Welcome to the June 12, 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."
|
|
|
Sound-Based Quantum Computers Could Be Built Using Chip-Sized Device
New Scientist
Karmela Padavic-Callaghan
June 8, 2023
University of Chicago (UChicago) researchers have shown that building sound-based quantum computers is feasible. UChicago's Andrew Cleland and colleagues built a chip-sized device from components of a conductive material that can generate individual phonons for transmission. The device is cooled to 0.01 kelvin (-459 degrees Fahrenheit) to induce quantum effects in the phonons, while a beam splitter causes any striking phonon to assume a quantum superposition state. The researchers also sent phonons from opposite directions into the beam splitter to be manipulated by their respective superposition states. Said Cleland, “Phonons are somehow more tangible, more ‘meaty’ than light, but they have been showing the same quantum behaviors.”
|
3D Computer Modeling Generates a Facial Approximation of Tutankhamun
The Jerusalem Post (Israel)
June 10, 2023
An international team of researchers has reconstructed the skull and soft tissue of the Egyptian pharaoh Tutankhamun using three-dimensional (3D) computer modeling methods. The researchers used tomographic scans of Tutankhamun acquired in 2022, along with "data published in the biomedical and Egyptological literature," to reconstruct his skull. They then produced a 3D facial approximation of the pharoah that was completed using features adapted from the Head of Nefertem, a sculpture thought to be a depiction of the young pharaoh.
|
MethaneMapper Poised to Resolve Underreported Methane Emissions
The Current (University of California, Santa Barbara)
Sonia Fernandez
June 8, 2023
An artificial intelligence-powered hyperspectral imaging tool developed at the University of California, Santa Barbara (UC Santa Barbara) can detect methane emissions in real time and trace them to their source. MethaneMapper analyzes survey images from NASA's Jet Propulsion Laboratory in the range of 400 to 2,500 nanometer wavelengths and identifies the presence of methane amid other hydrocarbons detected during the imaging process. The tool can see both the magnitude and the source of a hydrocarbon plume regardless of topography or landscape, with an accuracy rate of 91%. UC Santa Barbara's Satish Kumar said the curated datasets cover about 4,000 emissions sites.
|
Technology Aims to Prevent Falls for Elderly
BBC News
Aileen Moynagh
June 10, 2023
The Move More Live More project in Northern Ireland aims to prevent falls by the elderly by predicting risks that contribute to falls weeks ahead of time. The project, which involves Ulster University and other local institutions, involves wearing a smart watch that monitors activity levels, sleep, heart rate, and oxygen saturation, as well as providing predictive analytics to detect changes that can indicate an increased risk of falling. Said Vicki Caddy of the Age NI charity. "If there are any changes, the software will pick that up, even before the individual wearer would be aware that there was anything different."
|
Autonomous Robot to Stave Off Spotted Lanternflies
Carnegie Mellon University School of Computer Science
Kayla Papakie
June 8, 2023
Carnegie Mellon University researchers created the TartanPest robot to autonomously control the spread of crop-destroying spotted lanternflies. The robot roams fields and forests, identifying and destroying the insect's egg masses. TartanPest detects egg masses using a deep learning model honed on a dataset of 700 images from the iNaturalist social network, then scrapes them off surfaces with a robot arm mounted to an electric tractor from robotics firm Farm-ng. TartanPest can reduce the chemical pollution of crops while boosting farm efficiency and saving money on labor.
|
Nvidia's AI Software Tricked into Leaking Data
Financial Times
Mehul Srivastava; Cristina Criddle
June 9, 2023
At San Francisco-based Robust Intelligence, researchers found the "NeMo Framework" in Nvidia's artificial intelligence software can be manipulated into leaking private data. The framework enables developers to work with an array of large language models. The researchers were able to prompt language models to bypass safety guardrails. Instructing the system to swap the letter "I" with "J," for instance, triggered the release of personally unidentifiable information from a database. They also replicated Nvidia's example of a narrow discussion about a jobs report to get the model to shift to topics beyond the specific subjects set forth in the system's guardrails. Said Robust Intelligence's Yaron Singer, also a computer science professor at Harvard University, " These findings represent a cautionary tale about the pitfalls that exist."
*May Require Paid Registration
|
Bio-Printing Inside Hydrogels Could Help Understanding of How Cancer Spreads
University College London News (U.K.)
