News 2018

August 2018

Moore Steps Down as Dean of CMU's School of Computer Science

Byron Spice

Andrew W. Moore has announced he will step down as dean of Carnegie Mellon University's School of Computer Science and as a professor of computer science and robotics, effective at the end of the calendar year. Moore, who said he is stepping down to accept a new professional opportunity, joined the Carnegie Mellon faculty in 1993. He was recruited by Google in 2006 to establish its Pittsburgh office and serve as a vice president. He returned to CMU as dean of the world's top-ranked computer science school in August 2014. "When I took this post four years ago, I talked of how I was inspired by the world-changing ideas and technologies spawned by the School of Computer Science; today I can say I'm simply awed by what's happening here — everything ranging from our research in artificial intelligence to the increasing diversity of our student body. I am bursting with excitement to see where SCS is going next." "I want to express my deepest thanks to the amazing students, faculty and staff of SCS," Moore said. "This school is extremely strong and remains a thought-leader in computer science and robotics. With the technological changes now facing mankind, the world has never needed a strong SCS more than it does today." Carnegie Mellon President Farnam Jahanian said Moore's impact on CMU and Pittsburgh cannot be overstated. "Andrew Moore has been passionate about the impact of technology on society and a leader in the way technology enhances people's lives. At this pivotal time for both the School of Computer Science and Carnegie Mellon, we will make sure the momentum he built will continue," Jahanian said. Moore, known for his expertise in artificial intelligence and robotics, has pushed to increase the size and diversity of SCS's student body, in response to the overwhelming demand for the school's graduates. Women have reached parity with men in the last three incoming classes of undergraduates, and the size of the incoming class has increased from 139 in 2014 to a record 211 this year. New Carnegie Mellon undergraduate degrees in computational biology and artificial intelligence have been established. Outreach efforts to K-12 students begun under Moore's guidance are designed to increase the number of underrepresented minorities studying computer science. "Through his exceptional gift of explaining technology and presenting a vision of the future, Andrew has helped to establish Pittsburgh as a prominent hub of the new economy, and also greatly enhanced our institutional reputation," Jahanian said. Moore has sought to leverage the school's pioneering work in artificial intelligence, establishing a CMU AI initiative that now includes more than 200 faculty from across campus working together on some of today's most pressing issues. He has been an outspoken proponent of AI, testifying before Congress and consulting regularly with government and industrial leaders. Though AI sometimes is depicted as a possible menace, Moore has contended that AI holds great potential for good, and has urged researchers to consider the societal impact of their work in this area. Jahanian said in the next few weeks an interim dean will be appointed and a national search for Moore's successor will begin. Moore received a doctorate from the University of Cambridge in 1991. In 2005, he was elected a fellow of the American Association for Artificial Intelligence.

Carnegie Mellon Selected for Two Amazon Alexa Fellowships

Byron Spice

Carnegie Mellon University is one of 18 universities selected from around the world for an expanded Amazon Alexa Fellowship program, and one of just two universities in that group to receive both an Alexa Graduate Fellowship and a newly created Alexa Innovation Fellowship. CMU was one of four universities chosen last year by Amazon when it launched the fellowship program, which supports universities and researchers specializing in conversational AI. It is named for Alexa, the voice service that powers Amazon devices such as Echo. The fellowships during that initial year were awarded to graduate students, including Ran Zhao, a Ph.D. student in CMU's Language Technologies Institute. Ten universities will have Alexa graduate fellows in the expanded program. New this year is the Innovation Fellowship. Rather than presenting fellowships to graduate students, this award goes to faculty members at 10 universities who will serve as expert resources to help students start companies that integrate voice with their technology. "While graduate students are uniquely positioned to accelerate conversational AI research and education, we've observed students with diverse backgrounds inventing new products and services that can benefit from adding voice interfaces," Amazon said in announcing the expanded fellowship program. CMU's Alexa Innovation Fellow is Kit Needham, director of Project Olympus, a startup incubator program that is part of CMU's Swartz Center for Entrepreneurship. She is an advisory board member of Chatham University's Center for Women's Entrepreneurship and adjunct faculty in their MBA program. She was chief operating officer of Mellon Lab, which was responsible for creating revenue-producing businesses for Mellon Bank. She also provides consulting services to promote economic growth to chambers of commerce, individual entrepreneurs, and nonprofit organizations. Needham will travel to Amazon headquarters in Seattle for two days of training. She will receive funding and Alexa devices and hardware that can be used to help startups add voice interfaces to their technologies. Zhao will continue as CMU's Alexa graduate fellow in the second year of the program. The fellowship covers tuition, provides a stipend and pairs Zhao with an Alexa scientist. The fellowship also provides Alexa devices and developer kits for use in an LTI project course, Dialogue Systems. Alex Rudnicky, research professor emeritus, and Alan Black, professor of language technologies, teach the course with assistance from Zhao. Zhao's Ph.D. thesis focuses on extending task-oriented dialogue systems with social capabilities to better reflect the characteristics of human-human interaction. He is an author on 11 papers and received the Best Student Paper award at the 16th International Conference on Intelligent Virtual Agents. Learn more about the fellowship on the program's website.

