Welcome to the July 8, 2019 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."

Photoshopped people floating in a red space with boxes around them Why Experts Are Worried About the 2020 Census
The New York Times
Chris Hamby
July 3, 2019

Experts are concerned that the 2020 U.S. Census is vulnerable to hacking, software glitches, and online disinformation campaigns, due to the U.S. Census Bureau's growing reliance on digital technology. Soaring costs for traditional census methods, and declining response rates, prompted the bureau to modernize; technologies to be deployed for 2020 include smartphones for fieldworkers, greater use of data culled by other agencies, and fieldworker-tracking and route-guidance software. Problems cited in the run-up to the Census include the Bureau's failure to field-test the technology in rural communities, threatening a lack of knowledge for managing widescale glitches. The Commerce Department's Office of Inspector General also found a cloud security flaw that could render the Bureau powerless to prevent irreparable damage by hackers who access lost user credentials. The Bureau said it is working to identify and counter disinformation campaigns with big technology companies and social media networks.

Full Article
*May Require Paid Registration
New Method Makes Realistic Water Wave Animations More Efficient
IST Austria
July 2, 2019

Researchers at the Institute of Science and Technology Austria (IST Austria) have developed a technique to realistically simulate animated water waves that requires significantly less computational effort. Water wave models typically use either Fourier-based techniques that cannot simulate complicated interactions, or more versatile numerical techniques that support a wider range of simulations, but are less computationally efficient. The IST Austria team's methodology enables the rendering of large-scale realistic animated waves and their interaction with solid objects. Among the models the new technique can realize are scenes of boats moving past islands, and rain droplets striking water.

Full Article
Portugal Inaugurates First Supercomputer
July 5, 2019

The Minho Advanced Computing Center at Portugal’s University of Minho has inaugurated the country's first supercomputer. The system, called "BOB," will enable “ten times the national computing capacity,” according to Portugal’s Foundation for Science and Technology, and will encourage scientific and business cooperation in data science and artificial intelligence. The new Computing Center was designed to power the supercomputer mostly with renewable energy. The system will provide services for research in bioinformatics, climate, maritime safety, fisheries, mobility in cities, and forest fire risk management, according to the Portuguese Ministry of Science. A system called "Deucalion" that will be capable of running 10 billion transactions a second is scheduled to join BOB by the end of next year.

Full Article
AI Invents More Effective Flu Vaccine, Adelaide Researchers Say
ABC Online (Australia)
David Sparkes; Rhett Burnie
July 2, 2019

An artificial intelligence (AI) program created a "turbocharged" flu vaccine, marking the first time a computer program has created a new drug on its own, according to researchers at Flinders University in Australia. The researchers took existing drugs that are known to work, as well as examples of drugs that do not work or have failed, and fed that information to an AI program called Sam. The program generated a suggestion of what might be an effective therapy, which the researchers tested and found that it worked. A 12-month clinical trial of the vaccine will soon be conducted in the U.S., according to Flinders researcher Nikolai Petrovsky.

Full Article

Solar power storage Best Algorithms to Make Solar Power Storage Profitable
IEEE Spectrum
Charles Q. Choi
July 1, 2019

Scientists at the Public University of Navarre in Spain have identified optimal types of solar power-storage management algorithms. The researchers developed models based on a year's worth of power generated by a medium-sized, approximately 100kw solar cell array. An examination of dynamic, quadratic, and linear solar-cell energy management algorithms found dynamic algorithms demanded far more computational power than the others. The researchers found quadratic algorithms offered the optimum balance of accuracy and computational simplicity for solar power applications, boasting computational needs similar to linear algorithms, and also realizing revenues similar to dynamic algorithms for all battery sizes.

Full Article

Crystals produced from creatine phosphate Data Can Now Be Stored Inside Molecules That Power Our Metabolism
New Scientist
Ruby Prosser Scully
July 3, 2019

Brown University researchers have stored information within molecules smaller and less complex than DNA. The team created mixtures, with constituent sugars, amino acids, and other metabolites serving as binary 1s and 0s, to store and retrieve pictures of an Egyptian cat, an ibex, and an anchor. For the ibex, the team used arrays of six different metabolites dotted onto a standardized plate by liquid-manipulating robots; they generated 1,024 dots in which the six metabolites were either absent or present, providing sufficient binary data to encode the 6,142-pixel image. The team retrieved the information with about 99% accuracy, using a mass spectrometer to analyze the chemical mix in each dot. Said Brown's Jacob Rosenstein, "Compared to DNA, our metabolite data has low latency, in that we can write and read datasets quickly from start to finish."

