Welcome to the May 13, 2020 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."
|
|
|
ACM Service Awards Recognize Leaders Who Strengthen the Computing Community
ACM
May 13, 2020
ACM has named three individuals to receive 2019 service awards that reflect their longstanding efforts to strengthen the computing community. ACM named Mordechai Ben-Ari at Israel's Weizmann Institute of Science to receive the Karl V. Karlstrom Outstanding Educator Award for his pioneering textbooks, open source software tools, and research on learning concurrent programming, program visualization, logic, and programming languages. Michael Ley at Germany's University of Trier and Schloss Dagstuhl-Leibniz Center for Informatics was named to receive the ACM Distinguished Service Award for the creation, development, and curation of the DBLP online bibliographic resource. DBLP has made a massive corpus of published computer science research more accessible and useful to the computing field. Finally, ACM named North Carolina State University professor Arati M. Dixit to receive the Outstanding Contribution to ACM Award for contributing to the expansion and diversity of ACM programs in India, especially as chair of ACM-W India.
|
UCLA, Carnegie Mellon Researchers Develop First Real-Time Physics Engine for Soft Robotics
UCLA Samueli Newsroom
May 11, 2020
Researchers at the University of California, Los Angeles (UCLA) and Carnegie Mellon University have adapted refined computer graphics technology, used to simulate hair and fabrics in films and animated movies, into a real-time physics engine to model how soft limbed robots move. The researchers aimed to build the engine using the discrete elastic rods algorithm, which can model the movements of bio-inspired robots and robots in challenging environments. The physics engine can substantially shorten the time required to develop soft robots from conception to application. Said UCLA's Khalid Jawed, "We have achieved faster than real-time simulation of soft robots, and this is a major step toward such robots that are autonomous and can plan out their actions on their own."
|
Global Firms Push U.S. to Research 5G Tech That Would Minimize Chinese Influence
South China Morning Post
Jodi Xu Klein
May 6, 2020
A coalition of 31 global companies is urging U.S. lawmakers to fund research into next-generation technology, particularly open radio access networks (RANs), to shift 5G infrastructure away from costly proprietary hardware primarily controlled by China. The firms announced the establishment of the Open RAN Policy Coalition to support virtual, software-based 5G networks. James Lewis at the Center for Strategic and International Studies in Washington, DC, said an open RAN system "changes the business model for all the telecom suppliers, in that it moves telecom services to software and cloud computing rather than proprietary hardware." The coalition said the infrastructure switch will enable multiple vendors to interchangeably operate on the system, rather than having to rely on a single manufacturer for all equipment.
|
This Was Supposed to Be the Year Driverless Cars Went Mainstream
The New York Times
Cade Metz; Erin Griffith
May 12, 2020
The coronavirus pandemic has disrupted technology companies' plans to fulfill their promise of bringing autonomous vehicles into the mainstream this year. Even before the crisis, accidents and other safety issues made industry players realize driverless car technology was far from ready for the public. Perfecting the technology requires on-road testing with at least two people in each vehicle to avoid accidents, which social distancing restrictions make impossible. The delays further dampen investor enthusiasm for funding the technology's developers, although better-funded companies can tap cash reserves to wait out the pandemic. Some firms prevented from on-road trials, like the Silicon Valley-based Voyage startup, are using digital simulation to virtually test vehicles, but Voyage's Davide Bacchet said, "Simulation is not something you do in a vacuum, without any connection with the real world and real data."