Actuation Gives New Dimensions to an Old Material

Byron Spice

One of the oldest, most versatile and inexpensive materials — paper — seemingly springs to life, bending, folding or flattening itself, by means of a low-cost actuation technology developed at Carnegie Mellon University's Human-Computer Interaction Institute.A thin layer of conducting thermoplastic, applied to common paper with an inexpensive 3D printer or even painted by hand, serves as a low-cost, reversible actuator. When an electrical current is applied, the thermoplastic heats and expands, causing the paper to bend or fold. When the current is removed, the paper returns to a predetermined shape."We are reinventing this really old material," said Lining Yao, assistant professor in the HCII and director of the Morphing Matter Lab, who developed the method with her team. "Actuation truly turns paper into another medium, one that has both artistic and practical uses."Post-doctoral researcher Guanyun Wang, former research intern Tingyu Cheng and other members of Yao's Morphing Matter Lab have designed basic types of actuators, including some based on origami and kirigami forms. These enable the creation of structures that can turn themselves into balls or cylinders. Or they can be used to construct more elaborate objects, such as a lamp shade that changes its shape and the amount of light it emits, or an artificial mimosa plant with leaf petals that sequentially open when one is touched.In June, more than 50 students in a workshop at Zhejiang University in Hangzhou, China, used the paper actuation technology to create elaborate pop-up books, including interpretations of famous artworks, such as Van Gogh's "Starry Night" and "Sunflowers."The printed paper actuator will be exhibited Sept. 6–10 at the Ars Electronica Festival in Linz, Austria; Sept. 13–30 at Bozar Centre for the Fine Arts in Brussels; and from October through March at Hyundai Motorstudio in Beijing. Yao's group presented the technology in April at CHI 2018, the Conference on Human Factors in Computing Systems, in Montreal."Most robots — even those made of paper — require an external motor," said Wang, a CMU Manufacturing Futures Initiative fellow. "Ours do not, which creates new opportunities, not just for robotics, but for interactive art, entertainment and home applications."Creating a paper actuator is a relatively simple process, Cheng said. It employs the least expensive type of 3D printer, a so-called FDM printer that lays down a continuous filament of melted thermoplastic. The researchers use an off-the-shelf printing filament — graphene polyactide composite — that conducts electricity.The thermoplastic actuator is printed on plain copy paper in a thin layer, just half a millimeter thick. The actuator is then heated in an oven or with a heat gun, and the paper is bent or folded into a desired shape and allowed to cool. This will be the default shape of the paper. Electrical leads can then be attached to the actuator. Applying electrical current heats the actuator, causing the thermoplastic to expand and straighten the paper. When the current is removed, the paper automatically returns to its default shape.Yao said the researchers are refining this method, changing the printing speed or the width of the line of thermoplastic to achieve different folding or bending effects. They have also developed methods for printing touch sensors, finger-sliding sensors and bending-angle detectors that can control the paper actuators.More work remains to be done. Actuation is slow, which Yao and her team hope to address with some material engineering — using papers that are more heat conductive and developing printing filaments that are customized for use in actuators. The same actuation used for paper might also be used for plastics and fabrics.In addition to Yao, Wang and Cheng, authors of the CHI research paper include Youngwook Do and Byoungkwon An, HCII research affiliates; Jianzhe Gu, a Ph.D. student in the HCII; Humphrey Yang, a master's student in the CMU School of Architecture; and Ye Tao, a visiting scholar from Zhejiang University. More information is available on the project's web page.

