Welcome to the October 20, 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).
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."
|
|
|
'Low Cost Android' to Study the Brain
SINC (Information and Scientific News Service)
Laura Chaparro
October 19, 2017
Researchers working on the European Union-funded MoCoTi project say they have designed the prototype of an android that learns how to actuate its own limbs. The team says the device is equipped with an artificial brain that controls a tendon-driven robotic arm, and it could be the first step toward the development of low-cost humanoid robotics. "This is possible because its modular design permits a relatively efficient mass production," says Technical University of Munich researcher Christoph Richter in Germany, who is one of the leaders of the MoCoTi project. The researchers used the Myorobotics system to imitate the muscles and articulations of the human arm and give the robot increased mobility. The team then designed an artificial cerebellum to control the orders from the locomotor system. "We reproduce the most important neurons, their connectivity and--importantly--their adaptation and learning in our real-time simulation," Richter says.
|
For $1,000, Anyone Can Purchase Online Ads to Track Your Location and App Use
UW News
Jennifer Langston
October 18, 2017
Researchers at the University of Washington (UW) suggest it could cost only about $1,000 for someone to buy and target online ads in order to monitor the location of others as well as their application use. The team found individual ad purchasers can, under certain conditions, observe when a person visits a predetermined sensitive location within 10 minutes of their arrival, as well as tracking a person's movements across a city by serving location-based ads to the target's phone. By establishing a grid of location-based ads, adversaries can keep tabs on targets' movements if they have opened an app and stay in a location long enough for the ad to be served. "This is an issue that the online advertising industry needs to be thinking about," says UW professor Franzi Roesner. The UW team will present their findings on Oct. 30 at the ACM Workshop on Privacy in the Electronic Society (WPES 2017) in Dallas, TX.
|
CSforAll Announces Computer Science Pledges From Over 170 Organizations
Education Week
Sarah Schwartz
October 19, 2017
The CSforAll Consortium recently announced commitments from more than 170 organizations to develop and support computer science (CS) programming and train teachers. Although the initiatives vary widely in scope, many are focused on addressing persistent challenges in the field, including teacher-training pathways and professional development, curriculum resources, and accessible out-of-school time programs. For example, the Girls Scouts of the USA will introduce computing programs serving 400,000 girls ages 12 to 18 annually. Meanwhile, the Alliance for California Computing Education for Students and Schools is launching a CSforCA campaign to bring CS education to 6 million students across the state. In addition, public universities such as the University of Texas at Austin, City University of New York, and Michigan State University will expand CS development for pre-service and in-service teachers. The CSforAll Consortium's Ruthe Farmer says advocates need to find ways to bring CS to children where they already are, not the other way round.
|
QMUL and BBC Launch Major Partnership to Unlock Potential of Data
Queen Mary, University of London
Rupert Marquand
October 19, 2017
Queen Mary University of London (QMUL) in the U.K. and BBC Research and Development recently announced the Data Science Research Partnership, a five-year effort to analyze the potential of data in the media. The partnership will be at the forefront of machine learning in the media industry, helping create a more personal BBC platform that can inform, educate, and entertain. The effort will collaborate with media and technology organizations from across the U.K., Europe, and internationally on various projects, focusing on understanding audiences, understanding context, curation and personalization, and future content. The goal is to create a body of research, insights, and prototypes that can make a significant impact on the BBC and its audiences. "The partnership will also address the scarcity of data scientists in the U.K., training a new generation...on real media problems, to create new audience experiences that don't even exist yet," says BBC's Samantha Chadwick.
|
AlphaGo Zero Shows Machines Can Become Superhuman Without Any Help
Technology Review
Will Knight
October 18, 2017
DeepMind's upgrade to the AlphaGo algorithm, AlphaGo Zero, beat its predecessor in a 100-game Go match, acquiring skills without human data by playing millions of games against itself. "By not using human data or human expertise, we've actually removed the constraints of human knowledge," says University College London (U.K.) professor David Silver. "It's able to create knowledge for itself from first principles." DeepMind CEO Demis Hassabis says the methods for building AlphaGo Zero can be used in real-world scenarios in which vast possibilities must be explored. Both Silver and Hassabis note many situations may lack a sufficiently large body of examples to learn from, making self-learning a requirement. AlphaGo Zero, like AlphaGo, used a deep neural network and a search algorithm to determine its next move, but it differed from the original AlphaGo program in its use of a single network to execute both functions.