Full Article
How to Evaluate Computers that Don't Quite Exist
Science Magazine
Adrian Cho
June 26, 2019

A team of researchers led by Princeton University's Margaret Martonosi compared quantum computers from IBM, Rigetti Computing, and the University of Maryland (UMD), and found the UMD machine—which uses trapped ions—ran a majority of 12 test algorithms more accurately than the other two systems. The researchers found the five-qubit ion-based UMD machine solved most test problems correctly 90% of the time, compared with 50% or less for the superconducting qubit machines at IBM and Rigetti. The UMD machine has an advantage because every ion interacts with all the others in the system; in quantum computers that rely on superconducting chips (as IBM’s and Rigetti’s devices do), each qubit interacts only with its neighbors.

Full Article
Novel Computer Model Supports Cancer Therapy
University of Luxembourg
July 2, 2019

Researchers in the Life Sciences Research Unit at the University of Luxembourg developed a computer model that simulates the metabolism of cancer cells, and used it to study how combinations of drugs could be used more effectively to stop tumor growth. The researchers fed gene sequencing data from 10,000 patients in the Cancer Genome Atlas (TCGA) of the American National Cancer Institute (NCI) into digital models of healthy and cancerous cells. The models were used to simulate the effects of active substances on cell metabolisms so the researchers could identify which inhibited cancer growth while not affecting the healthy cells. Said University of Luxembourg researcher Thomas Sauter, "Our tool can help with the so-called 'drug repositioning,' which means that new therapeutic purposes are found for existing medication."

Full Article

A picture displaying a phone Code in Chinese Surveillance App Analyzed
Ruhr-University Bochum
Julia Weiler
July 3, 2019

Travelers entering China from Kyrgyzstan (on China’s western border) must install an app on their phones, and now researchers at Ruhr-Universität Bochum (RUB) in Germany have analyzed how the app works. The app scans the phone for about 73,000 specific files, and compiles a report for border officials that includes most recent phone activities, contacts, SMS, and social media accounts, among other types of information. The researchers studied the actual app, as well as two of the app's helper programs that were available only in machine code format. One of these other programs helps the app find out which Chinese social media apps are installed on the phone and which accounts are linked with them. The second program scans the phone for specific files, using a list of 73,315 so-called checksums, which are typically used to verify data integrity. Said RUB's Thorsten Holz, "The app is a surveillance tool used to scan mobile phones for specific information at the border, very fast and very efficiently."

Full Article
Logistics Recruiters Cast Virtual Net for Workers
The Wall Street Journal
Jennifer Smith
July 2, 2019

U.S. logistics firms are using a virtual technique called geofencing to scout potential recruits, focusing on people as they drive by sites, or park their cars in rivals' lots. Geofencing taps global positioning system and smartphone-emitted digital signals to serve up recruitment ads, texts, or push notifications to prospective employees via their mobile devices. For example, the Stericycle medical-waste management company's ManpowerGroup unit posts ads on the Waze navigation-app service, which pop up as someone drives past a company warehouse. Meanwhile, e-commerce logistics firm Radial has honed its digital recruitment campaign to target certain ZIP codes and those it considers likely employees.

Full Article
*May Require Paid Registration
Researchers Cast Neural Nets to Simulate Molecular Motion
Los Alamos National Laboratory News
Nancy Ambrosiano
July 2, 2019

The U.S. Department of Energy's Los Alamos National Laboratory (LANL), the University of North Carolina at Chapel Hill, and the University of Florida demonstrated that artificial neural nets can be taught to encode quantum mechanical laws that define molecular motion, potentially advancing simulations across many disciplines. Said LANL's Justin Smith, "We can now model materials and molecular dynamics billions of times faster compared to conventional quantum methods, while retaining the same level of accuracy." The researchers developed a machine learning technique to build empirical potentials—atomic dynamics descriptions that follow classical physical and Newtonian laws—from data collected about millions of compounds. The transfer learning technique can be applied to new molecules in milliseconds.

Full Article

A man in a library with ancient scrolls How Do You Read Ancient Scrolls Too Brittle to Unfurl? American Scientist May Have Answer
The Washington Post
Chico Harlan; Stefano Pitrelli
July 2, 2019

Ancient papyrus scrolls too fragile to unwrap are being scanned, with the hope of "digitally unwrapping" and reading them. The University of Kentucky's Brent Seales leads the effort, using high-resolution computed tomography (CT) scans, and analytical software. The CT scans generate a set of cross-sectional images to reveal the papyrus' internal contours, as well as surface data. Software extrapolates something readable from that data, mapping out the papyrus' configuration and any detectable text. Adding to the challenge is distinguishing text written in carbon-based ink from the carbonized papyrus, which was damaged by a volcanic eruption.

Full Article
*May Require Paid Registration
2019 SIGGRAPH Conference
ACM Queue Case Studies

Association for Computing Machinery

1601 Broadway, 10th Floor
New York, NY 10019-7434

ACM Media Sales

If you are interested in advertising in ACM TechNews or other ACM publications, please contact ACM Media Sales or (212) 626-0686, or visit ACM Media for more information.

To submit feedback about ACM TechNews, contact: [email protected]