Welcome to the September 29, 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."
|
|
|
11-Foot-Tall 3D Printer Is Helping Create a Community
The New York Times
Debra Kamin
September 28, 2021
Builders are constructing homes in a village on the outskirts of Nacajuca, Mexico, using an 11-foot-tall three-dimensional (3D) printer. San Francisco-based nonprofit New Story, Austin, TX-based construction technology firm Icon, and Mexican social housing production company Échale partnered on the project. Controlled by a tablet or smartphone, Icon's Vulcan II printer pours lavacrete, a proprietary concrete compound, into swirled layers; it can complete a single residence in less than 24 hours. The homes have already withstood a magnitude 7.4 earthquake. New Story's Alexandria Lafci said, "We're really looking at the biggest opportunities to have both impact and efficiency gains. There is a very significant gain in speed that you get with 3D printing, without sacrificing quality."
*May Require Paid Registration
|
FAA Technology Aims to Speed Takeoffs of Planes Idling on Runways
The Wall Street Journal
Micah Maidenberg
September 28, 2021
U.S. Federal Aviation Administration (FAA) officials said the agency intends to deploy new software designed by the National Aeronautics and Space Administration (NASA) to more easily estimate when planes can taxi onto a runway and take off. They said the FAA hopes to incorporate the software into a system for managing air traffic at airports in the next few years. The system reportedly reduced delays at North Carolina's Charlotte Douglas International Airport by over 900 hours during a four-year testing period, or an average 15 minutes of wait time for each of approximately 3,600 departing flights. FAA administrator Steve Dickson said the system will improve air-traffic controllers' visibility into data that airlines use to operate flights, enabling better prediction of on-time departures and avoidance of bottlenecks.
*May Require Paid Registration
|
Using AI, Old Reports to Understand Medical Images
MIT News
Steve Nadis
September 27, 2021
Scientists at the Massachusetts Institute of Technology (MIT), Harvard Medical School, and Boston’s Beth Israel Deaconess Medical Center combined artificial intelligence and written radiology reports accompanying medical images to enhance the interpretive abilities of machine learning algorithms. The process involves training a neural network to gauge disease extent using medical images and a physician's severity rating, while a second network performs a similar task for text, and a third integrates text and image data. MIT's Polina Golland said, "When the mutual information between images and text is high, that means that images are highly predictive of the text and the text is highly predictive of the images."
|
Teams Develop Code to Coordinate Robots, Win $535,000 in Space Robotics Challenge
NASA
September 27, 2021
Seven teams that competed in the U.S. National Aeronautics and Space Administration (NASA) Space Robotics Challenge were awarded a total of $535,000 for their development of code to advance autonomous robotics for future space exploration missions. The teams competed in a virtual lunar environment, where their virtual robots were tasked with autonomously locating, collecting, and hauling resources to their "home base" on the Moon. Scores were based on the number of resources the robots identified, excavated, and delivered to the targeted base. The first-place team, Olympus Mons, was awarded $185,000 for its squad of six rovers coordinated by a central system.
|
When Accidents Happen, Drones Weigh Their Options
University of Illinois at Urbana-Champaign
Debra Levey Larson
September 21, 2021
Gauging unmanned aerial vehicles' ability to recover from malfunctions and complete missions safely is central to research by scientists at the University of Illinois at Urbana-Champaign (UIUC). The “quantitative resilience” of a control system attempts to verify such systems' capabilities following an adverse event, according to UIUC's Melkior Ornik. He said that task requires the drone to solve four nested, possibly nonlinear, optimization challenges, and reduces the computation of quantitative resilience to a single linear optimization problem through control theory and two novel geometric results. Ornik added, “This ability to work through when equipment malfunctions has real-life implications.”
|
Risk Assessment Algorithms Can Unfairly Impact Court Decisions
Government Technology
Jule Pattison-Gordon
September 27, 2021
A study by the University of Michigan's Ben Green and Harvard University's Yiling Chen suggests pretrial risk assessment algorithms increase the likelihood that judges will change their priorities in making pretrial decisions. The algorithms use data on previous defendants' outcomes to make forecasts on the given arrestee, presenting them either as a numerical score, or designating them high-, medium-, or low-risk for failure to appear in court, or being arrested again. The researchers found viewing the algorithms' predictions caused participants to consider factors differently and to more highly prioritize the risk of defendants' failure to appear or getting re-arrested, with the result of more inequitable sentencing (since Blacks were more likely to be deemed higher-risk defendants, and to receive harsher decisions than Whites).