Carnegie Mellon Joins Meltwater To Advance Data Science

Byron Spice

Students and faculty at Carnegie Mellon University's School of Computer Science are collaborating with the digital media intelligence firm Meltwater to advance the state of the art in artificial intelligence education and research using the company's AI platform. Meltwater, which has the world's most diverse collection of open and licensed data, has opened its underlying AI platform, Fairhair.ai, to Carnegie Mellon and other select universities. The Fairhair.ai platform allows students and faculty to create, connect and organize web-scale information to generate real-time analytics that support decision-making from online data. CMU will use Fairhair.ai in graduate AI courses and as a resource for the university's data science and AI research community. "Sharing access to real-world data helps students, researchers and data scientists solve real-world problems more rapidly," said Eric Nyberg, director of CMU's Master of Computational Data Science program and a professor in the Language Technologies Institute. "In addition to realistic real-time data sources, the platform also includes AI modeling and integrated cloud computing to greatly simplify the process of building and optimizing new web-scale analytics." This past January, Nyberg joined Majd Sakr, a teaching professor in CMU's Computer Science Department, to launch the Accelerated Cloud for Artificial Intelligence Project with support from Meltwater. A team of MCDS capstone students — Shihui Li, Ganesh Palanikumar and Sida Wang — began developing a set of realistic benchmark challenges for natural language processing tasks. The ACAI project's initial focus is named entity recognition, an element of information extraction that classifies objects with proper names into categories, such as people, organizations and locations. In the coming semester, the team plans to create an open, web-scale named entity recognition challenge and benchmark possible solutions using Fairhair.ai resources. An international authority on the design of computer systems for answering questions, Nyberg worked with his students and members of IBM Research from 2007 to 2011 to develop Watson for the Jeopardy! Challenge. As a member of the Fairhair.ai Scientific Advisory Board, Nyberg helped to shape the development of the platform. He said he is focused on using Fairhair.ai to develop new methods for rapid, cost-effective development of specialized question-answering systems for specific information domains. "Eric Nyberg has played an instrumental role during the development of the platform that helped us build the right interface and toolkits for data scientists," said Aditya Jami, CTO of Meltwater. "CMU is well known for pushing the boundaries in the field of AI and this collaboration will undoubtedly foster a new wave of open innovation." "The ACAI project will crystallize important advancements in the engineering of cost-effective AI systems," Nyberg said. "The biggest challenge for current students is how to explore the large space of data, features and models available for developing a particular analytic in order to find an optimal or acceptable solution before they run out of time or computing resources," he continued. "The ACAI framework will allow students to explore this large solution space by providing a systematic approach that teaches cost-effective use of cloud resources to build AI systems." "ACAI's research outcomes are already strengthening Fairhair.ai's machine learning and text analytics capabilities across the board," said Giorgio Orsi, principal scientist and director of natural language processing at Meltwater. "They are enabling rapid model-building and customization of our text analytics to the needs of our customers." Nyberg said he hopes to use Fairhair.ai as part of his regular graduate course, Design and Engineering of Intelligent Information Systems. "Building a state-of-the-art AI system requires us to store, preprocess and annotate text collections with a variety of feature extractors, while simultaneously exploring the space of possible models that can be built from the data as they emerge," Nyberg said. "By providing a principled framework for storage, metadata and model training, along with a massive collection of open web data and metadata, Meltwater will make it possible for students to build advanced analytics in a classroom setting."

How Can AI Improve Humanitarian Assistance and Disaster Relief?

