Welcome to the March 8, 2017 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).

A robotic hand touches a human hand Robots Could Look, Feel "Human" on the Outside
Wireless Design Magazine
Michael Luciano
March 7, 2017


Researchers at the University of Oxford in the U.K. say technology currently is available to enable robots to "wear" human tendons, muscle, and skin. The researchers suggest robots could mold these human tissues as grafts before they are transplanted onto human patients. They say humanoid robots could play a major role in helping people prone to injuring their muscles, tendons, and skin, and propose a "humanoid-bioreactor system" that mimics the human body in framework, dimensions, and mechanics. The grafts would be placed over the robot's endoskeleton, and as the robot interacts with its environment, the grafts would experience natural strains and twists as if they grew on a human body. The researchers say the system could enable patients who need new tissue to interact with the robot that is wearing their future grafts, and it could be the basis for future "biohybrid humanoids."

Full Article

Hooded figure with computer code over face The Darknet Protects Itself by Being More Robust Against Attacks
URV Activ@
March 6, 2017


Researchers at Rovira i Virgili University (URV) in Spain have found the darknet, a hidden part of the Internet that is used for sensitive and often illegal purposes, is especially good at protecting users' privacy and anonymity because it can counter large cyberattacks on its own by spontaneously adding more network capacity. The researchers say the darknet is practically impenetrable because of its unique topology, which is significantly different from the rest of the Internet. The researchers demonstrated their discovery by using public data and network analysis to quantify the resilience of the darknet, and develop a model that shows how information is transmitted using the "onion router" (TOR), a technique that encrypts messages in multiple layers. The URV team found an attack on the darknet's nodes needs to be four times stronger than an attack against the Internet's nodes to succeed.

Full Article

Airplane flies beside an air traffic control tower Computer Linguists Are Developing an Intelligent System Aid for Air Traffic Controllers
Saarland University
March 7, 2017


Researchers at Saarland University in Germany have developed AcListant, a system that listens to air traffic controllers' conversations and makes informed suggestions for their current situation. The system uses automatic speech recognition technology so the controllers do not have to enter new commands themselves. AcListant also incorporates additional information from the "system aid" air traffic controllers use for planning proposes, so the controllers' display will only include commands that match the current situation, while unsuitable commands are filtered out. In addition, AcListant performs a kind of reality check by incorporating current information from the radar, which is used to generate probable word sequences. Other projects the researchers are working on include teaching computers to understand ambiguous expressions, as well as a system that automatically understands questions in different languages. They note that technology could be used for the early diagnosis and detection of depressive episodes.

Full Article
Top Obstacles for Women in Technology
Help Net Security
March 7, 2017


Women face major barriers when working in technology fields, including wage inequality compared to male colleagues, workplace gender bias, a shortage of female role models, and unequal growth opportunities, according to a new ISACA survey. However, ISACA's Jo Stewart-Rattray says progress can be made in ensuring that women are more equitably represented in the technology workforce by providing more opportunities, including career advancement programs. The survey found 75 percent of respondents said their employer lacks a gender leadership development program. In addition, 80 percent of women reported their supervisors are male, and just 8 percent reported never experiencing gender bias in the workplace. The survey also found women want more mentors, role models, and strong networking opportunities. "Women are vastly underrepresented in the global technology workforce," says Stewart-Rattray. "This is not only a societal concern, but also a workforce problem, given the critical shortage of skilled technology professionals faced by many enterprises."

Full Article
Virtual Reality Training for 'Safety-Critical' Jobs
University of Exeter
Alex Morrison
March 6, 2017


Researchers at the University of Exeter in the U.K. are working with experts from the nuclear industry to create a new training and technology organization called Cineon Training. The new organization has developed an immersive, 360-degree training platform using virtual reality headsets to prevent accidents and improve workers' performance. The system uses technology such as eye tracking and physiological monitoring to help understand how people learn and why they make errors. Exeter's Sam Vine says the goal is to use technology, scientific theory, and measurement techniques to train people in a safe environment to perform more effectively, and to provide feedback to trainers. He notes Cineon Training produces computer-generated, 360-degree filmed replications of dangerous training environments, which trainees can experience through their headsets. "Combined with our understanding of the psychology of learning and performing under pressure, we believe this to be a highly effective way to learn and perfect skills," Vine says.

Full Article
IBM Will Unleash Commercial 'Universal' Quantum Computers This Year
Scientific American
Davide Castelvecchi
March 6, 2017


IBM on Monday announced plans this year to release a commercial "universal" quantum-computing service called IBM Q. IBM physicist Jerry Chow says the goal is to cultivate "a community and ecosystem" around the technology, which builds on the insights IBM researchers gained from Quantum Experience, the company's freely available five-quantum-bit cloud computing service. Chow, who leads IBM's quantum-computing laboratory, says the service offers a way for scientists worldwide to build quantum algorithms without a quantum system of their own. He notes about 40,000 users have employed Quantum Experience to conduct 275,000 experiments and produce about 15 research papers. Chow says IBM Q, which will not outperform conventional computers yet, could enable researchers to focus on meeting the challenges of quantum coding. "The real challenge is whether you can make your algorithm work on real hardware that has imperfections," says the Massachusetts Institute of Technology's Isaac Chuang.