June 9, 2023
Researchers at University College London and Great Ormond Street Hospital in the U.K. worked with colleagues at Italy's University of Padova to develop a process for three-dimensionally (3D) bio-printing "mini-organs" within hydrogels to better understand the spread of cancer. The process can help manipulate the mini-organs' configuration and activities, and coax tissue to grow into "molds" in attempts to mimic organs and disease. The researchers produced hardened gel cages surrounding cancer cells and tracked density-directed changes to their movements to analyze the spread of cancer through different tissues.
|
Algorithm Aims to Project Future of Amazon Rainforest
Agencia FAPESP (Brazil)
Luciana Constantino
June 7, 2023
An algorithm developed by researchers from Brazil's State University of Campinas (UNICAMP) that can forecast the future vegetation of the Amazon rainforest indicates a drier climate that halves rainfall can boost diversity while reducing carbon capture. The CArbon and Ecosystem functional Trait Evaluation (CAETÊ) model simulates natural phenomena using mathematical equations trained on environmental data like rainfall, solar radiation, and carbon dioxide concentrations. UNICAMP's Bianca Fazio Rius explained CAETÊ's predictions primarily showed "that including diversity in vegetation models improves their ability to project ecosystem responses to climate change and enhances their credibility."
|
Secure Information Transfer Using Spatial Correlations in Quantum Entangled Beams of Light
University of Oklahoma
Chelsea Julian
June 5, 2023
A study led by University of Oklahoma (OU) researchers demonstrated that spatial correlations in quantum entangled beams of light can be used to encode information and transmit it securely. They accomplished this using two entangled beams of light, which achieve stronger correlations than classical light and maintain their interconnectedness regardless of their distance apart. The researchers transferred spatial patterns from one optical field to two new optical fields produced via four-wave mixing, a quantum mechanical process. OU's Gaurav Nirala explained, "The encoded spatial pattern can be retrieved solely by joint measurements of generated fields."
|
Self-Teaching Web App Improves Speed, Accuracy of Classifying DNA Among Cereal Varieties
U.S. Department of Agriculture
June 5, 2023
The BRIDGEcereal app, developed by researchers at the U.S. Department of Agriculture's Agricultural Research Service (ARS) and Washington State University, analyzes genomic data available for cereal crops with speed and accuracy. The app also can generate charts that identify patterns for genes of interest. ARS's Xianran Li said, "By simply providing BRIDGEcereal with the sequence of DNA you are interested in, it will complete the search process in less than one minute. And BRIDGEcereal will organize the data it finds and present it to you in easily understood charts that highlight any patterns of where that DNA is." Added Washington State University’s Bosen Zhang, "Researchers will find BRIDGEcereal to be an invaluable tool for selecting and prioritizing candidate genes that control specific traits in cereal crops."
|
ANDI Robot Breathes, Sweats, Walks
Interesting Engineering
Amal Jos Chacko
June 11, 2023
Thermal measurement technology company Thermatrics custom-built a thermal manikin robot that breathes, perspires, and walks for Arizona State University (ASU) researchers to emulate the human body's thermal functions. The ANDI robot features 35 surface areas outfitted with sweat pores and temperature and heat flux sensors; internal cooling channels carry cool water throughout the robot's body. The researchers are subjecting ANDI to simulated heat-exposure situations within a heat chamber to examine how heat stress impacts the body. They will pair ANDI with ASU's biometeorological heat robot MaRTy to understand human perspiration mechanics better. The team expects to use the collected data and knowledge to help design cooling apparel and exoskeletons for cooling backpacks.
|
Smart In-Memory Light Sensors Perform Image Recognition
KAUST Discovery (Saudi Arabia)
June 5, 2023
A team of researchers from Saudi Arabia's King Abdullah University of Science and Technology (KAUST) and United Arab Emirates' Khalifa University has created smart in-memory digital image sensors capable of image recognition. The researchers infused two-dimensional molybdenum disulfide within a metal-oxide semiconductor capacitor framework supporting a charge-coupled device sensor's charge-storing pixels. The resulting device is optically programmable. The team confirmed the sensor's ability to execute simple binary image recognition, differentiating between images of a dog or a car with 91% accuracy. KAUST's Nazek El-Atab said, "Our long-term goal is to be able to demonstrate in-memory sensors that can detect different stimuli and compute."
|
|
