Welcome to the April 20, 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 Reports Best Practices for Virtual Conferences
HPCwire
April 16, 2020
A new report from ACM outlines best practices for replacing live science and technology conferences with virtual ones during the COVID-19 pandemic. The report is a practical guide covering a wide range of topics that conference organizers contend with, including required technology, high-level planning, accessibility, nurturing social interaction, navigation, and finances. The guide was created by a task force that included ACM members with experience organizing online conferences and conducting virtual planning sessions. The task force will periodically update and revise the report, and organizers are encouraged to share their own experiences, or make comments or queries. ACM president Cherri M. Pancake said, "Our hope is that the report will also encourage conference organizers to think about reducing their reliance on face-to-face meetings in the future."
|
Raspberry Pi-Powered Ventilator to Be Tested in Colombia
BBC News
Zoe Thomas
April 13, 2020
Marco Mascorro, a robotics engineer with no prior experience creating medical equipment, developed and posted online plans for a ventilator made from a Raspberry Pi computer and easy-to-source parts. Now, researchers at Columbia's University Hospital of the Pontifical Xavierian University and Los Andes University are preparing to put the machine through a fast-tracked round of tests so that it may be used to help combat the COVID-19 pandemic. The Raspberry Pi computer is critical to the control of the ventilator; it regulates air pressure, opens and closes valves, and can determine whether a patient needs full or partial breathing assistance. Said Mascorro, "The beauty of developing a software-centric system is we can make changes to the processes without doing much to the hardware."
|
With Coronavirus Closing Schools, Here’s How Video Games are Helping Teachers
The Washington Post
Elise Favis
April 15, 2020
The quarantine caused by the COVID-19 pandemic has prompted teachers to utilize popular video games like Assassin's Creed, Minecraft, and Roblox to conduct lessons on a range of topics. Game publishers are facilitating this trend by making their platforms as accessible as possible to educators during the crisis. In 2018, Ubisoft added a new mode to Assassin's Creed: Origins, which is set in ancient Egypt, called Discovery Tour. This mode lets players embark on guided tours through famous historical sites and cities. This mode was recently adapted for Assassin's Creed: Odyssey—set in ancient Greece—complete with additional content like quizzes. Minecraft also comes with an education mode, which Microsoft has made free for educators and students through June 2020 due to the pandemic. Roblox is a platform that lets players create their own video games from scratch. The company has partnered with more than 170 educators from 35 countries to discuss applications for the game.
|
Silicon Chip 'Fingerprint' for Stronger Hardware Security at Low Cost
NUS News (Singapore)
April 15, 2020
Researchers at the National University of Singapore (NUS) have developed a method that allows Physically Unclonable Functions (PUFs) to generate silicon chip "fingerprint" outputs at low cost. The team essentially made PUFs self-repairing and self-concealing through an adaptation that employs on-chip sensors and machine learning algorithms to predict and identify PUF instability. The method adjusts the tunable correction level to the minimum required, and produces a more secure and stable PUF output; this reduces power consumption to the minimum possible, and allows the chip to be used to detect anomalous environmental conditions like temperature, voltage, or noise that attackers exploit. NUS' Massimo Alioto said, "On-chip sensing, as well as machine learning and adaptation, allow us to raise the bar in chip security at significantly lower cost. As a result, PUFs can be deployed in every silicon system on Earth, democratizing hardware security even under tight cost constraints."
|
Scientists Use a Little Math to Print the Strongest Kind of Steel
Popular Mechanics
Caroline Delbert
April 19, 2020
Scientists at Texas A&M University and the U.S. Air Force have developed a method for three-dimensionally (3D) printing the strongest kind of steel. The researchers used a steel powder melted into place by laser, while a mathematical model gauges which laser settings will best reduce defects. By varying the number of laser pulses per second and the laser's power, the researchers adapted a welding model to develop and refine their 3D printing model, and continued to tune the model in subsequent iterations until the final model could predict that certain settings would work well or not. "This framework utilizes the computationally inexpensive Eagar-Tsai model, calibrated with single track experiments, to predict the melt pool geometry,” the researchers wrote. "[T]he new idea is an alternative that saves a ton of time compared to a traditionally iterating or permutating algorithm that can take, well, almost forever."
|
Software Streams Apps to Save Space on Your Phone
IEEE Spectrum
Rina Diane Caballar
April 17, 2020
A group of researchers at Prince of Songkla University in Thailand, Purdue University, the University of Illinois at Urbana-Champaign, AT&T Labs Research, and the Georgia Institute of Technology have developed software to reduce app storage requirements through predictive streaming. The AppStreamer software assumes devices lack space for storing everything locally, and employs cloud storage or servers stationed closer to a network end point like a telecom provider's base station. When an app requires a resource, AppStreamer retrieves that resource from remote storage just before the app needs it; the software only fetches the resources for a specific state if that state has the highest probability of following the current state. Purdue's Saurabh Bagchi said, "Any application which has a large footprint in terms of memory and storage and uses only a small fraction of its overall functionality would be suitable for a technology like AppStreamer."
|
NASA Using Red and Blue 3D Glasses to Drive Mars Rover While Working From Home
Gizmodo
Andrew Liszewski
April 17, 2020
Planners at the National Aeronautics and Space Administration's Jet Propulsion Laboratory (JPL) are remotely piloting the Curiosity rover on Mars while working from home. Without access to JPL's powerful workstations and special three-dimensional (3D) goggles due to quarantine orders, the team must rely on red and blue 3D glasses. While antiquated by today's 3D standards, the cardboard glasses are essentially the same anaglyph 3D technology as the special goggles normally used by the team to plan the rover's movements and more accurately target its robotic arm and probes. The team successfully executed Curiosity's first mission planned outside of JPL's facilities just two days after relocating to home offices.