Full Article

Turtle standing between yellow lines of a road Steering a Turtle With Your Thoughts
Asian Scientist
March 2, 2017


Researchers at the Korea Advanced Institute of Science and Technology have developed a brain-computer interface (BCI) that can control a turtle using human thought. The interface includes a head-mounted display (HMD) paired with a BCI to immerse the human user in the turtle's environment. The user wears the BCI-HMD system, while the turtle's upper shell is equipped with a "cyborg system," which consists of a camera, Wi-Fi transceiver, computer control module, and a battery. The human operator receives images from the camera mounted on the turtle, which enables the operator to decide where the turtle should move. The operator provides thought commands that are recognized by the wearable BCI system as electroencephalography signals. The system can distinguish between three mental states: left, right, and idle. The researchers say their technology could be used to develop positioning systems, as well as improve augmented and virtual reality systems.

Full Article
How MIT's New AI System Lets You Control a Robot With Your Mind
TechRepublic
Hope Reese
March 6, 2017


Researchers at the Massachusetts Institute of Technology's (MIT) Computer Science and Artificial Intelligence Laboratory (CSAIL) and Boston University have devised a system that uses brain-signal feedback via electroencephalogram to control a robot and correct errors mid-task. Experiments involved a Baxter robot responding to human commands by picking up either paint cans or wires and sorting them into proper bins. The real-time interface monitors brain waves in 10 to 30 milliseconds, and the controller's transmission of an "error-related potential signal" when they see the robot making a mistake prompts its machine-learning algorithms to sense and correct the error. MIT CSAIL's Stephanie Gil says this breakthrough demonstrates that natural thought processes can control robots, which means humans do not have to be "trained" to think in any prescribed way. Gil notes the system could find potential use in factories "where the more dangerous tasks can be performed by robots under the supervision of a human operator."

Full Article
How to Upgrade Judges With Machine Learning
Technology Review
Tom Simonite
March 6, 2017


Software that could better inform a judge's decision to either jail defendants while awaiting trial or let them stay home could reduce the jail population or even the crime rate, according to a new study from the National Bureau of Economic Research. The researchers trained a machine-learning algorithm to predict whether defendants were a flight risk based on their crime history and court records using data from cases in New York City. When tested on more than 100,000 additional cases it was unfamiliar with, the algorithm proved better at predicting defendants' actions after release than judges. The team estimates in New York City, the software could reduce crime by defendants awaiting trial by up to 25 percent without enlarging the jail population, or slash the jail population awaiting trial by more than 40 percent without affecting the crime rate by defendants.

Full Article

A young, middle aged, and old woman Aging Faces Could Increase Security Risks
MSUToday
Jessi Adler
March 1, 2017


Michigan State University researchers led by Anil Jain are studying how facial aging affects the performance of automatic facial-recognition systems and what implications this could have on successfully identifying criminals or determining when identification documents need to be renewed. The researchers found 99 percent of facial images can be recognized as many as six years later. However, the results also showed after six years, recognition accuracy begins to decline due to natural changes that occur to the face. The researchers say this decrease in facial-recognition accuracy is person-dependent, as some people age faster than others due to lifestyle, health conditions, environment, or genetics. The team based their research on two police mugshot databases of repeat criminal offenders with each offender having a minimum of four images acquired over at least a five-year period, for a total of 23,600 images.

Full Article
Super-Fast Parkinson's App Will Track Symptoms More Closely
New Scientist
Timothy Revell
March 3, 2017


Researchers at Birkbeck, University of London in the U.K. have developed CloudUPDRS, a smartphone application that uses deep learning to enable people suffering from Parkinson's disease to quickly test their symptoms at home. The researchers say the app could help Parkinson's patients closely monitor the disease's progression, and learn how lifestyle factors may affect their symptoms. Patients suffer from a range of symptoms, but it is difficult to develop a solid picture of how those factors interact. One of the app's tests measures tremors by asking the user to hold the phone flat on their hand, while another measures gait by getting the user to walk five meters in a straight line and back with the phone in their pocket. The app is equipped with a deep-learning algorithm, which the researchers trained to recognize abnormal actions with an accuracy of 92.5 percent.

Full Article
Game Theory Could Improve Cyberwarfare Strategy
University of Michigan News
Nicole Casal Moore
March 1, 2017


A new framework based on game theory could help policymakers decide whether to retaliate against a cyberattack. Researchers at the University of Michigan, the University of New Mexico (UNM), and IBM recently conducted a study examining how victims should respond to cyberattacks. The researchers used historical examples to illustrate how the "Blame Game" applies to cases of cyber or traditional conflict involving the U.S., Russia, China, Japan, North Korea, Estonia, Israel, Iran, and Syria. They note it can be challenging for governments or organizations to identify the party responsible for a cyberattack, which complicates the strategic decision of when to assign blame. "Our model elucidates these issues and identifies key parameters that must be considered in formulating a response," says UNM professor Stephanie Forrest. She notes the Blame Game offers a series of questions that policymakers can ask as they determine how to respond to a cyberattack.

Full Article
Language Learning Robot Could Advance Autonomous Vehicles, Help Emergency Responders in the Future
Purdue University News
Curt Slyder
February 28, 2017


Purdue University researchers have developed three algorithms that enable a wheeled robot to learn the meanings of words from sentences that describe example paths taken by the robot, to use the words to generate a sentence to describe the path of movement, and to comprehend the sentence in order to produce a new path of movement. The small-wheeled robot, which is equipped with multiple cameras, ran numerous trials on an enclosed course containing several objects such as a chair, a traffic cone, and a table. The sentences describing the path for the robot to take were provided by anonymous online sources, and an operator steered the robot to follow the paths described by the sentences. The researchers say the robot used the algorithms to recognize words associated with objects within the course and words associated with directions of travel based on its sensory data.

Full Article
Smarter Than Their Machines
 
ACM Publications
 

Association for Computing Machinery

2 Penn Plaza, Suite 701
New York, NY 10121-0701
1-800-342-6626
(U.S./Canada)



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: technews@hq.acm.org