Association for Computing Machinery
Welcome to the July 18, 2016 edition of ACM TechNews, providing timely information for IT professionals three times a week.

Updated versions of the ACM TechNews mobile apps are available for Android phones and tablets (click here) and for iPhones (click here) and iPads (click here).


Artificial Intelligence Swarms Silicon Valley on Wings and Wheels
The New York Times (07/17/18) John Markoff

Artificial intelligence (AI) and robotics are leading a transformation in Silicon Valley, which is expected to be as significant as the personal computing industry and commercial Internet eras were. "Whenever there is a new idea, the valley swarms it," says Nvidia CEO Jen-Hsun Huang. "But you have to wait for a good idea, and good ideas don't happen every day." Some technologists see innovations such as deep learning establishing a platform for machines with human-level intelligence, and although past efforts in the field have come up short, many investors are confident the renewed interest in AI will bear fruit. There currently are at least 19 companies in the valley designing self-driving vehicles, and more than six kinds of mobile robots are undergoing commercialization. CB Insights estimates AI startup funding has climbed more than fourfold from $145 million in 2011 to $681 million in 2015, while new investments will grow 76 percent year-over-year to $1.2 billion. Singularity University's Paul Saffo says the entrepreneurial enthusiasm for AI is typical of Silicon Valley's tendency to reinvent itself and follow the latest technology trends. "The valley is built on the idea that there is always a way to start over and find a new beginning," he says.
View Full Article - May Require Free Registration | Return to Headlines | Share Facebook  LinkedIn  Twitter 

U.S. Pumps $400 Million Into Next-Generation Wireless Research
IDG News Service (07/15/16) Grant Gross

The U.S. National Science Foundation (NSF) will invest more than $400 million over the next seven years to fund next-generation wireless research to enable super-fast mobile service. U.S. officials expect the investments, announced last week, will accelerate the transition to next-generation 5G mobile service, potentially offering speeds of 10 Gbps and enabling a swift expansion of the Internet of things (IoT). John Holdren, assistant to President Barack Obama for science and technology, says these services will make self-driving cars, an "always on" IoT, smart cities, new virtual reality offerings, and video to assist police, firefighters, and emergency medical responders a reality. The NSF grants include $50 million as part of an alliance with more than 20 mobile companies and trade groups to deploy advanced wireless testing sites in four U.S. cities, which will feature the rollout of small cells to increase signals of high-band, millimeter-wave spectrum. The announcement follows a vote by the U.S. Federal Communications Commission to unlock almost 11 GHz of high-band spectrum to 5G and IoT services. NSF also will invest $350 million over the next seven years on basic research and testing of next-generation wireless technologies. "History has shown us that when we make sustained federal investments in fundamental academic research and in public-private partnerships...we as a nation reap the benefits," Holdren says.

HoloLens Augmented Reality to Foil Hack Attacks in Factories
New Scientist (07/16/16) Paul Marks

An augmented reality (AR) "IT help desk" could enable factories and industrial plants to identify and mitigate cyberattacks on industrial control systems by giving production engineers immediate access to the expertise of security professionals. Cybersecurity researcher Kevin Jones and his team at the Airbus research center designed the system using Microsoft HoloLens AR visors. When network anomalies or machine behavioral changes are detected, the system alerts the remote security team, which instructs local engineers on how to fix the problem through the headset. Jones says this significantly reduces attack response time and protects infrastructure controls. "If you forget everything you know about how to interact with a computer and their limited visualization capabilities, then augmented reality and haptic touchy-feely interfaces make complete sense," says Falanx Cyber Defense's Jay Abbott. "This has to be the direction of travel for the security industry--it will make remote support and operation more intuitive and more efficient." The system currently is being used by Airbus, but the team plans to develop it for the open market. Airbus says it will hold further trials later this year to quantify the speed advantages of AR.

Sandia Labs Researchers Build DNA-Based Encrypted Storage
Dark Reading (07/14/16) Steve Zurier

Researchers George and Marlene Bachand at Sandia National Laboratories are experimenting with using DNA as an encrypted storage medium for archival material. "I see this as a potentially robust way of storing classified information in the future to preserve it for multiple generations," says George Bachand. The Synthetic DNA for Highly Secure Information Storage and Transmission project was inspired by a European Bioinformatics Institute initiative to record all of Shakespeare's sonnets into 2.5 million base pairs of DNA. Bachand says this method could hypothetically store 2.2 petabytes of information in a single gram of DNA. In addition to being obsolescence-proof, Bachand says DNA sequences can reduce the amount of repetition to make DNA synthesis run smoother; they also make brute-force hacking tougher to achieve. Using an algorithm, the Sandia team encrypts a message into a sequence of DNA, and then chemically synthesizes the DNA. The DNA is read via DNA sequencing and translated and decoded using the same algorithm.