|
Researchers Design Intelligent Microsystem for Faster, More Sustainable Industrial Chemistry
NYU Tandon School of Engineering
April 13, 2020
New York University (NYU) Tandon School of Engineering researchers have developed a machine learning intelligent microsystem for modeling chemical reactions that could potentially offer faster, more sustainable industrial chemistry. The group combined a custom-designed, rapidly prototyped microreactor with automation and in-situ infrared thermography to study exothermic polymerization. The researchers created and applied artificial neural networks to simulate and optimize zirconocene-catalyzed exothermic polymerization based on experimental results. By coupling efficient microfluidic technology with machine learning algorithms to compile high-fidelity datasets based on minimal iterations, the researchers reduced chemical waste by two orders of magnitude and shortened catalytic discovery from weeks to hours. NYU Tandon's Ryan Hartman said the technique could lead to more efficient design and environmentally friendly plastics.
|
Dark_Nexus Botnet Outstrips Other Malware with New, Potent Features
ZDNet
Charlie Osborne
April 8, 2020
Researchers at cybersecurity firm Bitdefender announced the discovery of a new botnet, dark_nexus, with capabilities that exceed those of most current botnets. Although dark_nexus features code links to the Qbot and Mirai botnets, most of its functions are original, which amplifies its robustness and potency. The botnet has been in existence for three months, with three distinct iterations released and linking to at least 1,372 bots mainly hosted in China, the Republic of Korea, Thailand, and Brazil. Dark_nexus uses credential-stuffing and exploits to compromise machines after discovery; both synchronous and asynchronous modules are in use, and attempt to employ Telnet protocol and predefined credential lists to secure access. Attacks are mostly typical, but Bitdefender called dark_nexus' browser_http_req command, which tries to masquerade the traffic as harmless, browser-generated traffic, “highly complex and configurable."
|
Can't Scout Players in Person? The NFL Turns to Brooklyn Startup
The New York Times
Ken Belson
April 16, 2020
The National Football League (NFL) is using the Brooklyn-based Slants startup's artificial intelligence technology to remotely scout prospective players during the coronavirus restrictions. Slants analyzes video of college football games to identify formations, routes, and tracking metrics so coaches can better assess players. Slants' founders aimed to replicate the statistics that NFL teams cull from their own players, using radio frequency identification (RFID) chips to track their movements, and later teaching the software to compensate for different camera angles and other factors via machine learning. In the past month, teams have been sending Slants video of plays from college games, requesting predictive statistics on players; some teams are using the startup's findings to verify data that scouts collected in games or drills.
*May Require Paid Registration
|
These Fold-Up Robots Fly Just Like Ladybugs
Popular Science
Kate Baggaley
April 15, 2020
Researchers at Seoul National University (SNU) in South Korea have built remote-controlled robots inspired by ladybugs, with wings that snap open, lock, and fold up like origami upon landing. Kyu-Jin Cho and colleagues at SNU's Soft Robotics Research Center constructed fabric wings with artificial plastic veins, and affixed them to robotic insects previously engineered to jump or crawl. The team tested the robots by having one leap from a roof, whip its wings open in midair, and glide down. They threw another robot off the second floor of a building, and after gliding safely to the ground, it refolded its wings and crawled along the Earth. Said Northeastern University roboticist Alireza Ramezani, “The way they are trying to integrate origami designs in this kind of bio-inspired robot is interesting.”
|
Speeding Quantum Computing Using Giant Atomic Ions
University of Nottingham (UK)
April 15, 2020
A team of scientists from the University of Nottingham in the U.K. and Sweden's University of Stockholm (SU) accelerated trapped-ion quantum computing with an experimental approach known as trapped Rydberg ions. SU's Chi Zhang and colleagues used the entangling interaction to perform a quantum computing operation about 100 times faster than trapped-ion systems can typically conduct. Said Zhang, "Our [quantum] gate might allow quantum computers to be scaled up to sizes where they are truly useful." Theoretical calculations performed by researchers at Nottingham and Germany's University of Tubingen confirmed no anticipated slowdown once the ion crystals become larger, highlighting the possibility of a scalable quantum computer.
|
Uber Claims Its AI Enables Driverless Cars to Predict Traffic Movement with High Accuracy
VentureBeat
Kyle Wiggers
April 15, 2020
Researchers at Uber's Advanced Technologies Group have proposed an artificial intelligence methodology to enhance driverless vehicles' traffic-movement predictions. The scene-compliant generative adversarial network (SC-GAN) produces trajectories that follow limitations within scenes, given access to high-definition maps of scenes and detection and tracking systems informed by on-vehicle LiDAR, radar, and camera sensors. The GAN outputs nearby cars' frames of reference, and for each car whose potential future trajectories the GAN predicts, the scene data and map constraints are bundled within an RGB image presented by a mathematical object called a matrix. An SC-GAN deployment in Google's TensorFlow machine learning framework reduced off-road false positives by 50% compared with a baseline, and predicted cars' movements even in fairly challenging edge cases. The researchers said, "The method outperforms the current state-of-the-art in GAN-based motion prediction of the surrounding actors, producing more accurate and realistic trajectories."
|
|
