Welcome to the January 24, 2018 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."
|
|
|
China Wants to Make the Chips That Will Add AI to Any Gadget
Technology Review
Yiting Sun
January 24, 2018
China's technology industry is exploring innovative processor designs for the purpose of incorporating artificial intelligence into ordinary devices. For example, researchers at Tsinghua University in Beijing have developed Thinker, a low-power chip that is capable of dynamically tailoring its computing and memory requirements to meet the requirements of the software being run. Thinker's creators say the chip could be embedded in a broad spectrum of devices, including smartphones, watches, home robots, or equipment stationed in remote areas. The Tsinghua team is planning to launch the first product equipped with Thinker in March. Also coming soon is the production of Dadu, a dual-core chip for robots that runs neural networks and controls motion. The Dadu development team says the chip's neural core operates algorithms for vision, while the chip's motion core plans the optimal route for reaching a destination or the best motion for grabbing an object.
|
Drones Learn to Navigate Autonomously by Imitating Cars and Bicycles
University of Zurich
January 23, 2018
Researchers at the University of Zurich (UZH) and the National Center of Competence in Research Robotics in Switzerland have developed the DroNet algorithm, which enables drones to fly by themselves through the streets of a city and in indoor environments. They say the deep-learning algorithm was trained on traffic rules and examples from cyclists and car drivers. DroNet generates two outputs for each single camera input image, including a steering angle to keep the drone navigating while evading obstacles, and a collision probability so the drone can identify and promptly respond to dangerous situations. The researchers have demonstrated that their DroNet-equipped drones learned not only to navigate through city streets, but also to do so in completely different environments, where they were never trained. "With this algorithm we have taken a step forward towards integrating autonomously navigating drones into our everyday life," says UZH professor Davide Scaramuzza.
|
Clemson Attempts to Crack Code of Culturally Responsive Computer Science Teaching
The Newsstand (SC)
Michael Staton
January 23, 2018
Faculty researchers at Clemson University are using a U.S. National Science Foundation grant to help computer science educators across South Carolina develop teaching methods that better serve its diverse population. "As educators, we're selling students and our discipline short if we're educating a population of computer science students that isn't reflective of our state," says Clemson professor Megan Che. "Computer science can be a tool for any student to express problems around them as well as possible solutions to those problems." The researchers will collaborate with upstate schools to measure the effectiveness of their instructional strategies. The project applies a culturally responsive approach to teaching and learning so students can connect their daily life experiences with computer science content. "The hope is that we can become a model for any schools or districts in any state that want to make computer science rigorous, engaging, and more culturally responsive," says Clemson professor Murali Sitaraman.
|
Coders Are Shifting to Silicon Valley-Backed Programming Languages
ITPro
Adam Shepherd
January 23, 2018
Developers are planning to focus on languages created and popularized by Silicon Valley technology companies, according to HackerRank's 2018 Developer Skills Report. The report found Google's Go language topped the list of languages developers are planning to learn next, and Mozilla's Rust language and Apple's Swift also placed in the top 10. However, more than 30 percent of respondents said they planned to learn the older Python language next, while 15 percent noted they wanted to learn C# and Perl. Javascript was found to be the most practical language, and more than 45 percent of employers said they are actively looking for Java or Javascript knowledge in prospective employees, while 33 percent of employers said they are looking for Python knowledge. The purpose of the report is to help companies become more developer-focused. "Very few companies are doing tech hiring well because there's a gap in knowledge about engineers," says HackerRank CEO Vivek Ravisankar.
|
DARPA Funds Six Centers Working on Computer Design Alternatives
TechRepublic
Evan Koblentz
January 22, 2018
The U.S. Defense Advanced Research Projects Agency (DARPA) is funding six new centers to develop alternative computing architectures such as embedding logic within memory, and modular designs that make motherboards redundant. "We need to rethink the overall system stack," says University of Virginia professor Kevin Skadron. "So we need new programming frameworks that insulate the programmer from these new architectural capabilities and allow them to write code that is highly portable." Skadron also says operating systems must be reimagined with new methods for organizing and protecting memory. A DARPA-funded center at the University of Michigan (U-M) aims to simplify application development by making processor and memory components modular in the same assembly. "When many more people can make the applications, that's what makes an ecosystem bloom," says U-M professor (and ACM Distinguished Scientist) Valeria Bertacco, who wants to work on processors customized to algorithms so developers can choose the processor best suited to their task.
|
New Autonomous Car Can Handle Snow and Ice
R&D Magazine
Kenny Walter
January 18, 2018
Researchers at the VTT Technical Research Center in Finland have developed Martti, a self-driving vehicle designed to maneuver in rough and icy conditions. The team says they created filtering technology for processing environment perception data and improving the performance of LiDAR in snowy conditions. "In addition, the control strategy of driving this car has special features for controlling the vehicle when friction due to snow or ice makes driving conditions challenging," notes VTT's Matti Kutila. Martti includes cameras, antennae, sensors, and laser scanners, and was created using a Volkswagen Touareg. The autonomous car's three laser scanners sense the environment in front of the vehicle, and a global-positioning system with correction signaling and an inertia unit helps align the car on the road. The researchers tested Martii on a snow-covered road in Muonio, Finland, during which it reached a speed of 25 miles an hour, marking a major breakthrough for autonomous cars in rough conditions.
|
Engineers Design Artificial Synapse for 'Brain-on-a-Chip' Hardware
MIT News
Jennifer Chu
January 22, 2018
Engineers at the Massachusetts Institute of Technology (MIT) have designed an artificial synapse that enables them to finely control the strength of an electric current flowing across it. Instead of employing amorphous materials as an artificial synapse, the researchers selected single-crystalline silicon with a one-dimensional dislocation through which ions could predictably flow. A silicon germanium lattice on top of a silicon chip formed a funnel-like dislocation, supporting single-path ionic flow. The team applied voltage to each synapse, and all synapses exhibited an approximately identical current, with about a 4-percent variation between synapses. Simulation testing suggests the chip and its synapses could be capable of identifying handwriting samples with 95-percent accuracy. The researchers see this as a key advance toward portable, low-power neuromorphic chips for pattern recognition and other learning tasks. "Ultimately, we want a chip as big as a fingernail to replace one big supercomputer," says MIT professor Jeehwan Kim.
|
How Do We Protect Today's Secrets From Tomorrow's Quantum Computers?
