Welcome to the March 6, 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."
|
|
|
How Computer Modeling of COVID-19's Spread Could Help Fight the Virus
NPR Online
Nell Greenfieldboyce
March 4, 2020
Scientists are using computer models to predict how the global COVID-19 outbreak could spread. Efforts include one team's analysis of airplane flights and travel data to identify the 15 most at-risk cities (a list released in January); another group used the number of infected persons identified outside China to determine the actual number of cases in Wuhan city, where the disease originated, far outnumber official counts. Researchers at the London School of Hygiene & Tropical Medicine in the U.K. modeled the feasibility of controlling the pathogen by isolating sick individuals and tracking down each person they came into contact with; they found a high percentage of infectee contacts would need to be traced for this tactic to be effective.
|
FDA Airs Potential Cybersecurity Vulnerabilities in Medical Devices with Bluetooth Low Energy
U.S. Food and Drug Administration
March 3, 2020
The U.S. Food and Drug Administration (FDA) has alerted patients, medical providers, and manufacturers about potential cybersecurity vulnerabilities in medical devices that use Bluetooth Low Energy (BLE). BLE lets two devices "pair" and share data while saving battery life. "SweynTooth" exploits may allow hackers to wirelessly hijack devices like pacemakers, glucose monitors, and ultrasound probes and crash them, stop them from working, or access functions on them normally reserved for authorized users, with publicly available software. The FDA's Suzanne Schwartz said the agency recommends device manufacturers proactively correct such vulnerabilities via coordinated disclosure and mitigation. The agency also requested that device manufacturers communicate to healthcare providers and patients which devices could be impacted by SweynTooth, and how they could ameliorate associated risks.
|
Honeywell to Roll Out Quantum Computer
The Wall Street Journal
Sara Castellanos
March 3, 2020
Honeywell will introduce an early-stage quantum computer for commercial experiments within about three months, with JPMorgan Chase as the first public user. The new machine is expected to be the world's most powerful quantum computer, based on its expected quantum volume (a measure of the performance of a quantum system) of at least 64. Honeywell’s Tony Uttley anticipates the technology will be used by organizations interested in developing new materials or new trading strategies for financial services, or by those looking to speed up calculations. Marco Pistoia of JPMorgan Chase said he expects to use quantum computing to speed up computing-intensive calculations, including Monte Carlo simulations, which are commonly used to calculate the theoretical value of an option. Quantum computing could also be used in portfolio optimization.
*May Require Paid Registration
|
Software Tool Fosters Quality Control of Genome-Scale Models of Metabolism
Technical University of Denmark
Anders Osterby Monsted; Anne Warme Lykke
March 5, 2020
Scientists at the Novo Nordisk Foundation Center for Biosustainability at the Technical University of Denmark (DTU) collaborated with biotechnology researchers in many countries to develop a software tool to perform quality tests on computational models that predict metabolic responses to genetic changes. The suite of standardized metabolic model tests—called Memote—can uncover specific problems in models for continuous improvement and versioning, using best practices derived from software engineering. Memote can rapidly compare any two computational models to determine which is optimal for a specific host organism, in the hope of improving designs for “cell factories” faster than currently possible. Said DTU’s Moritz Beber, “We conducted a quantitative assessment of thousands of published models. While the majority of them are in a reasonable state, Memote was able to reveal specific problems in all of the models.”
|
NOAA to Triple Supercomputing Capacity in Bid to Dramatically Improve Forecast Accuracy
The Washington Post
Andrew Freedman
March 4, 2020
The U.S. National Oceanic and Atmospheric Administration has contracted with Cray Computing for two new supercomputers, which would triple the computing capacity running the agency's weather forecasting and research models. By the time the machines are operational in 2022, the agency will have a total of 40 petaflops of computing capacity dedicated to running about two dozen operational computer models. Peter Bauer of the European Center for Medium-Range Weather Forecasts says this will bring U.S. forecasting computing power closer to that of his agency. Said the University of Washington’s Cliff Mass, “This enhanced resource would give [NOAA] the capacity to make better forecasts, but it is important to note that the proposed increased resources are far less than needed to provide the American people with state-of-science weather forecasting.”
