Welcome to the November 11, 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."
|
|
|
Geometry Reveals Tricks Behind Gerrymandering
Scientific American
Manon Bischoff
November 10, 2022
Researchers explored geometry to uncover the mechanics of gerrymandering voting districts for electoral gain. Arriving at the best possible division of districts so each party has the same likelihood of gaining a seat is a so-called NP problem, which computer scientists and mathematicians have long considered beyond the capability of ordinary computers. Researchers at the University of Illinois at Urbana-Champaign developed an algorithm that divides maps into voting districts based on the rules set by the given state. In applying that algorithm to voting results in Maryland in 2011, they noted that most of the 250 million results gave Democrats an advantage, indicating both natural conditions and district requirements automatically put Republicans at a disadvantage.
|
IBM Launches Quantum Computer with 433 Qubits
Reuters
Jane Lanhee Lee
November 9, 2022
IBM’s new Osprey quantum computer boasts 433 qubits (quantum bits), triple the number in its predecessor. IBM's Dario Gil said the company is on pace to develop a quantum computer with more than 1,000 qubits, noting that "we have been designing and engineering the whole architecture for quantum computing based on modularity." Gil said IBM’s Quantum System Two will be “the first truly modular quantum computing system so that you can continue to scale to larger and larger systems over time. Modularity means the chips themselves are going to have to be interconnected to one another." The system, expected to be online by late 2023, would serve as the foundation for "quantum-centric supercomputing,” according to the company.
|
AI Model Can Help Prevent Data Breaches
Imperial College London (U.K.)
Gemma Ralton
November 8, 2022
The QuerySnout artificial intelligence (AI) algorithm designed by researchers at the U.K.'s Imperial College London (ICL) can check privacy-safeguarding systems for potential breaches. QuerySnout can automatically identify attacks on query-based systems (QBS) used by analysts to search data and retrieve aggregate information. The model learns which questions to ask the QBS to obtain answers, then learns to integrate the answers automatically to detect potential privacy bugs. QuerySnout uses machine learning to produce a query-based attack that combines answers in order to expose specific private data, following an evolutionary search to find the correct query sets. ICL's Ana-Maria Cretu said, " QuerySnout finds more powerful attacks than those currently known on real-world systems. This means our AI model is better than humans at finding these attacks."
|
Using Vibrations to Control Swarm of Tiny Robots
Georgia Tech Research
November 9, 2022
Researchers at the Georgia Institute of Technology (Georgia Tech) developed a system to control swarms of microbristle robots (microbots) without onboard sensing. Georgia Tech's Zhijian Hao explained, "Microbots are too small to interpret and make decisions, but by using the collision between them and how they respond to frequency and the amplitude of global vibration actuation, we could influence how individual robots move and the collective behaviors of hundreds and thousands of these tiny robots." The researchers created computational models and a live tracking system to analyze swarm behavior and motion data, in order to comprehend motility-induced phase separation. The researchers then tuned vibration levels to coax the microbots into clusters, or to disperse.
|
Smarter Sensor Sniffs Out Target Gases
KAUST Discovery (Saudi Arabia)
November 10, 2022
A smart chemical sensor developed by researchers at Saudi Arabia's King Abdullah University of Science and Technology (KAUST) uses machine learning to identify certain gases in the air based on the temperature changes they cause when interacting with the sensor. The device features a microbeam resonator (a heated strip of silicon) clamped at both ends, which demonstrates significant sensitivity to various gases that have a heat conductivity lower or higher than the air. The researchers used a microsystem analyzer vibrometer to detect shifts in resonant frequency, which are increased by gases like helium and hydrogen (higher thermal conductivity) and decreased by gases like argon (lower thermal connectivity).
|
Chinese Chip Designers Slow Processors to Avoid U.S. Sanctions
Ars Technica
Qianer Liu; Ryan McMorrow; Nian Liu
November 7, 2022; et al.
