Welcome to the December 6, 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."
|
|
|
Climate Change Data Deluge Sends Scientists Scrambling for Solutions
The Wall Street Journal
Robert Lee Hotz
December 5, 2021
Scientists are racing to store, analyze, and preserve massive volumes of climate change data collected by sensor-equipped satellites, aircraft, and other instruments. Officials managing the main U.S. repositories of climate-related data expect their total volume to balloon from about 83 petabytes to over 650 petabytes in the next decade. The managers are moving that information into the cloud, developing new analytical techniques, and adapting a standard format for the data, no matter its origin. Data managers also are restructuring digital climate information to ease autonomous access for machine learning and artificial intelligence systems, which could detect nuanced patterns that humans miss.
*May Require Paid Registration
|
The Power of Graphs
USC Viterbi School of Engineering
Ben Paul
December 3, 2021
University of Southern California Viterbi School of Engineering (USC Viterbi) researchers have shown a graph of a network can yield insights on individual nodes or node groups, and their function over space and time. The researchers used a new algorithm on a partially observed graph to produce a compressed representation of insect neuronal networks, and to differentiate among brain regions inapplicable to traditional network science metrics. They learned not just how insects distinguished between different things they encountered, but also their classifications of food, danger, or mates. Said USC Viterbi's Paul Bogdan, "This is a tool to mine the biological complexity of the brain, discover its network generators (or the rules by which neuronal networks emerge for accomplishing a cognitive task), to mimic how it functions, and recreate that process for machine learning."
|
Nuro, 7-Eleven Launch CA Autonomous Delivery Service
Reuters
Jane Lanhee Lee; Hyunjoo Jin; Paresh Dave
December 1, 2021
Self-driving technology startup Nuro said it is launching California's first commercial autonomous delivery service with convenience store operator 7-Eleven, using converted Toyota Priuses manned by safety drivers to monitor the technology. "The Priuses are the first step in Nuro's multi-phase approach to introducing commercial autonomous delivery to select service areas," said Nuro, adding that the company and 7-Eleven will jointly decide when to introduce Nuro’s R2 driverless vehicles, which they said will happen "as soon as possible." The service will be made available to customers in Mountain View, CA, while Nuro’s autonomous Priuses and R2s already are delivering Domino's Pizza products in Houston, TX.
|
Research Twists Elusive 'Exotic' Quantum Particles
University of Nottingham (UK)
December 2, 2021
Research by scientists at the U.K.'s University of Nottingham, Germany's Technical University of Munich, and Google Quantum AI yielded new insights into "exotic" quantum particles called anyons that could improve the operation of quantum computers. The researchers employed a highly controllable quantum processor to model the ground state of a toric code Hamiltonian. They programmed Google's Sycamore quantum processor to simulate two-dimensional states of quantum matter, and used a quantum algorithm to achieve a state with topological order. Modeling the generation of anyon excitations and twisting them around one another confirmed long-range quantum entanglement, and such topologically ordered states can be tapped to improve quantum computers by facilitating new error correction techniques.
|
Automating the War on Noise Pollution
Bloomberg CityLab
Linda Poon
December 2, 2021
Paris soon will start testing Medusa noise sensors in two neighborhoods, to track and reduce noise pollution. Each device uses microphones to detect and quantify noise levels, and cameras to help authorities track down culprits; French newspaper Liberation said the test will assess the potentials and constraints of automated noise pollution countermeasures. Meanwhile, New York University researchers have been studying New York City's soundscape to engineer a sensor network that uses machine learning to help officials better address noise complaints and proactively establish policies to minimize sound pollution-causing activity. The researchers have partnered with the city's Department of Environmental Protection (DEP) to test a sensor network that can stream real-time data on neighborhood decibel levels and identify sources of noise disturbance to the DEP.
|
Using Statistics to Aid the Fight Against Misinformation
American University
Rebecca Basu
December 2, 2021
A machine learning-based statistical model developed by American University (AU) researchers can detect misinformation in social posts while addressing issues of algorithmic transparency. AU's Zois Boukouvalas said, "We would like to know what a machine is thinking when it makes decisions, and how and why it agrees with the humans that trained it." The researchers trained the model on tweets labeled as either misinformation or real, and included pre-defined rules about language used in misinformation; nuances in human language and linguistic features associated with misinformation also were used in the training. The model correctly classified a test set of 112 real and misinformation tweets with nearly 90% accuracy, "while offering transparency about how it detected the tweets that were misinformation," Boukouvalas said.
