Welcome to the September 12, 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."

In January 2021, then-chair of the Coffee County, Georgia, Republican Party Cathy Latham greeted a team of computer experts from SullivanStrickler at the county elections office in Douglas, GA. Computer Experts Urge Georgia to Replace Voting Machines
Associated Press
Kate Brumback
September 9, 2022


In a Sept. 8 letter to Georgia's State Election Board and Secretary of State Brad Raffensperger, a group of 13 computer and election security experts said they should replace the state's Dominion Voting Systems touchscreen voting machines with hand-marked paper ballots prior to the November midterm elections. The letter cited "serious threats" from a breach of voting equipment in Coffee County, which remains under investigation. A documented incident that occurred in January 2021 reportedly involved unauthorized copying of election equipment in the county. The letter said the copying and sharing of election data and software from Coffee County "increases both the risk of undetected cyber-attacks on Georgia, and the risk of accusations of fraud and election manipulation."

Full Article

Ian Char, right, works in the control room of the DIII-D National Fusion Facility. Exploring Reinforcement Learning to Control Nuclear Fusion Reactions
Carnegie Mellon University News
Aaron Aupperlee
September 8, 2022


Carnegie Mellon University (CMU) doctoral candidate Ian Char, the first CMU researcher to run an experiment on the DIII-D National Fusion Facility's tokamak machine, demonstrated that reinforcement learning algorithms can control the rotation of the machine's hydrogen plasma. Char developed two algorithms: one was trained using data from the tokamak on how the plasma reacts, while the other calculates the rate and direction at which to add hydrogen particles to affect the speed of the plasma's rotation. Said CMU's Jeff Schneider, "This work shows a path to using reinforcement learning to control other parts of the plasma state and ultimately achieve the temperatures and pressures long enough to have a power plant. That would mean limitless, clean energy for everyone."

Full Article

Harbor seals lounge on the seaweed-covered rocks of the Maine coast. Facial Recognition May Reveal Maine Harbor Seal Lifestyles
UPI
Brooks Hays
September 7, 2022


Facial recognition software could provide scientists with insights into the behavior of harbor seals in Maine. Colgate University researchers built the SealNet program using a modified version of software designed to recognize primates. Colgate's Krista Ingram said the researchers also combined features from other facial recognition algorithms into their model, then trained SealNet to identify seal faces and distinguish between individual harbor seal. The algorithm was 95% accurate when tested on more than 1,700 photos of over 400 individual seals. Said the University of Maine's Kristina Cammen, "I think [facial recognition software] will give us a better sense of seals as individuals, and that can offer us a better sense of their ecologies and behavioral patterns."

Full Article
Researchers Improve Method for Removing Gender Bias in NLP
Folio (University of Alberta, Canada)
Adrianna MacPherson
September 8, 2022


A new method developed by researchers at Canada's University of Alberta (U of A) eliminates gender bias from text while retaining other critical contextual data in natural language processing models. Word embedding converts words into numbers that researchers can plot on a graph and visualize their relationships to one another, in order to measure gender bias and its elimination. U of A's Bei Jiang said debiasing often removes semantic information that could be needed in future tasks involving the word embeddings. The new method preserves semantic information, and also outperformed leading debiasing techniques in various tasks that were assessed according to word embedding.

Full Article
The Supply Chain Broke. Robots Are Supposed to Help Fix It
The New York Times
Peter S. Goodman
September 7, 2022


Companies facing supply-chain disruptions are investing in robots and automation as solutions. Autonomous mobile robot manufacturer Locus Robotics makes carts that travel through warehouses to complement humans selecting goods off shelves, doing a job that Locus' Kary Zate said people do not want. The company has deployed its vehicles in 200 warehouses globally, and promotes them as a remedy for worker shortages. Meanwhile, Silicon Valley-based Gatik is running 30 driverless delivery vehicles between distribution centers and Walmart outlets in Texas, Louisiana, and Arkansas. Israeli startup SafeMode has developed an application that monitors truck drivers to incentivize performance improvement, while North Carolina-based V-Track uses similar technology to surveil and alert drivers when they engage in unwanted behaviors.

Full Article
*May Require Paid Registration

A 3D-printed light-sensing medical device could help millions of people worldwide with lupus and other light-sensitive diseases. Light-Sensing 3D-Printed Device Could Help People with Lupus
University of Minnesota College of Science & Engineering
September 8, 2022


A three-dimensionally (3D) printed wearable device developed by researchers at the University of Minnesota Twin Cities features an ultraviolet and visible light detector that could be placed on the skin to correlate light exposure with symptom flare ups in people with lupus. The device is comprised of multiple layers of electrodes and optical filters printed on a biocompatible silicone base. Integrated with a custom-built portable console, the device provides continuous monitoring and real-time feedback. Said University of Minnesota's Michael McAlpine, "The light is converted to electrical signals to measure it, which in the future can then be correlated with the patient's symptoms flare ups."