Dave Smalley

The  topic of artificial intelligence for humanitarian assistance and disaster relief (HA/DR) was in the spotlight last week, as leading minds from academia, industry and the federal government met to discuss how modern technology can help victims of disasters around the globe. The Artificial Intelligence and Autonomy for Humanitarian Assistance and Disaster Relief Workshop, co-hosted by the Office of Naval Research (ONR) and Carnegie Mellon University, took place at the university's Pittsburgh campus Aug. 2-3. The parties also signed an Education Partnership Agreement to encourage and strengthen studies in the scientific disciplines, particularly as they relate to humanitarian assistance and disaster relief. ''The problem of catastrophes affecting humanity will unfortunately always be among us,'' said Chief of Naval Research Rear Adm. David J. Hahn. ''The great minds in this room are here to figure out how we can best leverage artificial intelligence and autonomy to better deliver resources and people to those in urgent need.'' As Hahn addressed the group, pictures of naval relief efforts over many years scrolled behind him, including Sailors and Marines assisting victims in Haiti after an earthquake, Japan after an earthquake and tsunami, New Orleans and New York after hurricanes, and more. Hahn noted the Navy and Marine Corps are uniquely suited to support rescue and relief work done by different federal agencies, when called upon. Naval assets include ships, planes and helicopters; being positioned globally; and being highly mobile. Andrew Moore, dean of CMU's School of Computer Science, thanked Hahn for helping to inspire the meeting, which included renowned AI-focused academics; senior representatives from the Federal Emergency Management Agency (FEMA); leaders from the new Joint AI Center; relief directors at the U.S. State Department; representatives from NASA, the U.S. Army and Air Force; and many more. Some in attendance had been on the ground providing assistance on multiple occasions, from natural disasters to war zones. ''We've brought in some of the best roboticists and technologists in the world, with the single goal of keeping people safe during disasters,'' said Moore. ''Technology to save lives is inspiring for everyone in the AI field.'' AI and autonomy are wide-ranging terms that include multiple fields of study and are increasingly relevant to disaster relief work. Some of those focus areas include machine learning, human-computer interaction, big data analytics, computational modeling and robotics. Several speakers at the event discussed real-world examples of how AI could help victims of disasters, including better using unmanned aerial vehicles to find survivors; using robots to communicate with trapped victims, for example, under the rubble of a building; generating the most accurate data on emerging damaged areas, using crowd-sourced social media reports; and computer models and simulations to predict the best ways and tools to help in different disaster situations. Both Hahn and Moore encouraged workshop attendees to not let the meeting's value fade. A series of working groups addressed multiple topics and were requested to come up with short-, mid- and long-term ideas on how to utilize AI in real-world HA/DR situations. Follow-up meetings are currently being slated to track progress.

Robotics Institute Featured on “In Search Of”

Byron Spice

The Friday Aug. 10 episode of the HISTORY series ''In Search Of,'' hosted by actor and Carnegie Mellon University alum Zachary Quinto, will feature a segment that was shot last year at the CMU Robotics Institute. In the segment, Quinto talked with Nathan Michael, assistant research professor of robotics, and Ellen Cappo, Ph.D. student in robotics, about their research involving swarms of small drones that can interact with people. Quinto entered the netted drone enclosure in the Planetary Robotics Lab, where several drones circled benignly around him as he plucked them individually out of the air. The segment is part of an episode on artificial intelligence, which also included a visit by Quinto to the Uber Advanced Technologies Center in Pittsburgh for a look at self-driving cars. The new series, inspired by the original ''In Search Of'' hosted by Rod Serling and later by Leonard Nimoy from 1977-82, delves into all manner of mysterious phenomena. Quinto serves as executive producer of the series, which airs at 10 p.m. ET Fridays on HISTORY. The network now reaches more than 96 million homes.