*May Require Paid Registration
|
Obituary: Jaime Carbonell, Professor at CMU Pioneered Language Technologies Research
Pittsburgh Post-Gazette
Lauren Rosenblatt
March 4, 2020
Carnegie Mellon University (CMU) professor Jaime Carbonell, whom colleagues called the "godfather of language technologies," died Feb. 28 at age 66. Over four decades at CMU, Carbonell received more than $100 million in research grants and contracts, and founded the Language Technologies Institute in CMU's School of Computer Science. Carbonell was one of the first to concentrate on machine translation, and CMU's Robert Frederking said apps like Google Translate are "the descendants of that work." Carbonell in 2015 received the Okawa Prize for outstanding contributions to the information and telecommunications field, and also received a Recognition of Service award from ACM for his stint as president of SIGAI, the special interest group on artificial intelligence, from 1983 to 1985.
|
Majority of Promising AI Startups Still Based in the U.S.
Bloomberg
Susan Decker; Alexandre Tanzi
March 3, 2020
Technology research group CB Insights found that the most promising artificial intelligence (AI) startups remain concentrated in the U.S., constituting 65% of the top 100 such companies worldwide. However, the percentage of U.S.-based AI startup investments has declined from 71% to 39% since 2014, which CB Insights analyst Deepashri Varadharajan attributed to other countries funding AI technology research. CB Insights highly ranked "unicorn" startups privately valued at $1 billion or more, like Butterfly Network, which is building an ultrasound device with AI-assisted diagnostics; Faire Wholesale, which helps local retailers predict goods that will sell best in specific locations; and DataRobot, which helps businesses develop their own AI applications. Varadharajan said these startups are using AI to address a specific industry need.
|
Researchers Identify Cybersecurity Approach to Protect Army Systems
U.S. Army Research Laboratory
March 4, 2020
Researchers at the University of California, Riverside and the U.S. Army Combat Capabilities Development Command's Army Research Laboratory (ARL) have developed an approach to protecting Army systems from attack in ways that don't require much manual intervention. The approach, called SymTCP, can be used to identify previously unknown ways to bypass deep packet inspection (DPI) checks in networked devices. Internet service providers often use DPI checks to prevent malicious attacks from being launched or to censor certain content. The research provides an automated method to identify potential vulnerabilities in the Transmission Control Protocol (TCP) state machines of DPI implementation. Said ARL's Kevin Chan, "This research will improve the security of Army networks in terms of being able to protect against future intrusion and evasion strategies. It has developed an efficient way to find and patch vulnerabilities in future Army network infrastructure."
|
Simple Tool Could Solve One of Democracy's Worst Problems
Fast Company
Mark Sullivan
March 5, 2020
Political scientists at Harvard and Boston Universities used mapping algorithms to help develop a tool for addressing gerrymandering that avoids politically biased redistricting designed to maximize the electoral sway of the majority's base. The researchers proposed what they termed Define-Combine district mapping, a process in which the majority party maps a state by drawing twice the number of desired districts, followed by the minority party recombining neighboring "subdistricts" back into final districts. The minority must predict how the majority will likely draw the subdistricts, and have a corresponding recombination strategy. The researchers algorithmically demonstrated that implementing Define-Combine generates more moderate maps by reducing the majority's advantage.
|
Scientists Develop Computer Program to Map Blood Flow 'Landscape' in Tumors
Johns Hopkins Medicine Newsroom
March 5, 2020
Johns Hopkins University School of Medicine (Johns Hopkins Medicine) researchers have developed a computer program that lets scientists measure structural and functional abnormalities in the blood flow networks feeding tumors, based on imaging data. Johns Hopkins Medicine's Janaka Senarathna suggested the HemoSYS toolkit, which the researchers are allowing scientists to use without cost, could accelerate new therapies that target blood vessels feeding tumors to restrict the flow of nutrients and oxygen, and could lead to more effective delivery of already available medications by mapping the blood flow "landscape." Johns Hopkins Medicine's Arvind Pathak said, "We hope that this tool can be adapted to provide a noninvasive way to analyze the blood flow fluctuations in an individual patient's cancer and help to customize their therapy."
*May Require Paid Registration
|
|