FedScoop
Carten Cordell
January 22, 2018
To ensure the future protection of networks from quantum computers, the U.S. National Institute of Standards and Technology (NIST) has approached private stakeholders to help develop new cryptographic standards. The short-term goal of the effort is to continue the secure data encryption of billions of computers, including those deployed across critical infrastructure systems. NIST's Post-Quantum Cryptography contest solicited algorithms from the larger cryptography community, and they will be assessed in the coming years prior to the issuance of draft standards. "There's really no guarantee that a quantum computer can't break these problems necessarily," says NIST's Dustin Moody. "It's just that nobody has been able to find a way to do it so far, which might not sound great, but that's kind of actually what we rely on today. It's really the same paradigm, it's just that we are putting quantum computers into the mix as well."
|
Students Use Sandbox to Demonstrate Gravity
Daily Iowan
Brooklyn Draisey
January 18, 2018
Researchers at the University of Iowa (UI) have developed Gravbox, an augmented reality (AR) sandbox to help explain how gravity affects objects when they travel through a certain environment. They say Gravbox is created with the sand, and then a computer program projects a particle moving through that environment to show gravity's effects. The UI team based Gravbox on the University of California, Davis' AR sandbox that simulates water flow in an environment. "Essentially, we want all of this to be open source, so anyone can go to our website, download it, look at the hardware, the specifications they need, then they can go and try to build their own," says UI's Mason Reed. He notes Gravbox makes demonstrations of gravity less tedious and more interesting to see. The researchers want to use these features at local schools and children's museums to help young people understand gravity in a more hands-on way.
|
Brainy Little UT Robot May Hold Key to Smart, Unmanned Drones and Vehicles
Associated Press
Brittany Crocker
January 23, 2018
A team at the University of Tennessee and Oak Ridge National Laboratory (ORNL) has created Neon, a neuromorphic robot vehicle that can navigate while evading objects. They say once booted up with a computer, the prototype Neon requires no outside programming to move. The researchers note Neon is trained to beware of collisions, learning as it goes, and becoming smarter and more accurate in its performance every time it approaches something in its path. "We ran it through a simulation where the next step is to train the brain not only to avoid objects but to target certain objects, and it is incentivized to go as many places as it can, without hitting anything," says ORNL's Katie Schuman. Neon's generational training process involved many versions of its brain perishing, which Schuman notes taught it survival strategies. The team envisions Neon's technology applied to aerial drones and other devices that are environmentally adaptable.
|
NIST Takes a Deep Look at Memristors
NIST News
Ben P. Stein
January 18, 2018
Researchers at the U.S. National Institute of Standards and Technology (NIST) and the University of California, Santa Barbara have detailed the creation of memristors, semiconductor elements that function like the short-term memory of nerve cells. They say a memristor's resistance relies on the amount of current that recently flowed through it, and it retains that memory even when electrical power is switched off. However, until now, scientists have lacked a detailed understanding of how these devices work and have yet to develop a standard toolset to study them. The NIST team identified such a toolset and used it to more thoroughly examine how memristors operate. The researchers studied the electrical function of memristors by aiming a tightly focused beam of electrons at different locations on a titanium dioxide memristor and the beam induced four distinct currents to flow within the device, which are associated with the multiple interfaces between materials and memristors.
|
Using Data Mining to Make Sense of Climate Change
Georgia Tech News Center
Jason Maderer
January 17, 2018
Researchers at the Georgia Institute of Technology (Georgia Tech) have developed an open source and self-contained data-mining method for analyzing climate change model datasets. The tool's development involved collaboration between both climate scientists and computational researchers, and its developers say the technique extracts commonalities of datasets while demanding less user expertise, creating more trustworthy data and stronger and more transparent results. Georgia Tech professor Athanasios Nenes notes the methodology follows a more robust and holistic strategy than conventional approaches, eliminating the bottlenecks that other model evaluation and analysis algorithms suffer from. "The methodology reduces the complexity of millions of data points to the bare essentials--sometimes as few as 10 regions that interact with each other," Nenes says. "We need to have tools that reduce the complexity of model output to understand them better and evaluate if they are providing the correct results for the right reasons."
|
Trust Me, I'm a Robot
PC Magazine
Sophia Stuart
January 22, 2018
In an interview, Ayanna Howard discusses her work directing the Georgia Institute of Technology's Human-Automation Systems Lab (HumAnS), a multidisciplinary effort to bring humanized intelligence and human cognitive capability to autonomous systems. Howard says her background in both classical engineering and computer science inspired her to study problems from a systems thinking perspective, with an emphasis on assistive robots for pediatric health therapy. "We found that children want to cooperate with the robot," she notes. "Also the robot doesn't get tired or lose patience with the child and that helps with maintaining longer-term interactions." Howard recognizes the importance of trust in human-in-the-loop robotics, and says educating robots to mimic human behavior fosters interaction and trust. However, she warns excessive trust in robots can be dangerous, and HumAnS is working to mitigate this hazard to discourage overreliance. "The key is to maximize the rewards while minimizing any potential risk," Howard says.
|
|