|
Simulating Hydrogen Storage Efficiency with High Accuracy
Japan Advanced Institute of Science and Technology
September 27, 2021
Japanese and U.S. scientists modeled hydrogen storage efficiency on silicon-carbide nanotubes more accurately than conventional methods, through the use of diffusion Monte Carlo (DMC) computation. Japan Advanced Institute of Science and Technology (JAIST) researchers employed DMC to produce a model that incorporated van der Waals forces, reconstructing the interaction mechanism by tracking the arrangement of individual electrons. The model allowed the researchers to predict the amount of energy needed to dislodge hydrogen from storage, and its likely distance from the nanotube's surface. JAIST's Kenta Hongo said, "Although the DMC method is computationally expensive, it can be used to clarify the peculiarities [tendencies of prediction error] of each prediction method. This will help us understand which prediction to trust, and also how to modify prediction methods to make them more useful."
|
Technique Speeds Measurement of Ultrafast Pulses
University of Rochester
September 24, 2021
Researchers at the University of Rochester, China's Tianjin University, and Canada's University of Ottawa have developed a time-domain single-pixel imaging (TSPI) technique that speeds the measurement of ultrafast pulses. In TSPI, the photodiode serves as the single-pixel detector in the time domain and is used in conjunction with a programmable temporal fan-out gate based on a digital micromirror device. The technique detected five femtojoule ultrafast light pulses with a temporal sampling size down to 16 femtoseconds. University of Rochester's Jiapeng Zhao said a computational hyperspectral imaging system can be created by combining TSPI with single-pixel imaging, resulting in faster detection and analysis of images at broad frequency bands.
|
Researchers Create Integrated Pipeline for Genetic Screening Studies
News-Medical Life Sciences
Emily Henderson
September 28, 2021
The Sequencing-based Target Ascertainment and Modular Perturbation Screening (STAMPScreen) protocol is an integrated pipeline for genetic screening studies developed by scientists and engineers at Harvard University’s Wyss Institute for Biologically Inspired Engineering, Harvard Medical School (HMS), and the Massachusetts Institute of Technology (MIT) Media Lab. Harvard Medical School's Pranam Chatterjee calls STAMPScreen "a streamlined workflow that makes it easy for researchers to identify genes of interest and perform genetic screens without having to guess which tool to use or what experiments to perform to get the results they want." Harvard and MIT scientists contributed algorithms to analyze and extract information from datasets compiled from next-generation sequencing. Harvard's George Church said, "STAMPScreen will help researchers get to eureka moments faster and speed up the pace of innovation in genetic engineering."
|
Far Fewer Qubits Required for 'Quantum Memory' Quantum Computers
APS Physics
Erika K. Carlson
September 28, 2021
The French Alternative Energies and Atomic Energy Commission's Élie Gouzien and Nicolas Sangouard demonstrated that a quantum memory-equipped quantum computer could calculate with fewer quantum bits (qubits) in its processor than standard architectures. The researchers conceived a two-dimensional grid of qubits connected to a three-dimensionally-arranged quantum memory, and analyzed its ability to find prime factors for RSA integers. Their design could factor a 2,048-bit RSA integer with just 13,436 qubits, while a standard architecture might require 20 million. The quantum memory architecture would require 177 days to accomplish this, compared to about eight hours for a standard architecture, but Gouzien and Sangouard said the smaller number of qubits makes their approach more feasible in the near term.
|
Pandemic Pushes Chinese Tech Giants to Roll Out More Courier Robots
Reuters
Yilei Sun; Yingzhi Yang; Sophie Yu
September 28, 2021; et al.
Chinese technology giants Alibaba, Meituan, and JD.com plan to increase their courier robot fleets fourfold to more than 2,000 by 2022, as the pandemic boosts demand for contactless services and the costs of manufacturing robots declines. JD.com's Kong Qi said, "We want people and vehicles to work better together and not for vehicles to replace people. It is just the most boring section of the delivery guy's work that we will try to replace." Alibaba currently operates over 200 robots and plans to have 1,000 in the field by March, and as many as 10,000 over the next three years. JD.com operates about 200 robots, and plans add as many as 800 to its fleet by year’s end. Meituan now has around 100 delivery robots in use.
|
How AI Completed Beethoven's Unfinished Tenth Symphony
Smithsonian
Ahmed Elgammal
September 24, 2021
Computer scientists, music historians, musicologists, and composers collaborated with the startup Playform AI to apply artificial intelligence (AI) to the task of completing Beethoven's unfinished Tenth Symphony. The researchers taught an AI system Beethoven's entire body of work and his creative process, including the methods he used to develop certain musical forms, how to harmonize a melodic line, how to bridge two sections of music together, how to compose a coda, and how to orchestrate the full composition. Over a period of more than two years, the research team’s efforts yielded two movements (symphonies typically include four), each more than 20 minutes long. The release of the full recording of the Tenth Symphony, as well as its world premiere performance in Bonn, Germany, are scheduled for Oct. 9.
|
|