*May Require Free Registration
|
Software Improves Captioning for Those With Hearing Deficits
University of Michigan News
Laurel Thomas
October 17, 2017
Researchers at the University of Michigan say they have developed Scribe, a program that makes getting real-time captions on demand possible by simultaneously engaging multiple, non-expert captionists. Scribe takes content from several less-skilled translators and intelligently forms captions in less than four seconds, according to an article in the current issue of Communications of the ACM. Deaf and hard-of-hearing students that currently require assistance must notify an office dedicated to helping them, which then hires a translator at an hourly rate. The translator often is not someone with subject-matter expertise, but with the new Scribe system, several peers or work-study students can translate the same material more accurately and for less money. On average, people can only type about 10 percent to 20 percent of what is being said, but when the notes of many captionists are combined, the full message becomes more complete, says University of Michigan professor Walter Lasecki.
|
Self-Driving Car Experts Recall Blips Along the Road to Winning DARPA Urban Challenge
Pittsburgh Post-Gazette
Courtney Linder
October 17, 2017
At an event celebrating the 10th anniversary of the winning of the U.S. Defense Advanced Research Projects Agency's (DARPA) Urban Challenge, driverless car experts reflected on the often-stumbling progress in their quest to achieve victory. Carnegie Mellon University (CMU) professor William L. Whittaker remembered that during the inaugural DARPA Challenge on desert terrain in 2004, CMU's "Sandstorm" vehicle traveled the farthest, but did not finish the course due to an obstructing rock. He noted in his keynote speech that his team's 2007 win and the $2-million prize it earned were "transformational" to the self-driving industry. Many of the Urban Challenge's visionaries have since gone on to lead competing autonomous vehicle companies such as Aurora Innovation, Delphi, and Argo AI. The experts said the competition's ultimate success laid the groundwork for the current autonomous vehicle industry, which has exploded and branched into ancillary markets for driverless cars' hardware systems.
|
New Techniques Boost Performance of Non-Volatile Memory Systems
NC State News
Matt Shipman
October 17, 2017
Researchers at North Carolina State University (NCSU) say they have developed new software and hardware designs that should limit programming errors and improve system performance in devices using non-volatile memory (NVM) technologies. The researchers note dynamic random-access memory (DRAM) has significant limitations, making it difficult to scale up to deal with next generation systems, which means future systems will likely rely on emerging NVM technologies for both operations and data storage. "Our work here is focused on addressing some of the programming and performance challenges inherent in shifting from a DRAM computing paradigm to NVM," says NCSU professor Yan Solihin. The researchers developed Proteus, a system that includes a software model and complementary hardware. "Compared to existing techniques, Proteus was able to log memory almost for free, in terms of writing to memory," Solihin says. The researchers presented their work at the recent IEEE/ACM International Symposium on Microarchitecture (MICRO 50) in Boston, MA.
|
Project Aims to Increase STEM Access for Native American Students
Campus Technology
Adam Stone
October 12, 2017
Native Americans comprise 1.2 percent of the overall U.S. population, but only account for 0.4 percent of all engineering bachelor's degrees, according to Sandia National Laboratories. The University of Montana wants to solve this problem by using a U.S. National Science Foundation grant to launch the American Indian Traditional Science Experience, a pilot project to encourage Native American participation in science, technology, engineering, and mathematics (STEM) fields. "Our focus is to work with middle school students to help create pathways into STEM that will continue through high school and then on to higher education," says University of Montana researcher Aaron Thomas. He notes the project aims to combine after-school, hands-on learning opportunities and long-term educational programming in order to generate better cultural awareness in STEM fields. The Montana pilot program will eventually expand across seven Montana reservations.