*May Require Paid Registration
|
Preventing AI From Divulging Its Own Secrets
IEEE Spectrum
Jeremy Hsu
May 6, 2020
North Carolina State University (NC State) researchers have demonstrated the first countermeasure for shielding artificial intelligence from differential power analysis attacks. Such attacks involve hackers exploiting neural networks' power signatures to reverse-engineer the inner mechanisms of computer chips that are running those networks. The attack relies on adversaries physically accessing devices in order to measure their power signature, or analyze output electromagnetic radiation. Attackers can repeatedly have the neural network run specific computational tasks with known input data, and eventually determine power patterns associated with the secret weight values. The countermeasure is adapted from a masking technique; explains NC State's Aydin Aysu, "We use the secure multi-part computations and randomize all intermediate computations to mitigate the attack."
|
Intel Confirms Critical New Security Problem for Windows Users
Forbes
Zak Doffman
May 11, 2020
Intel has verified a newly disclosed security flaw in Windows that exposes an apparently critical vulnerability on millions of computers. The Thunderspy exploit reportedly allows a hacker to read and copy all system data by physically wiring into the machine, even if a drive is encrypted and the computer is locked or set to sleep, while leaving no trace. Bjorn Ruytenberg at Eindhoven University of Technology in the Netherlands said the flaw affects all computers with Thunderbolt ports. Although computers shipped in the last year or so are equipped with Kernel Direct Memory Access (DMA) to patch the bug, it is uncertain how many units have this feature enabled.
|
Coronavirus Has Upended Election Security Training with Just Months Before November
The Washington Post
Joseph Marks; Tonya Riley
May 8, 2020
The University of Southern California (USC)'s Election Security Initiative is scrambling to virtually train campaign and election officials across the U.S. before the November elections, an effort upended by the coronavirus pandemic. The project originally aimed to host in-person trainings nationwide and allow attendees to connect with experts at local universities to help them prepare for cyberattacks, disinformation campaigns, and related threats from adversaries. New challenges presented by the pandemic include preparing for more voting by mail, and ensuring officials have mailing, envelope stuffing, and sorting technology to accommodate that spike. The initiative also is holding shorter training sessions via videoconference than it used to host in person, while postponed primaries add another layer of security problems by compounding skeptical voters' vulnerability to disinformation. Initiative executive director Adam Clayton Powell III said, “Security concerns now are more urgent in almost all cases because the virus has really exacerbated security issues.”
*May Require Paid Registration
|
NIST Scientists Create New Recipe for Single-Atom Transistors
NIST
Ben P. Stein
May 11, 2020
Researchers at the National Institute of Standards and Technology (NIST) and the University of Maryland have developed a step-by-step recipe to produce single-atom transistors, a breakthrough that could lead to a new generation of computers with unmatched memory and processing power. The team used these steps to construct a single-atom transistor and fabricate a series of single-electron transistors with atom-scale control over the devices' geometry. The researchers demonstrated they could precisely adjust the rate at which individual electrons flow through a physical gap or electrical barrier in their transistor, known as quantum tunneling. Precise control over quantum tunneling enables the transistors to become "entangled" in a way only possible through quantum mechanics and allows for the possibility of creating qubits that could be used in quantum computing.
|
Predictive Text Systems Change What We Write
Harvard University John A. Paulson School of Engineering and Applied Sciences
Leah Burrows
May 11, 2020
A study by researchers at the Harvard John A. Paulson School of Engineering and Applied Sciences (SEAS) investigating how predictive text systems alter human writing found that captions written with suggestions from such systems were shorter, with fewer unexpected words than captions written without suggestions—specifically, fewer adjectives, like color words. The team compared the impact of predictive text in three scenarios: captions written with suggestions concealed; captions written with programs that always suggest three next words, and captions written with programs that make suggestions only when the system has high confidence in the following word. SEAS' Kenneth Arnold said, "While, for the most part, people wrote more efficiently with predictive text systems, this may have come at the cost of thoughtfulness."
|
Computational Imaging Benefits From Untrained Neural Network
Chinese Academy of Sciences
Zhang Nannan
May 6, 2020
Investigators at the Chinese Academy of Sciences (CAS) in Beijing combined an untrained neural network with knowledge of physics to remove constraints from deep learning-based computational imaging techniques. CAS' Situ Guohai said the new method requires only raw measurement data for an object captured by the detector, and compares it to the physical model of the object. The researchers based the technique on a deep neural network, which is used to fit different mapping functions from many training data pairs, and the free space propagation principle. They applied this method to a lensless quantitative phase imaging problem, which requires reassembling phase data lost in the detection stage; the team successively resolved the information using a single intensity diffraction pattern. Guohai said, "The new approach for phase imaging is a single shot, non-interferometric method, which has great potential in microscopy and optical metrology."
|
Can Mosquitoes Stop Us Going Bump in the Night?