NYU Researchers Report Cybersecurity Risks in 3D Printing
NYU Tandon School of Engineering (07/12/2016)

Additive manufacturing, also known as three-dimensional (3D) printing, faces some of the same cybersecurity risks as the electronics industry, according to a team of cybersecurity and materials engineers at the New York University (NYU) Tandon School of Engineering. Since computer-assisted design files do not give instructions for printer head orientation, malefactors could deliberately alter the process without detection, the researchers note. They say an attacker could hack into a printer connected to the Internet to introduce internal defects as the component is being printed. When the researchers introduced sub-millimeter defects between printed layers, they found the defects were undetectable by common industrial monitoring techniques. They report new cybersecurity methods and tools will be needed to protect critical parts from such compromise. The researchers say the best orientation for the printer is one that minimizes the use of material and maximizes the number of parts that can be printed in one operation, when there is no clear directive from the design team. Printing orientation and insertion of fine defects "are possible foci for attacks that could have devastating impact on users of the end product, and economic impact in the form of recalls and lawsuits," says NYU professor Nikhil Gupta.

Can We Protect Against Computers Being Fingerprinted?
University of Adelaide (07/18/16) Robyn Mills

Computer browser fingerprints are often unique to an individual user, and can be monitored, tracked, and identified by organizations and hackers. University of Adelaide researchers are working to find new methods of protecting against the fingerprinting of personal computers, and they are giving members of the community a chance to see their own browser fingerprints. "Fingerprinting on computers is invisible to most people but there are companies out there who are already using these techniques to learn more information about individuals, about their interests and their habits," says University of Adelaide postdoctoral researcher Lachlan Kang. Although computer users are growing more aware of privacy issues, currently there is little that can be done to counter fingerprinting. Kang notes fingerprints built up in between the browsing history and personal information can be pooled in the gaps between those websites. "Simply clearing your browsing history won't make any difference to this, because the information is already out there," Kang says. The researchers want the public's help to better understand which fingerprinting techniques are the most powerful, as this information will help to build defenses against those techniques. "Eventually, we hope that people will be able to protect themselves from being fingerprinted, or tracked without their consent," Kang says.

90 Percent of Software Developers Work Outside Silicon Valley (07/13/16) Michael J. Coren

A recent App Association study analyzed federal government and private sector data to map where software developers live, identifying 223,054 open positions across the U.S. The study found most developers live far from Silicon Valley, and the job openings follow a similar pattern. "You can find places where you didn't expect software developers to be, but they are part of the local economy," says the App Association's Jonathan Godfrey. "It's pretty much everywhere." Godfrey says one reason for this trend is "software developer" is now the most common job in Washington, Utah, Colorado, and Virginia, according to 2014 U.S. Census Bureau data. In addition, the enterprise, finance, and other tech-heavy sectors have been able to expand their footprints in their hometowns, offering salaries that are far above local averages. However, if developers want to be surrounded by other technology professionals, large metro areas such as Boston, the Bay Area, Seattle, and Washington, DC, are the most attractive areas. The App Association built its maps with data from the U.S. Bureau of Labor Statistics, the National Center for Education Statistics, the Census Bureau, the AP College Board, and Glassdoor.

Greater Privacy and Security Measures Needed to Protect Patients' Information in Mobile Health Technology
Dartmouth College (07/13/16)

Mobile health (mHealth) companies must implement greater privacy and security measures to fully realize the benefits of their technology, according to researchers at Dartmouth College. They pose a series of research challenges in the areas of data sharing and consent management, access control and authentication, confidentiality and anonymity, mHealth smartphone apps, policies and compliance, accuracy and data provenance, and security technology. The researchers highlight the need for mHealth systems to provide users with the opportunity to specify how their protected health information will be used, which will prevent mHealth systems from collecting information that extends beyond the clinical setting. Access control and continuous authentication measures, such as building biometric sensors into a device, may be needed to verify that a personal device reporting health-related information is being used by the rightful owner. The researchers also note anonymizing data would help mitigate the risk that remote sensor data could disclose an individual's location and other private information. "We encourage colleagues with research expertise in mobile health, medical devices, and secure computing to engage with these issues and help bring pervasive mobile-health technology to the world," says Dartmouth professor David Kotz.