Robotics Institute Featured on “In Search Of”

Byron Spice

The Friday Aug. 10 episode of the HISTORY series ''In Search Of,'' hosted by actor and Carnegie Mellon University alum Zachary Quinto, will feature a segment that was shot last year at the CMU Robotics Institute. In the segment, Quinto talked with Nathan Michael, assistant research professor of robotics, and Ellen Cappo, Ph.D. student in robotics, about their research involving swarms of small drones that can interact with people. Quinto entered the netted drone enclosure in the Planetary Robotics Lab, where several drones circled benignly around him as he plucked them individually out of the air. The segment is part of an episode on artificial intelligence, which also included a visit by Quinto to the Uber Advanced Technologies Center in Pittsburgh for a look at self-driving cars. The new series, inspired by the original ''In Search Of'' hosted by Rod Serling and later by Leonard Nimoy from 1977-82, delves into all manner of mysterious phenomena. Quinto serves as executive producer of the series, which airs at 10 p.m. ET Fridays on HISTORY. The network now reaches more than 96 million homes.

Robotics Institute Featured on “In Search Of”

Byron Spice

The Friday Aug. 10 episode of the HISTORY series ''In Search Of,'' hosted by actor and Carnegie Mellon University alum Zachary Quinto, will feature a segment that was shot last year at the CMU Robotics Institute. In the segment, Quinto talked with Nathan Michael, assistant research professor of robotics, and Ellen Cappo, Ph.D. student in robotics, about their research involving swarms of small drones that can interact with people. Quinto entered the netted drone enclosure in the Planetary Robotics Lab, where several drones circled benignly around him as he plucked them individually out of the air. The segment is part of an episode on artificial intelligence, which also included a visit by Quinto to the Uber Advanced Technologies Center in Pittsburgh for a look at self-driving cars. The new series, inspired by the original ''In Search Of'' hosted by Rod Serling and later by Leonard Nimoy from 1977-82, delves into all manner of mysterious phenomena. Quinto serves as executive producer of the series, which airs at 10 p.m. ET Fridays on HISTORY. The network now reaches more than 96 million homes.

Robotics Institute Delivers Pipe-Crawling Robot To DOE

Byron Spice

A team from Carnegie Mellon University’s Robotics Institute on June 25 delivered a robot called RadPiper to the U.S. Department of Energy’s former uranium enrichment plant in Piketon, Ohio, where it will be used to identify uranium deposits on pipe walls. RadPiper is an autonomous, tracked robot that crawls inside miles of 30-inch and 42-inch-diameter pipe at the Portsmouth Gaseous Diffusion Plant, which is being decommissioned and prepared for demolition. The robot is equipped with radiation detectors and can identify sections of pipe that have hazardous uranium deposits. This technology, developed by a team led by William “Red” Whittaker, professor of robotics, could potentially save millions of dollars in decommissioning costs by identifying those sections of pipe that need special handling and disposal. The rest of the pipe could remain in place and be demolished safely and less expensively with the rest of the process buildings – three gigantic structures that span the size of 158 football fields. The CMU team has worked closely with DOE and Fluor-BWXT Portsmouth, the decommissioning contractor, to develop the robot. Fluor-BWXT staff members are now testing the robot as they work toward its deployment. The robot also could be used in the decommissioning of a similar plant in Paducah, Kentucky. The 270-pound RadPiper robot runs through straight sections of pipe, automatically returning to its starting point when it reaches a bend or other obstruction. Sensor data is recorded on a USB drive; nuclear analysts download the information from the drive, freeing them from time-consuming calculations and making reports available the same day. More information is available on the project web site.

