Welcome to the June 21, 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."
|
|
Testing AI Fairness in Predicting College Dropout Rate
Cornell Chronicle Tom Fleischman June 17, 2021
Cornell University and University of California, Irvine researchers have found that removing protected student characteristics from machine learning models used to identify college students at risk of dropping out does not improve the accuracy or fairness of the models' predictions. The research compared predictive models with and without such protected attributes using a dataset of 564,104 residential course records and 81,858 online course records. The dataset was used to build 58 student attributes across four categories: student gender, first-generation college status, membership in an underrepresented minority group, and high financial need. The researchers found that including protected attributes had no significant effect on three common measures of overall prediction performance when the model already included commonly used features like academic records.
|
Compact Quantum Computers for Server Centers
University of Innsbruck (Austria) June 18, 2021
A prototype ion trap quantum computer developed by researchers at Austria's University of Innsbruck can fit into two 19-inch server racks commonly used in datacenters. Current quantum computers can fill entire laboratories, but Innsbruck's Christian Marciniak said the team demonstrated that "compactness does not have to come at the expense of functionality." The researchers were able to ensure stability in the compact quantum computer, which can be operated autonomously, and successfully entangled up to 24 ions with the device. Innsbruck's Thomas Monz said, "By next year, we want to be able to provide a device with up to 50 individually controllable quantum bits."
|
Algorithm Uses Mass Spectrometry Data to Predict Identity of Molecules
Carnegie Mellon University School of Computer Science Aaron Aupperlee June 17, 2021
Researchers at Carnegie Mellon University (CMU) and Russia's St. Petersburg State University have developed an algorithm that uses mass spectrometry data from molecules to predict the identity of unknown substances. This could keep researchers seeking new naturally occurring products for use in medicine from wasting time and money by isolating molecules that already are known. There is no database of known molecules to compare with mass spectrometry data, but the algorithm, called MolDiscovery, can predict the identity of a molecule using mass spectrometry data and a predictive machine learning model. Said CMU???s Hosein Mohimani, "Detecting whether a molecule is known or not early on can save time and millions of dollars, and will hopefully enable pharmaceutical companies and researchers to better search for novel natural products that could result in the development of new drugs."
|
Science Denial, Partisanship on Social Media Indicate Where COVID-19 Strikes Next
University of Southern California June 17, 2021
A machine learning-assisted social media analysis by researchers at the University of Southern California (USC) could help predict where COVID-19 will emerge next, based on anti-science views and political ideology. USC's Kristina Lerman said the study determined entirely from social media data that "anti-science views are aligned with political ideology, specifically conservatism." Using 27 million Twitter tweets posted by 2.4 million U.S. users from Jan. 21 to May 1, 2020, the researchers compared public discourse on COVID-19 with epidemiological outcomes. They found that anti-science attitudes were high from January through April 2020 in Mountain West and Southern states that experienced COVID-19 surges.
|
If I Had a Hammer: A Simple Tool to Enable Remote Neurological Examinations
Georgia Tech Research Horizons June 16, 2021
Researchers at the Georgia Institute of Technology, in collaboration with Japan's Tohoku University and NITI-ON Co., have developed a mobile app for use with their "smart" tendon hammer to allow remote deep tendon reflex exams, the first step in identifying neurological illnesses. To determine whether the hammer has hit the correct spot on the tendon to elicit a proper reflex response, the researchers added a small wireless inertial measurement unit to a standard commercially available reflex hammer. The hammer's acceleration measurements from on-tendon and off-tendon locations are run through a classification algorithm, and the patient receives instant feedback as to whether the hammer has hit the correct spot. Physicians grade patients' reflexes by reviewing videos of the tendon tapping and reflex response uploaded by the patient through the app.
|
Smart Tires Hit the Road
The Wall Street Journal Sara Castellanos June 16, 2021
Tire manufacturers Goodyear Tire & Rubber and Bridgestone are launching new smart tire features for last-mile delivery vehicles transporting packages from e-commerce sites like Amazon.com. Goodyear's SightLine solution runs data from a sensor through proprietary machine learning algorithms to capture tire wear, pressure, road-surface conditions, and other variables to forecast flats or other problems days ahead of time. Goodyear's Chris Helsel said SightLine could detect 90% of tire-related issues ahead of time in a test that involved about 1,000 vehicles operated by 20 customers. Meanwhile, Bridgestone Americas is developing an intelligent tire system that combines sensors, artificial intelligence algorithms, and digital twins to predict tire wear and readiness for retreading.