Breakthrough in Powering Wireless Sensors
Australian National University (07/13/16)

Australian National University (ANU) researchers are making progress toward harvesting renewable or ambient energy from mobile phone base stations to power battery-operated wireless sensors used in a range of industries. The researchers were able to accurately model how much energy it takes to sense and transfer information by wireless sensors, and they are working on further ways to analyze the problem. "A major problem hindering the widespread deployment of wireless sensor networks is the need to periodically replace batteries," says ANU professor Salman Durrani. The researchers found it was feasible to replace batteries with energy harvested from solar or ambient radio frequency sources such as communication towers or other mobile phone base stations, with communication delays limited to less than a few hundred milliseconds. "If we can use energy harvesting to solve the battery replacement problem for wireless sensors, we can implement long-lasting monitoring devices for health, agriculture, mining, wildlife, and critical national infrastructure, which will improve the quality of life," Durrani says.

Duke Researchers Create Videoconferencing App to Prevent Security Leaks
The Chronicle (07/12/2016) Nidhila Masha

A new videoconferencing application developed by Duke University researchers is designed to prevent accidental security leaks during videoconferencing calls. Participants in videoconferencing calls can draw rectangular markers around the segments of the scene they want to make public, and the software will black out the excluded regions, says Duke graduate student Animesh Srivastava. "We specifically designed the rectangle marker because it is easy to detect and fast to detect by our system," notes fellow graduate student Nisarg Raval. The researchers say the privacy software is efficient enough to work with real-time video feeds and can display modified broadcasts without causing a lag. Srivastava and Raval also note the software is user friendly and compatible with multiple third-party apps. Eyeing the commercial market, the Duke team is planning enhancements that include modeling the software to be compatible with multiple operating systems to reach a broader audience. Srivastava says the software currently interoperates with Android and Linux operating systems, but it is possible to modify the software to fit other devices.

Shedding the Fat: ONR Explores Ways to Trim Software Bloat, Improve Security
CHIPS (07/12/16) Warren Duffie Jr.

The U.S. Office of Naval Research (ONR) is designing the Naval Tactical Cloud, an initiative to harness the power of cloud computing and bring big data capabilities to the military. To ensure the Navy's cloud and other computing efforts run more securely and efficiently, ONR is supporting the work of Pennsylvania State University (PSU) professor Dinghao Wu and University of California, Irvine (UC Irvine) professor Harry Xu, who are studying software bloat. A bloated software system contains a larger code base that could lead to more vulnerabilities and greater entry platforms for hackers and cyber terrorists, according to Wu. To solve these issues, the PSU researchers have created JRed, a tool that reads thousands of lines of code in seconds. Using an algorithm, JRed applies predefined rules to the code of software upgrades and identifies and removes bloated, repetitive code. The researchers have shown JRed can shrink software bloat by about 50 percent. Meanwhile, the UC Irvine team has designed Library Auto-Selection, an optimization technique that creates "shadow libraries" that can update existing software by identifying areas of bloat and adding only the code and data needed for an upgrade.

UMass Amherst Research Advances Self-Driving Car Design, Other Shared Human- and Machine-Controlled Systems
University of Massachusetts Amherst (07/12/16) Janet Lathrop

University of Massachusetts, Amherst (UMass Amherst) researchers describe a new approach to managing the challenge of transferring control between a human and an autonomous system. The research, which was conducted using a driving simulator, should help to advance the development of safe semi-autonomous systems (SAS) such as self-driving cars. These systems rely on human supervision and occasional transfer of control between the user and the automated system, according to UMass Amherst professor Shlomo Zilberstein. He is leading a team that is working on new approaches to SAS that are controlled collaboratively by a person and a machine while each capitalizes on their distinct abilities. "Paradoxically, as we introduce more autonomy, people become less engaged with the operation of the system and it becomes harder for them to take over control," Zilberstein says. The researchers applied the theoretical framework to semi-autonomous vehicles using a hierarchical or step-wise approach with two levels of reasoning. The high-level route planning accounts for the occasional need to transfer control, while the actual transfer of control is managed by a detailed "high-fidelity" model that alerts drivers of their expected actions and monitors their reactions. Zilberstein says the research will help validate the new approach with human drivers controlling a self-driving car while performing a variety of tasks.

Abstract News © Copyright 2016 INFORMATION, INC.
Powered by Information, Inc.

To submit feedback about ACM TechNews, contact: [email protected]
Current ACM Members: Unsubscribe/Change your email subscription by logging in at myACM.
Non-Members: Unsubscribe