|
Exascale and the City
Argonne National Laboratory
Joan Koka
October 16, 2017
As the leader of the Multiscale Coupled Urban Systems project, Argonne National Laboratory researcher Charlie Catlett seeks to produce a computational framework for urban developers and planners to assess integrated models of city systems and processes. "We're focused on coupling models for urban atmosphere, building energy, socioeconomic activity, and transportation, and we'll later expand to energy systems models," Catlett says. "The framework will define what data will be exchanged between these models and how that data will be structured." Once the framework is complete, city planners can partner with scientists to address issues, raise their own, and optimize design proposals. Catlett and colleagues are working to merge EnergyPlus, a program for modeling buildings' energy demands, with Nek5000, a turbulence model for tracking urban heat and airflow. The coupling framework also will be designed to incorporate sensor data to help validate and improve existing models. Catlett says the project will tap exascale computing to develop these modeling solutions.
|
Using Artificial Intelligence to Improve Early Breast Cancer Detection
MIT News
Adam Conner-Simons
October 16, 2017
Researchers at the Massachusetts Institute of Technology's (MIT) Computer Science and Artificial Intelligence Laboratory (CSAIL) are working on the first project to apply artificial intelligence (AI) to improving the detection and diagnosis of breast cancer. The researchers developed an AI system that uses machine learning to predict if a high-risk lesion identified by a needle biopsy after a mammogram will upgrade to cancer at surgery. The researchers tested the system on 335 high-risk lesions, correctly diagnosing 97 percent of the breast cancers as malignant and reducing the number of benign surgeries by more than 30 percent compared to existing approaches. The machine-learning model was trained on more than 600 existing high-risk lesions, and now is able to look for patterns among many different data elements such as demographics, family history, past biopsies, and pathology reports. The model uses a method known as a "random-forest classifier," which results in fewer unnecessary surgeries.
|
IUPUI Developing, Testing Tools to Predict Crime, Other Social Harms
IUPUI Newsroom
Candace Beaty Gwaltney
October 12, 2017
Researchers at Indiana University-Purdue University Indianapolis (IUPUI) have received a grant from the U.S. National Science Foundation's (NSF) Smart and Connected Communities program to develop a system that can predict where and when crime and other social-harm events are likely to occur. The researchers will develop algorithms and a software system to collect and analyze data, enabling users to make dynamic predictions of social-harm incidents. The researchers say this type of analysis will empower stakeholders to make informed decisions about how to allocate resources to maintain and improve the quality of life in their communities. The proposed system will be based on trusted distributed systems, which is an area of focus for IUPUI professor Rajeev Raje. He says the team will use the new technology to study social systems, estimate the probability of specific harmful events, and determine effects of specific actions. The final phase of the project will be a randomized controlled field trial in Indianapolis.
|
What CMU's Snake Robot Team Learned While Searching for Mexican Earthquake Survivors
IEEE Spectrum
Evan Ackerman
October 13, 2017
Researchers at Carnegie Mellon University (CMU) used their snake robots in search-and-rescue missions in Mexico City shortly after a major earthquake struck the region in September. In an interview, CMU's Matt Travers and Howie Choset discuss the insights gained from the deployment. "What makes this snake robot unique is that it has a small cross-sectional diameter," Choset says. "It has the benefit of being small, but the potential locomotive capabilities of much larger robots, making it versatile." The snake robot is equipped with a camera on the front, but Travers says dogs and microphones used by the human crews were more effective, as sound and smell travel through concrete better than light does. "It was interesting to see how much better our system performed next to the closest system in the [Red Cross'] toolbox, which was the camera on the stick," Travers notes. Choset says the snake robot also is the best option for entering tight spaces.
|
|