University of Leeds
May 7, 2020
Scientists at the U.K.’s universities of Leeds and Brighton, and the Royal Veterinary College (RVC), working with colleagues at Japan's Chiba University, applied the principles of the male Culex quinquefasciatus mosquito's nocturnal flight to an autonomous quadcopter's collision-avoidance system. Mosquitoes orient themselves when they cannot see or feel surfaces by detecting changes in airflow patterns, using special antennae receptors. Computational fluid dynamics simulations showed the researchers how these receptors detect airflow changes. The team mimicked this function in the quadcopter with a sensor made from an array of probe tubes linked to differential pressure modules, enabling the drone to sense the greatest changes in airflow when approaching surfaces, in order to avoid obstacles. RVC's Richard Bomphrey said, "It could be useful to take some inspiration from mosquitoes to make our machines safer when operating close to buildings or other infrastructure."
|
Researchers Find Bitcoin's Lightning Network Susceptible to Cyberattack
FIU News
Diana Hernandez-Alende
May 11, 2020
Florida International University (FIU) researchers warn that hackers can exploit Bitcoin's Lightning network to launch cyberattacks that include controlling botnets. Lightning, launched in 2017, allows for faster, more affordable transfers of bitcoin cryptocurrency. The researchers created a proof-of-concept botnet called LNBot to commandeer Lightning by exploiting unrecorded transactions. FIU's Ahmet Kurt said, "Since transactions aren't recorded on the blockchain, a botmaster can communicate with the C&C (Command and Control), and would never be discovered because there is no way to trace it back to the original botmaster." Lightning lacks a central model to authorize or reject messages on what can or cannot be passed. Kurt said possible countermeasures include taking down Lightning to prevent future attacks and compromises, and deactivating a C&C server.
|
Technology Sector Shed Record Number of Jobs in April
The Wall Street Journal
Angus Loten; Sara Castellanos
May 8, 2020
A report by information technology (IT) trade group CompTIA estimated that U.S. IT employers lost a record 112,000 jobs in April, based on U.S. Labor Department data. IT job attrition is occurring despite strong demand for technology, as companies race to deploy remote-working during pandemic-driven lockdowns. Nearly 50% of the country's 12 million tech employees work in the enterprise tech sector, with the rest in IT-related jobs, and the total IT workforce constitutes about 8% of U.S. labor. Although enterprise IT hiring by non-tech firms fared better, with 80,000 new jobs in April, these gains were undercut by overall IT job losses. Market researcher Canalys said many companies are sidelining IT projects, including artificial intelligence or automation implementations, while Tom Gimbel at tech staffing firm LaSalle Network expects a rebound after the crisis as employers seek to implement technologies to boost savings and efficiencies.
*May Require Paid Registration
|
Machine Learning Method Streamlines Particle Accelerator Operations
SLAC National Accelerator Laboratory
Erika K. Carlson
April 29, 2020
Researchers at the U.S. Department of Energy's Stanford Linear Accelerator Center (SLAC) National Accelerator Laboratory used machine learning to design a streamlined technique for speeding up experiments with the Linac Coherent Light Source x-ray laser. The resulting algorithm combines machine learning with knowledge about the physics of the particle accelerator by harnessing the Gaussian process, which forecasts the effect a specific tuning has on the quality of the x-ray beam. This also generates uncertainties for its predictions, and the algorithm then decides which adjustments to attempt to yield the biggest improvements. SLAC's Jane Shtalenkova said, "Our ability to increase our tuning efficiency is really ... critical to being able to deliver a beam faster and with better quality to people who are coming from all over the world to run experiments."
|
|
