlep
October 2, 2022
U.S. pharmacy chain Walgreens is deploying robots to fill out prescriptions at drugstores amid a nationwide shortfall of pharmacists and pharmacist technicians. The robot arms operate at automated drug-filling centers, sorting and bottling pills, and placing them on conveyor belts. The goal is to free up pharmacists to provide medical services like vaccinations, which are a new and expanding revenue channel. Walgreens says the technology slashes pharmacist workloads by at least 25%, and will save the company over $1 billion annually. Pharmacy automation software and technology supplier iA provides the robots, and iA's Alecia Lashier said more drugstores are looking to automate and centralize drug fulfillment.
*May Require Paid Registration
|
Computer Made from Magnets Can Recognize Handwritten Digits
New Scientist
Karmela Padavic-Callaghan
September 30, 2022
Researchers led by Tomoyuki Yokouchi at Japan's University of Tokyo have programmed a computer made from magnetic materials to recognize handwritten digits. The researchers built the prototype from blocks of platinum, cobalt, and iridium layers containing skyrmions, connected to emulate an artificial neural network. They feed the computer a problem by changing the magnetic field to rearrange the skyrmions, inducing the voltage to change across each block, which is translated as the computer's answer. The researchers converted numbers ranging from 0 to 9 into variations in the magnetic field and inputted them into the computer; the device recognized the digits about 95% of the time across 5,000 different tests, outperforming other non-traditional computers employing the same neural network structure by roughly 5%. Reducing the computer's size and accelerating its speed could enable more energy-efficient processing.
|
Algorithms Help Four-Legged Robots Run in the Wild
University of California, San Diego
Liezel Labios
October 4, 2022
University of California San Diego (UC San Diego) researchers helped develop a system of algorithms that allow four-legged robots to walk and run on sand, gravel, grass, and other challenging terrain. The system also enables robots to navigate areas without hitting static or moving obstacles. The system is unique in that it allows robots to use proprioception and vision simultaneously. Said UC San Diego's Xiaolong Wang, "We combine proprioception with computer vision to enable a legged robot to move around efficiently and smoothly—while avoiding obstacles—in a variety of challenging environments, not just well-defined ones." The algorithms combine data collected by depth cameras on the robot's head and sensors on its legs. Because the data may not arrive at the same time, the researchers used multi-modal delay randomization to train a reinforcement learning model that helps the robot anticipate changes in the environment and make decisions quickly as it navigates.
|
3D-Printed, Ultrastrong, Ductile Alloys Form Nanostructures
Texas A&M University
Michelle Revels
September 30, 2022
A multi-institutional team of scientists led by Texas A&M University has defined the microstructure of three-dimensionally (3D)-printed ultrastrong and ductile high entropy alloys (HEAs). The researchers used laser powder bed fusion (L-PBF) to produce the dual-phase materials, which are layered in a self-organized nanostructure. The L-PBF process's rapid cooling rate generates high-density interfaces that contribute to exceptional strength and ductility. HEAs could potentially reduce energy costs. Said Texas A&M's Kelvin Xie, "Our findings show new avenues for materials design, which eventually can lead to lightweight alternatives to many of the materials we currently use in manufacturing."
|
Scientists Use ML to Help Fight Antibiotic Resistance in Farmed Chickens
University of Nottingham (U.K.)
September 29, 2022
Researchers at the U.K.'s University of Nottingham leveraged machine learning to identify diseases at poultry farms, with the goal of predicting and detecting bacterial infection to reduce antibiotic use and prevent antimicrobial resistance (AMR) transfer to humans. The researchers identified 11 types of clinically important antibiotic resistance genes in human and chicken samples taken from a Chinese farm and slaughterhouse. Using a machine learning approach that combined metagenomics data with culture-based methods, the researchers detected the presence of a core chicken gut resistome correlated with AMR circulating in the farms. They also found an association between the AMR-related core resistome and external factors like temperature and humidity. Said Nottingham's Tania Dottorini, "Precision farming, cost-effective DNA sequencing, and the increased adoption of machine learning technologies offer the opportunity to develop methods giving a better understanding and quantification of AMR risks in farming environments."