From high school hacker to DefCon champ

Daniel Tkacik

While many kids dream of a shopping spree at their favorite toy store or the local candy shop, Zach Wade's preference was a bit uncommon for a youngster: RadioShack. ''I had a RadioShack kit that had bajillions of components, and I kept hand-drawn books on how to connect circuits and how to build stuff,'' says Wade, now a fourth-year undergraduate in the School of Computer Science. ''That kit was my absolute favorite. I spent hours and hours playing with that.'' That RadioShack kit may have been the beginning of a path that would eventually lead Wade to the ranks of the Plaid Parliament of Pwning (PPP), Carnegie Mellon's internationally acclaimed competitive hacking team. The team is heading to Las Vegas this week to compete for its fifth "World Series of Hacking" championship at the DefCon security conference. Wade helped PPP win its fourth DefCon title last year, giving PPP more DefCon titles than any other team in the competition's 21-year history. Wade describes the way he first got into security as ''pretty much an accident.'' In high school, a friend of his dragged him to a meeting for the school's Cybersecurity Club. Despite initially feeling indifferent about the subject material, Wade recognized that the club was losing most of its leadership to graduating seniors and decided to step up and take over the club. ''That's how I got into security – because in order to lead the club, I had to first learn what I was doing,'' he says. ''Through teaching others, I gained a much deeper understanding of the subject material, and I found I really liked it.'' Wade says he's always enjoyed puzzle games of all kinds. Hacking, he learned, isn't so different. ''I think what made me enjoy security so much is it felt like playing those games,'' he says. ''The way it makes you think, the joy you get after solving something, I really enjoy that.'' In 2014 as a high school senior, Wade and some friends competed in picoCTF, the Carnegie Mellon hacking competition for middle and high school students. His team finished third place overall, beating thousands of other teams around the country. ''PicoCTF was really an important part of my high school career,'' Wade says. ''It not only helped me learn the security skills I needed to run the club, but it also introduced me to security as a profession – something that people spend a lot of time on.'' Wade says hacking competitions like these – especially the world championships at DefCon – are much more than just a game. ''It's practice for the kinds of attacks and the kinds of defenses we need in the real world,'' he says. ''But it's also a coming together of the security community to make the world safer. It's passing ideas back and forth, in the form of a game, yes, but it's so that we can, as a world, be more secure.''

From high school hacker to DefCon champ

Daniel Tkacik

While many kids dream of a shopping spree at their favorite toy store or the local candy shop, Zach Wade's preference was a bit uncommon for a youngster: RadioShack. ''I had a RadioShack kit that had bajillions of components, and I kept hand-drawn books on how to connect circuits and how to build stuff,'' says Wade, now a fourth-year undergraduate in the School of Computer Science. ''That kit was my absolute favorite. I spent hours and hours playing with that.'' That RadioShack kit may have been the beginning of a path that would eventually lead Wade to the ranks of the Plaid Parliament of Pwning (PPP), Carnegie Mellon's internationally acclaimed competitive hacking team. The team is heading to Las Vegas this week to compete for its fifth "World Series of Hacking" championship at the DefCon security conference. Wade helped PPP win its fourth DefCon title last year, giving PPP more DefCon titles than any other team in the competition's 21-year history. Wade describes the way he first got into security as ''pretty much an accident.'' In high school, a friend of his dragged him to a meeting for the school's Cybersecurity Club. Despite initially feeling indifferent about the subject material, Wade recognized that the club was losing most of its leadership to graduating seniors and decided to step up and take over the club. ''That's how I got into security – because in order to lead the club, I had to first learn what I was doing,'' he says. ''Through teaching others, I gained a much deeper understanding of the subject material, and I found I really liked it.'' Wade says he's always enjoyed puzzle games of all kinds. Hacking, he learned, isn't so different. ''I think what made me enjoy security so much is it felt like playing those games,'' he says. ''The way it makes you think, the joy you get after solving something, I really enjoy that.'' In 2014 as a high school senior, Wade and some friends competed in picoCTF, the Carnegie Mellon hacking competition for middle and high school students. His team finished third place overall, beating thousands of other teams around the country. ''PicoCTF was really an important part of my high school career,'' Wade says. ''It not only helped me learn the security skills I needed to run the club, but it also introduced me to security as a profession – something that people spend a lot of time on.'' Wade says hacking competitions like these – especially the world championships at DefCon – are much more than just a game. ''It's practice for the kinds of attacks and the kinds of defenses we need in the real world,'' he says. ''But it's also a coming together of the security community to make the world safer. It's passing ideas back and forth, in the form of a game, yes, but it's so that we can, as a world, be more secure.''

