Welcome to the January 27, 2023, 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 TechBrief: Policies Needed for Safer Algorithmic Systems
ACM
January 26, 2023
A TechBrief released by ACM's global Technology Policy Council (TPC) warns the spread of algorithmic systems carries with it unaddressed risks. The brief describes perfectly safe algorithmic systems as not possible, but suggests achievable steps can improve their safety and should be prioritized by governments and stakeholders. The council urges the organizational development of a "safety culture that embraces human factors engineering" be "woven" into algorithmic system design. The University of Maryland's Ben Shneiderman, lead author of the TechBrief, said algorithmic systems require safeguards similar to the review of new food products and drugs. Said TPC TechBriefs Committee chair Stuart Shapiro, "As artificial intelligence and other complex consequential systems become more and more prevalent, the need to act to make them safer becomes more and more urgent."
|
Blockchain, JavaScript Most In-Demand Programming Skills
ITPro Today
Nathan Eddy
January 24, 2023
DevSkiller's Digital & IT Skills Report 2023 reported that the most in-demand programming skills are JavaScript, Java, SQL, and Python. The report, based on 209,249 skills assessments sent through the DevSkiller platform to candidates and employees globally, also revealed a 552% jump in demand last year for blockchain programming skills. Said DevSkiller's Tomasz Nurkiewicz, "Either the market still didn't account for a crypto crash or enterprises began using blockchains for non-financial use cases. We don't see a drop yet in 2022 data—quite the opposite—but recruitment processes may be lagging behind the market." Despite concerns about a recession and recent layoffs by major tech companies, Nurkiewicz said companies are still hiring.
|
Soft Robots Harness Viscous Fluids for Complex Motions
Cornell Chronicle
David Nutt
January 23, 2023
Scientists at Cornell University, the University of Delaware, and Israel's Technion-Israel Institute of Technology have enabled a soft robot to achieve complex motions via fluid-impelled actuators. The six-legged robot includes two syringe pumps and a linked series of elastomer bellows with slender tubes running in two parallel columns to facilitate antagonistic push-pull motions. The tubes induce viscosity, which distributes pressure unevenly and bends the actuator into different contortions and motion patterns. Cornell's Yoav Matia created a descriptive model to anticipate the actuator's possible motions induced by different input pressures, geometries, and tube and bellow shapes from a single fluid input. Said Matia, "This work represents a novel scalable framework to manifest control of the material by the material, where otherwise inert structural matter acts on their own behalf."
|
Laid-Off Foreign Tech Workers Set to Find Jobs in Chicago
Bloomberg
Isis Almeida
January 25, 2023
Chicago is courting laid-off foreign technology employees amid a shortage of tech talent, with more than 35 businesses willing to hire workers on H-1B visas in specialties like software engineering. The companies have partnered with the city and the P33 nonprofit on a job-listing website unveiled this week; postings will come from employers already sponsoring H-1B visas. The site currently lists about 900 openings, and organizers see potential for expansion as more companies participate. The effort could help Chicago fulfill its ambition of becoming a tech hub, as well as helping to fill more than 400,000 job vacancies in Illinois.
*May Require Paid Registration
|
Tillage Estimates Get Tech Boost
University of Illinois Urbana-Champaign
Lauren Quinn
January 23, 2023
Researchers at the University of Illinois (U of I) integrated ground, airborne, and satellite imagery to provide real-time mapping of tilled land. The researchers used 6,719 GPS-tagged ground images to train a computer to distinguish between bare ground and crop residue. The computer then was used to interpret and predict hyperspectral images from an airborne sensor, with an accuracy rate of 82%. The researchers upscaled from ground to satellite data, with an accuracy rate of just 22%, and from air to satellite data, which increased the accuracy rate to 67%. Said U of I's Kaiyu Guan, "In remote sensing, we're always trying to link ground-truth data with spectral signals from satellites, but that represents a big scale mismatch. The intermediate-scale hyperspectral data helps to augment ground-truth data because it can provide both high resolution and accuracy. It's a major innovation; nobody has done this in the agricultural world."
|
University of Texas to Offer Online Master's in AI
The New York Times
Natasha Singer
January 26, 2023
The University of Texas at Austin (UT Austin) has launched a large-scale online Master of Science degree program in artificial intelligence (AI). University officials said they intend to train thousands of graduate students in skills like machine learning, natural language processing, and AI health applications, beginning in spring 2024. The program's $10,000 tuition aims to provide more affordable AI education compared to institutions like Johns Hopkins University, whose online AI master's degree costs students more than $45,000. UT Austin's Peter Stone said students will be asked to consider the potential benefits and harms of AI technologies. UT Austin's Don Fussell said the school intends to begin accepting applications for the program in June, and will enroll roughly 2,000 students in the program annually.