|
Supermarkets Grapple with Checkout-Free Stores
Financial Times
Jonathan Eley; Ian Johnston
December 1, 2021
J Sainsbury, the first supermarket chain in Europe to license Amazon's "just walk out" technology, recently opened a checkout-free shop in London's Holborn district, where Amazon and Tesco also operate checkout-free stores. Sainsbury's store is the first in which Amazon's technology was retrofitted to an existing store. A QR code is needed for customers to enter the store, after which they need only place their items in a bag and exit. The customer's purchase is determined using weight sensors, motion detectors, cameras, and software, and they are billed automatically as they leave. Michael Gabay of Trigo, the Israeli retail technology company that provided the technology for the Tesco store, said the technology could be rolled out in rural areas where the costs of running staffed stores pose challenges for retailers.
*May Require Paid Registration
|
These Researchers Wanted to Test Cloud Security. They Were Shocked by What They Found
ZDNet
Danny Palmer
December 1, 2021
Cybersecurity researchers at Palo Alto Networks created a honeypot comprised of 320 nodes around the world to determine how quickly they would be compromised by hackers. The researchers found that 80% of the honeypots were compromised within 24 hours, and all had been compromised within a week. SSH was the most attacked application, with each SSH honeypot compromised an average of 26 times per day; one was compromised 169 times in one day. Palo Alto Networks' Jay Chen said, "The fact that attackers could find and compromise our honeypots in minutes was shocking. This research demonstrates the risk of insecurely exposed services." Chen said the findings underscore "the importance of mitigating and patching security issues quickly."
|
High-Speed Cell Holography Spots Beacons of Disease
Duke Pratt School of Engineering
Ken Kingery
November 30, 2021
Duke University scientists have engineered a holographic system that can image and analyze tens of thousands of cells per minute to identify disease indicators. Sample cells can be washed off the collector into a biocompatible solution and deposited into a microfluidic chip; the sample flows into channels under a line camera. The camera reads the illuminated cells' topography and calculates their features, which when combined with data points and deep learning, can serve as disease-spotting markers. The system differentiated between healthy and cancerous or carcinogen-exposed pre-cancerous cells with 98% to 99% accuracy. Duke's Cindy Chen said, "The idea is that, by having this wealth of quantitative data, you can separate the cells better than if you were using one single metric."
|
Investors Snap Up Metaverse Real Estate in Virtual Land Boom
The New York Times
Debra Kamin
November 30, 2021
A digital land boom is on the horizon as investors purchase concert venues, shopping malls, and other properties in the metaverse. Digital currency investor Grayscale estimates the value of the global market for goods and services in the metaverse will hit $1 trillion soon. Blockchain technology firm Tokens.com recently purchased 50% of the virtual real estate company Metaverse Group and broke ground on a tower in virtual Decentraland in Crypto Valley, the Silicon Valley of the metaverse. The tower could generate revenue from leases and advertising for luxury brands in the metaverse. Tokens.com made the biggest land deal in metaverse history last week, a $2.5-million investment in an area in Decentraland's fashion district that it plans to develop into a virtual commerce hub for luxury fashion brands.
*May Require Paid Registration
|
Discovery Opens the Way for Brain-Like Computers
University of Gothenburg (Sweden)
November 29, 2021
Researchers at Sweden's University of Gothenburg and Japan's Tohoko University successfully combined a memory function with a calculation function in the same component, the first time researchers have been able to link oscillator networks and memristors. Gothenburg's Johan Åkerman said, "These components work more like the brain's energy-efficient neural networks, allowing them to become important building blocks in future, more brain-like computers." Åkerman explained, "The more energy-efficiently that cognitive calculations can be performed, the more applications become possible. That's why our study really has the potential to advance the field."
|
ML Model Reduces Uncertainty in Detection of Breast Cancer
News-Medical Life Sciences
December 1, 2021
A machine learning model developed by researchers at Michigan Technological University (Michigan Tech) can reduce the risk of false positives and false negatives by evaluating the uncertainty in its predictions as it classifies tumors as benign or malignant. The model can classify images more quickly than a human expert, and refers images to a human expert when it is less confident in its predictions. Said Michigan Tech’s Susanta Ghosh, "Despite the promise of machine learning-based classification models, their predictions suffer from uncertainties due to the inherent randomness and the bias in the data and the scarcity of large datasets. Our work attempts to address these issues and quantifies, uses, and explains the uncertainty."
|
|