*May Require Paid Registration
|
Amazon Brings Cashierless Tech to Full-Size Grocery Store
GeekWire Taylor Soper June 15, 2021
Amazon has deployed its Just Walk Out cashierless retail system in its newest Seattle-based Amazon Fresh physical grocery outlet, its first use in a full-size store. Just Walk Out utilizes cameras and sensors to log items selected by customers, eliminating the need for checkout lines; shoppers scan their phones when they enter the store, and just "walk out" after loading their basket or cart. The new store, in the Factoria neighborhood of Bellevue, WA, also will feature Amazon One, the retailer's palm-scanning ID system that recently was adopted by Whole Foods stores.
|
Underwater Robot Offers Insight into Mid-Ocean 'Twilight Zone'
Woods Hole Oceanographic Institution June 16, 2021
Scientists at the Woods Hole Oceanographic Institution (WHOI), the Monterey Bay Aquarium Research Institute (MBARI), and Stanford University used the underwater robot Mesobot to track and capture high-resolution images of slow-moving organisms inhabiting the mid-ocean "twilight zone" region. Mesobot uses an array of oceanographic and acoustic survey sensors, and can be operated remotely through a fiberoptic cable linked to a ship, follow pre-programmed missions, or autonomously track targets at depths of up to 1,000 meters (3,300 feet). MBARI's Kakani Katija said, “Mesobot has the potential to change how we observe animals moving through space and time in a way that we've never been able to do before."
|
Autonomous Walking Excavator Can Build Walls, Dig Trenches
New Scientist Adam Vaughan June 18, 2021
A construction vehicle can operate autonomously on rough terrain, thanks to a team of Swiss-German engineers that adapted a walking excavator to perform various tasks. Researchers at ETH Zurich in Switzerland made the prototype Hydraulic Excavator for an Autonomous Purpose (HEAP) autonomous through the use of algorithms, control mechanisms, and Light Detection and Ranging (LiDAR). The 12-ton HEAP was programmed to use an excavator bucket and a two-finger gripper, and was able to construct a four-meter (13-foot)-high stone wall, grab trees for mock forestry work, and dig out a trench containing live ammunition from World War II. ETH Zurich's Dominic Jud said one of the biggest challenges in switching the excavator from human operation to a computer running open source Ubuntu software was reengineering the cabin controls to drive the hydraulic pumps. Jud said HEAP is roughly as accurate as human operators in executing tasks, although not yet as quick.
|
ThroughTek Flaw Opens Millions of Connected Cameras to Eavesdropping
The Hacker News Ravie Lakshmanan June 16, 2021
An advisory issued by the U.S. Cybersecurity and Infrastructure Security Agency (CISA) warns of a major software supply-chain flaw in cloud security provider ThroughTek's point-to-point (P2P) software development kit (SDK), which could allow unauthorized access to the audio and video streams from millions of connected cameras. The flaw stems from insufficient protection when transferring data between the local device and ThroughTek's servers; it impacts ThroughTek P2P product versions 3.1.5 and before, and SDK versions with NoSSL tag. Security firm Nozomi Networks reported the bug in March, warning that vulnerable security cameras could place critical infrastructure operators at risk by compromising sensitive business, production, and employee data.
|
Clearing the Way Toward Robust Quantum Computing
MIT News Michaela Jarvis June 19, 2021
A technique developed by researchers at the Massachusetts Institute of Technology (MIT) removes common errors in the operation of a two-quantum bit (qubit). MIT's Youngkyu Sung and colleagues extended the concept of tunable couplers, which enable on-and-off switching of two-qubit interactions to control operations while maintaining the qubits. The tunable couplers were prone to errors from residual unwanted interactions between the qubits and between the qubits and the coupler, and the new method reduces such errors. The researchers tapped higher energy levels of the coupler to neutralize the interactions. MIT's William D. Oliver said the technique realized 99.9% fidelity for the two major two-qubit gate models, the Controlled-Z gates and iSWAP gates.
|
Biomimetic Resonant Acoustic Sensor Detecting Far-Distant Voices Accurately to Hit the Market
KAIST (South Korea) June 14, 2021
Researchers at South Korea's Korea Advanced Institute of Science and Technology (KAIST) have developed a bioinspired flexible piezoelectric acoustic sensor with a multi-resonant ultrathin piezoelectric membrane that acts like the basilar membrane of the human cochlea to achieve accurate and far-distant voice detection. The miniaturized sensor can be embedded into smartphones and artificial intelligence speakers for machine learning-based biometric authentication and voice processing. Compared to a MEMS condenser microphone, the researchers found the speaker identification error rate for their resonant mobile acoustic sensor was 56% lower after it experienced 150 training datasets, and 75% lower after 2,800 training datasets.
|
|