From high school hacker to DefCon champ

Daniel Tkacik

While many kids dream of a shopping spree at their favorite toy store or the local candy shop, Zach Wade's preference was a bit uncommon for a youngster: RadioShack. ''I had a RadioShack kit that had bajillions of components, and I kept hand-drawn books on how to connect circuits and how to build stuff,'' says Wade, now a fourth-year undergraduate in the School of Computer Science. ''That kit was my absolute favorite. I spent hours and hours playing with that.'' That RadioShack kit may have been the beginning of a path that would eventually lead Wade to the ranks of the Plaid Parliament of Pwning (PPP), Carnegie Mellon's internationally acclaimed competitive hacking team. The team is heading to Las Vegas this week to compete for its fifth "World Series of Hacking" championship at the DefCon security conference. Wade helped PPP win its fourth DefCon title last year, giving PPP more DefCon titles than any other team in the competition's 21-year history. Wade describes the way he first got into security as ''pretty much an accident.'' In high school, a friend of his dragged him to a meeting for the school's Cybersecurity Club. Despite initially feeling indifferent about the subject material, Wade recognized that the club was losing most of its leadership to graduating seniors and decided to step up and take over the club. ''That's how I got into security – because in order to lead the club, I had to first learn what I was doing,'' he says. ''Through teaching others, I gained a much deeper understanding of the subject material, and I found I really liked it.'' Wade says he's always enjoyed puzzle games of all kinds. Hacking, he learned, isn't so different. ''I think what made me enjoy security so much is it felt like playing those games,'' he says. ''The way it makes you think, the joy you get after solving something, I really enjoy that.'' In 2014 as a high school senior, Wade and some friends competed in picoCTF, the Carnegie Mellon hacking competition for middle and high school students. His team finished third place overall, beating thousands of other teams around the country. ''PicoCTF was really an important part of my high school career,'' Wade says. ''It not only helped me learn the security skills I needed to run the club, but it also introduced me to security as a profession – something that people spend a lot of time on.'' Wade says hacking competitions like these – especially the world championships at DefCon – are much more than just a game. ''It's practice for the kinds of attacks and the kinds of defenses we need in the real world,'' he says. ''But it's also a coming together of the security community to make the world safer. It's passing ideas back and forth, in the form of a game, yes, but it's so that we can, as a world, be more secure.''

From high school hacker to DefCon champ

Daniel Tkacik

While many kids dream of a shopping spree at their favorite toy store or the local candy shop, Zach Wade's preference was a bit uncommon for a youngster: RadioShack. ''I had a RadioShack kit that had bajillions of components, and I kept hand-drawn books on how to connect circuits and how to build stuff,'' says Wade, now a fourth-year undergraduate in the School of Computer Science. ''That kit was my absolute favorite. I spent hours and hours playing with that.'' That RadioShack kit may have been the beginning of a path that would eventually lead Wade to the ranks of the Plaid Parliament of Pwning (PPP), Carnegie Mellon's internationally acclaimed competitive hacking team. The team is heading to Las Vegas this week to compete for its fifth "World Series of Hacking" championship at the DefCon security conference. Wade helped PPP win its fourth DefCon title last year, giving PPP more DefCon titles than any other team in the competition's 21-year history. Wade describes the way he first got into security as ''pretty much an accident.'' In high school, a friend of his dragged him to a meeting for the school's Cybersecurity Club. Despite initially feeling indifferent about the subject material, Wade recognized that the club was losing most of its leadership to graduating seniors and decided to step up and take over the club. ''That's how I got into security – because in order to lead the club, I had to first learn what I was doing,'' he says. ''Through teaching others, I gained a much deeper understanding of the subject material, and I found I really liked it.'' Wade says he's always enjoyed puzzle games of all kinds. Hacking, he learned, isn't so different. ''I think what made me enjoy security so much is it felt like playing those games,'' he says. ''The way it makes you think, the joy you get after solving something, I really enjoy that.'' In 2014 as a high school senior, Wade and some friends competed in picoCTF, the Carnegie Mellon hacking competition for middle and high school students. His team finished third place overall, beating thousands of other teams around the country. ''PicoCTF was really an important part of my high school career,'' Wade says. ''It not only helped me learn the security skills I needed to run the club, but it also introduced me to security as a profession – something that people spend a lot of time on.'' Wade says hacking competitions like these – especially the world championships at DefCon – are much more than just a game. ''It's practice for the kinds of attacks and the kinds of defenses we need in the real world,'' he says. ''But it's also a coming together of the security community to make the world safer. It's passing ideas back and forth, in the form of a game, yes, but it's so that we can, as a world, be more secure.''