Full Article
Stealthy Shikitega Malware Targets Linux Systems, IoT Devices
The Hacker News
Ravie Lakshmanan
September 7, 2022


Researchers at AT&T Alien Labs have identified a Linux malware that can compromise endpoints and Internet of Things devices via a multi-stage infection chain. After the Shikitega malware is deployed, the attack chain downloads and executes the Metasploit's "Mettle" meterpreter to, among other things, escalate its privileges and launch the Monero cryptocurrency miner. It remains unclear how the initial compromise occurs, but Shikitega can download next-stage payloads from a command-and-control server and execute them directly in memory. The use of a polymorphic encoder makes it harder for antivirus engines to detect the malware. Said AT&T Alien Labs' Ofer Caspi, "Shiketega malware is delivered in a sophisticated way, it uses a polymorphic encoder, and it gradually delivers its payload where each step reveals only part of the total payload."

Full Article
Collaborative ML That Preserves Privacy
MIT News
Adam Zewe
September 7, 2022


Researchers at the Massachusetts Institute of Technology (MIT) and MIT-originated startup DynamoFL have enhanced federated learning to better train a privacy-preserving machine learning model. Federated learning involves hundreds or thousands of users training their own model with their own data on their own device, then transferring the models to a central server that pools them into a better model. The researchers' FedLTN system improves the accuracy of the combined model while shrinking its size, which accelerates communication between users and the central server while ensuring each user receives a model tailored for their environment. FedLTN follows the lottery ticket hypothesis, positing that within large neural network models are smaller subnetworks that can realize the same performance. The researchers reduced model size by nearly an order of magnitude versus other methods, yielding four- to six-fold lower communication costs for individual users, and boosting model accuracy 10%.

Full Article

One side of the John Deere See & Spray Ultimate’s 120-foot sprayer boom, which features 36 cameras. Deere Invests Billions in Self-Driving Tractors, Smart Crop Sprayers
The Wall Street Journal
Bob Tita; Jacob Bunge
September 11, 2022


Agricultural machinery manufacturer Deere is investing billions of dollars to launch self-driving tractors and intelligent crop sprayers this year. Deere CEO John May predicts software use fees will constitute 10% of the company’s annual revenue by 2030. The sprayers use computer vision to differentiate crops from weeds and to more precisely target herbicide sprays. By 2026, Deere aims to connect 1.5 million machines in operation and 500 million acres in use to its cloud-based John Deere Operations Center, which will collect and retain crop data. Last year, Deere bought the Bear Flag Robotics startup to provide software to make older tractors capable of autonomous use. The company is wagering it can alleviate farmers' anxiety of losing control over their equipment by offering software-as-a-service on an as-needed basis for specific tasks.

Full Article
*May Require Paid Registration
Low-Cost Solution Viable for Self-Driving Cars to Spot Hacked GPS
University of Alabama
September 6, 2022


University of Alabama (UA) researchers have developed a cost-effective system that uses existing software code and sensors in self-driving vehicles to prevent GPS signal spoofing, which could send cargo or people to the wrong destination. The researchers developed an algorithm that relies on the vehicle's built-in sensors to detect a change in location in real time and to return the vehicle to the correct route. The researchers used data from the Honda Research Institute Driving Dataset to simulate how self-driving vehicles would respond to a spoofed GPS signal, finding the models to be accurate in detecting spoofs. Said UA's Sagar Dasgupta, "We think this will be one of the security modules in the next generation of self-driving vehicles."

Full Article
Revolutionizing Image Generation by AI: Turning Text into Images
Ludwig-Maximilian University of Munich (Germany)
September 1, 2022


Scientists at Germany's Ludwig-Maximilian University of Munich (LMU Munich) have developed an algorithm that can convert text into images in seconds. The Stable Diffusion artificial intelligence (AI) model can accomplish this with a conventional graphics card and without the need to use supercomputers. The researchers trained the model on the servers of their startup Stability.Ai, and the algorithm refines the essence of billions of training images into an AI model of just a few gigabytes. Said LMU Munich's Björn Ommer, "Once such AI has really understood what constitutes a car or what characteristics are typical for an artistic style, it will have apprehended precisely these salient features and should ideally be able to create further examples, just as the students in an old master's workshop can produce work in the same style."

Full Article

A woman selects an African language on the 'Superphone' made by Cerco. Voice-Operated Smartphones Target Africa's Illiterate
Voice of America News
September 10, 2022


The Cerco company, in partnership with the French telecommunications company Orange, is rolling out a "superphone" equipped with a voice assistant for the approximately 40% of the Ivory Coast's population who cannot read or write. The phone, priced at 60,000 CFA francs ($92), aims to help residents more easily perform everyday tasks like checking bank balances. It features Cerco's Kone operating system and responds to commands in the Ivory Coast's 17 languages, as well as 50 other African languages. Cerco ultimately plans to reach half of the continent's population with the phone. Cerco's Alain Capo-Chichi said, "Various institutions set down the priority of making people literate before making technology available to them. Our way skips reading and writing and goes straight to integrating people into economic and social life."

Full Article
Democratizing Cryptography: The Work of Whitfield Diffie and Martin Hellman
 
ACM Distinguished Speakers Program
 

Association for Computing Machinery

1601 Broadway, 10th Floor
New York, NY 10019-7434
1-800-342-6626
(U.S./Canada)



ACM Media Sales

If you are interested in advertising in ACM TechNews or other ACM publications, please contact ACM Media Sales or (212) 626-0686, or visit ACM Media for more information.

To submit feedback about ACM TechNews, contact: [email protected]