*May Require Paid Registration
|
Colliding Particles, Not Cars: CERN Machine Learning Could Help Self-Driving Autos
CERN
Priyanka Dasgupta
January 25, 2023
Researchers at CERN (the European Organization for Nuclear Research, based in Switzerland) and Swedish car-safety software firm Zenseact examined machine learning models to find ways they could help self-driving cars avoid accidents by enabling faster, better decision-making. The researchers selected field-programmable gate arrays (FPGAs), configurable integrated circuits with the ability to execute complex decision-making algorithms in micro-seconds, as the hardware benchmark. By optimizing existing resources, the researchers found they could increase the functionality of FPGAs significantly. They also learned that even when a processing unit had limited computational resources, the FPGAs performed tasks with high accuracy and short latency.
|
Stacking Turns Organic Transistors Up
IEEE Spectrum
Charles Q. Choi
January 25, 2023
Scientists in the U.S. and China have adopted a vertical framework to produce high-performance organic electrochemical transistors (OECTs). Stacking OECTs like sandwiches supports switching speeds that exceed 1 kilohertz, stability across more than 50,000 cycles, less than 100-nanometer-long channel lengths, and transconductances of 200 to 400 millisiemens. The researchers sandwiched the channel between a bottom source electrode and a top drain electrode; the channel consists of a semiconducting ion-permeable compound combined with a polymer that adds robustness and stability, while an electrolyte lies atop the channel and drain electrodes. The researchers said they could manufacture these transistors in a simple, scalable manner using standard fabrication methods.
|
Computer Model of Influenza Virus Shows Universal Vaccine Promise
UC San Diego Today
Michelle Franklin
January 25, 2023
An atomic-level computer model of the H1N1 flu virus developed by University of California, San Diego (UC San Diego) researchers shows "breathing" and "tilting" movements of surface glycoproteins, the main targets of the flu vaccine. Identification of these vulnerabilities could be valuable in designing new flu vaccines and antivirals. The model highlighted a breathing movement by the hemagglutinin (HA) protein that uncovered the epitope, an immune response site previously unknown. The model also revealed tilting movements made by the neuraminidase (NA) protein, allowing researchers to observe how antibodies in convalescent plasma access the epitope. The simulation, run on the Oak Ridge National Laboratory's Titan supercomputer, generated an enormous 160 million atoms worth of data.
|
Tennessee High School Students Build Robotic Hand for Classmate
The Guardian (U.K.)
Ramon Antonio Vargas
January 27, 2023
Students at Henderson high school in Tennessee constructed a robotic hand for classmate Sergio Peralta, whose right hand was not fully formed. Encouraged by teacher Jeff Wilkins, the students designed, three-dimensionally-printed, and sized the prosthesis over four weeks with Peralta's participation. Student Leslie Jaramillio said the project reflected the class's philosophy that "you're supposed to be engineering, coming up with new ideas, solving issues." Henderson principal Bob Cotter said Wilkins and his class challenge themselves to transform abstract concepts "into reality," adding that Peralta's robotic hand "is a testament to the students."
|
How Non-Linear Dynamics Can Augment Edge Sensor Time Series
Tokyo Institute of Technology News (Japan)
January 25, 2023
Engineers at Japan's Tokyo Institute of Technology (Tokyo Tech) and Shinshu University have shown they can augment sensor time series-based classification by using neural networks in a new way. The approach uses the recorded signal as an external forcing to a basic non-linear dynamic system, providing temporal responses to this network disturbance parallel to the original data. Explained Tokyo Tech's Chao Li, "Basically, it is about finding creative and innovative ways of generating additional data to help get the very best performance out of neural networks that necessarily have to be quite small to meet power and size requirements." The researchers considered classification of basic cattle behaviors using a collar-mounted accelerometer, then developed protocols for filtering, preprocessing, and injecting kinematic signals so the simulated dynamic system would accept and respond to them without divergence.
|
Research Team One Step Closer to Printing 3D Models of Organs
CU Boulder Today
Daniel Strain
January 26, 2023
A method developed by University of Colorado (CU) researchers can turn medical images into detailed three-dimensional (3D) computer models. The method utilizes scan data to create maps of organs comprised of billions of volumetric pixels (voxels). The researchers used custom software to transform a standard 3D Digital Imaging and Communications in Medicine (DICOM) file produced by computed tomography and magnetic resonance imaging scans into voxels, which permits organs to be sliced into tiny cubes without losing any information about them. CU Boulder's Robert MacCurdy said, "Surgeons are constantly touching and interacting with tissues, so we want to give them models that are both visual and tactile and as representative of what they're going to face as they can be."
|
Low-Cost Sensor Detects Heavy Metals in Sweat
Agencia FAPESP
Julia Moióli
January 25, 2023
A portable sensor engineered by researchers at Brazil's University of São Paulo (USP), Germany's University of Munich, and Sweden's Chalmers University of Technology from simple materials can detect heavy metals in perspiration. The base of the device is polyethylene terephthalate, on top of which is conductive flexible copper adhesive tape, a label with the sensor printed on it, and a protective layer of nail varnish or spray. The sensor is linked to a potentiostat that reads each metal's concentration, with the result displayed on a computer or smartphone using application software. Said USP's Paulo Augusto Raymundo Pereira, “The world needs flexible sensors that are easily, cheaply, and rapidly mass-produced, as our device is, for on-site detection, continuous monitoring and decentralized analysis of hazardous compounds."
|
|