|
Microsoft Exchange 0-Day Attack Threatens 220,000 Servers
Ars Technica
Dan Goodin
September 30, 2022
Microsoft researchers said numerous servers have been compromised and approximately 220,000 additional servers worldwide are threatened by two critical vulnerabilities in its Exchange application. One is a server-side request forgery vulnerability, and the other enables remote code execution via PowerShell. The unpatched flaws were identified in August by researchers at the Vietnamese security firm GTSC, who found that an Exchange vulnerability was exploited to infect customer networks with malicious webshells. The GTSC researchers said, "After successfully mastering the exploit, we recorded attacks to collect information and create a foothold in the victim's system. The attack team also used various techniques to create backdoors on the affected system and perform lateral movements to other servers in the system." Microsoft is working on a patch for the new vulnerabilities.
|
Rare Electrical Recordings of Human Brain Detail Neural Activity
New York University
October 4, 2022
An international team of neuroscientists used medical data to record human neural activity of visual processing in exceptional detail. The researchers examined volunteer epilepsy patients who had been implanted with electrodes in order to measure seizure-associated neural activity. Recordings made as patients watched pictures on a laptop showed computational models that were designed to explain neural responses in non-human primates are applicable to human brains. The results indicated these models can accurately predict changes in human neural activity for various changes in a visually presented image. "We found that both human and animal brains seem to be using a similar 'toolkit' of neural calculations to make sense of the continuous stream of inputs arriving from our senses," said Iris Groen at the Netherlands' University of Amsterdam.
|
Learning on the Edge
MIT News
Adam Zewe
October 4, 2022
Scientists at the Massachusetts Institute of Technology (MIT) developed an energy-efficient method that allows artificial intelligence (AI) models to continually learn from new data on intelligent edge devices. The technique uses less than 0.25 megabyte (MB) of memory, while other training methods can consume over 500 MB, surpassing most microcontrollers' 256-kilobyte capacity. The process can be used to train a machine learning model on a microcontroller in minutes, and shields privacy by keeping data on the device. "Our study enables IoT [Internet of Things] devices to not only perform inference but also continuously update the AI models to newly collected data, paving the way for lifelong on-device learning," said MIT's Song Han. "The low resource utilization makes deep learning more accessible and can have a broader reach, especially for low-power edge devices."
|
Matter's Internet of Things Standard, Certification Ready
TechCrunch
Christine Hall
October 4, 2022
The Connectivity Standards Alliance has officially approved the Matter Internet of Things standard. The Matter 1.0 standard and certification program aims to make commercially sold devices easily deployable and able to communicate with each other from one place via a local controller. Over 280 technology companies collaborated on requirements and specification development, reference design, test events, and final specification authentication. The program launches with eight authorized test laboratories to assess Matter and its underlying Wi-Fi and Thread network technologies. The certification, specifications, testing tools, and software development kit allow companies to accelerate the market rollout of hardware and innovations designed to improve privacy and security, and to reach a wider consumer base.
|
Team Advances Automatic Forest Mapping Technology
Purdue University
Steve Koppes
October 4, 2022
Purdue University researchers devised an algorithm to enhance automatic forest mapping by digitally separating individual trees from their forests. Purdue's Joshua Carpenter said, "If I could somehow treat all of the points in this [forest data] point cloud like a path of least resistance, that will tell me something about where the tree is located." The algorithm's accuracy exceeds that of current state-of-the-art solutions, according to most metrics. Although point cloud data exhibits an identical structure, the data from the researchers' collection methods (photogrammetry, aerial LiDAR, and ground-based LiDAR) has different anomalies. "The goal is to use all of the different point clouds that are available to make a flexible algorithm," Carpenter said.
|
Screening Tool Could Mean More Effective Idiopathic Pulmonary Fibrosis Treatment
Weill Cornell Medicine
September 29, 2022
A multi-institutional team of scientists designed an automated electronic medical records screening tool to identify patients at high risk of having or developing idiopathic pulmonary fibrosis (IPF). "This tool takes literally zero additional minutes [of patients' time] and can recognize features of the disease before symptoms manifest," said the University of Chicago's Ishanu Chattopadhyay. The software's algorithm calculates an IPF risk score by searching for known IPF risk factors and IPF-connected events in the patient's health records covering the prior two years. The researchers trained the algorithm on an insurance claims dataset spanning millions of U.S. patients from 2003-2018, then validated it on three additional claims datasets. Patients with risk scores exceeding the detection threshold were over 30 times more likely than unscreened patients to be diagnosed with IPF in the ensuing year.
|
|