Alibaba, Biren Technology, and other Chinese chip designers are taking steps to avoid U.S. sanctions by changing their chip designs to lower processing speeds. The sanctions cap the processing power of any semiconductor shipped into China without a license. Sources say further production has been halted while design changes are made as a result of the new rules. Because nearly all semiconductor fabrication plants use U.S. components or software, the rules extend to third-country chip manufacturers and are equivalent to an embargo on high-end processors entering China. Said Paul Triolo at consulting group ASG, "Attempting to freeze a country in place for a technological level of hardware is a big deal. That is what the U.S. is trying to do by restricting sales and closing off the manufacturing road map to get to these advanced levels of hardware."
|
Hacking the Metaverse
Louisiana State University
November 8, 2022
Louisiana State University (LSU)'s Abe Baggili and colleagues tested the security of immersive virtual reality (VR) and X-reality (XR) systems for weaknesses. The researchers hacked a popular XR application used to watch movies with others in a virtual room, and found they could hijack a user's VR headset, view their screen, activate their microphone, and install a virus on their computer invisibly. The virus infected other users who interacted with the compromised user, while the researchers also could enter the virtual room to eavesdrop. They notified the app's developer, which accepted their recommendations. Former LSU researcher Martin Vondrácek said the team open-sourced its hacking software, example exploits, and vulnerability signatures "to improve the state of the art of vulnerability detection and prevention in VR."
|
Engineers Model Moving Cars, Tires
Rice University News
Mike Williams
November 8, 2022
Mechanical engineers at Rice University and Japan's Waseda University simulated moving cars and tires in high resolution by enhancing computational fluid dynamics models of airflow. Rice's Tayfun Tezduyar said the models, which incorporate road contact and tire deformation, expose previously unseen aerodynamic interaction between the front and rear tires. "The simulation is also unique because the level of accuracy reached in representing the car and tire geometries and flow patterns has never been previously attained," Tezduyar observed. He explained the models around the tires required 24 times the computational intensity of those used in earlier experiments.
|
Amazon Launches Warehouse Robot That Can Do Human Jobs
Financial Times
Dave Lee
November 10, 2022
A new warehouse robot rolled out by Amazon in order to automate more jobs as the company seeks to reduce logistics costs. Amazon said the Sparrow robotic arm is the first robot that can "detect, select, and handle individual products in our inventory," jobs once performed only by warehouse employees. The robotic arm identifies and picks up small objects with the help of computer vision technology and suction cups. Amazon, which has developed 700 new robotics-related job "categories," said Sparrow will "benefit" employees by allowing them to focus more on less-repetitive warehouse tasks. Said Amazon's Joe Quinlivan, "Robotics technology enables us to work smarter — not harder — to operate efficiently and safely."
*May Require Paid Registration
|
Engineers Create Algorithm for Powerful Imaging in Low Light
UCLA Samueli School of Engineering
November 8, 2022
An algorithm created by engineers at the University of California, Los Angeles (UCLA) can generate high-resolution images in real time from pictures and videos, even when they are recorded in poor lighting. UCLA's Bahram Jalali said the Vision Enhancement via Virtual-diffraction and coherent Detection (VEViD) algorithm "builds on more than two decades of research in optical physics in our laboratory and has direct applications in augmented reality, night-time driving, biology, and space exploration.” The researchers demonstrated the algorithm's ability to enhance colors in videos, and combined it with a neural network trained in object detection.
|
Warnings Over Growing Threat of 3D-Printed Firearms
BBC News
Chris Vallance
November 9, 2022
Experts warn of the escalating threat posed by three-dimensionally (3D)-printed firearms. Last month, officers from the Specialist Crime Command of the London-area’s Metropolitan Police discovered what they claimed was a "suspected makeshift 3D firearms factory" at a London residence. The U.K. National Crime Agency's Matthew Perfect said the guns' 3D-printed components currently form 80% to 90% of the “latest 3D weapons,” while metal parts such as the gun’s barrel require more traditional fabrication, so the guns can still be identified by "normal security processes." Perfect said there were 21 3D-printed weapon seizures since the start of 2021, a "significant increase" from previous years.
|
Atomistic Experiments, ML-Inspired Modeling Lead to Better Understanding of 2D Materials
Northwestern University McCormick School of Engineering
Brian Sandalow
November 7, 2022
Northwestern University McCormick School of Engineering's Horacio Espinosa and colleagues say they have advanced experimentation and machine learning-inspired simulation for assessing two-dimensional (2D) materials. The researchers conducted in situ transmission electron microscopy atomistic experiments, comparing 2D materials' atomic structures and material toughness with characteristics predicted by molecular dynamics models using inter-atomic potentials based on ab initio data. Said Espinosa, "The integrated framework provides a robust approach to obtaining intrinsic mechanical properties of 2D materials—in their pristine and defective states—and informs the analysis of device reliability with unprecedented accuracy."
|
|