How a Computer Learns To Dribble: Practice, Practice, Practice

Byron Spice

Basketball players need lots of practice before they master the dribble, and it turns out that's true for computer-animated players as well. By using deep reinforcement learning, players in basketball video games can glean insights from motion-capture data to sharpen their dribbling skills.Researchers at Carnegie Mellon University and DeepMotion Inc., a California company that develops smart avatars, have for the first time developed a physics-based, real-time method for controlling animated characters that can learn dribbling skills from experience. In this case, the system learns from motion capture of the movements performed by people dribbling basketballs.This trial-and-error learning process is time consuming, requiring millions of trials, but the results are arm movements that are closely coordinated with physically plausible ball movement. Players learn to dribble between their legs, dribble behind their backs and do crossover moves, as well as how to transition from one skill to another."Once the skills are learned, new motions can be simulated much faster than real-time," said Jessica Hodgins, Carnegie Mellon professor of computer science and robotics.Hodgins and Libin Liu, chief scientist at DeepMotion, will present the method at SIGGRAPH 2018, the Conference on Computer Graphics and Interactive Techniques, Aug. 12–18, in Vancouver."This research opens the door to simulating sports with skilled virtual avatars," said Liu, the report's first author. "The technology can be applied beyond sport simulation to create more interactive characters for gaming, animation, motion analysis, and in the future, robotics."Motion capture data already add realism to state-of-the-art video games. But these games also include disconcerting artifacts, Liu noted, such as balls that follow impossible trajectories or that seem to stick to a player's hand.A physics-based method has the potential to create more realistic games, but getting the subtle details right is difficult. That's especially so for dribbling a basketball, because player contact with the ball is brief and finger position is critical. Some details, such as the way a ball may continue spinning briefly when it makes light contact with the player's hands, are tough to reproduce. And once the ball is released, the player has to anticipate when and where the ball will return.Liu and Hodgins opted to use deep reinforcement learning to enable the model to pick up these important details. Artificial intelligence programs have used this form of deep learning to figure out a variety of video games, and the AlphaGo program famously employed it to master the board game Go.The motion capture data used as input was of people doing things such as rotating the ball around the waist, dribbling while running, and dribbling in place both with the right hand and while switching hands. This capture data did not include the ball movement, which Liu explained is difficult to record accurately. Instead, they used trajectory optimization to calculate the ball's most likely paths for a given hand motion.The program learned the skills in two stages — first it mastered locomotion and then it learned how to control the arms and hands and, through them, the motion of the ball. This decoupled approach is sufficient for actions such as dribbling or perhaps juggling, where the interaction between the character and the object doesn't have an effect on the character's balance. Further work is required to address sports, such as soccer, where balance is tightly coupled with game maneuvers, Liu said.Learn more about this research on the Carnegie Mellon Graphics website.

Reading, Writing, and Robotics

Pauline Dolle

Carnegie Mellon University teamed up with the Pittsburgh-based educational company WholeRen Education LLC to host the inaugural World Artificial Intelligence Competition for Youth, or WAICY 2018, at the end of July. The week of workshops and the final competition demonstrated that students of all ages can understand artificial intelligence if given the proper equipment and training. ''At first, Max could identify what he wanted to do but couldn’t connect it to the robot,'' said Sha Onyang, the mother of the competition’s youngest participant, 4-year-old Max Tongxi Zhao. ''Then he saw that Cozmo could do a lot of things. The professors guided him to collaborate with others and design his project.'' One hundred and thirty students between the ages of 4 and 18 some from as far away as China, Sweden and England, competed in person on 30 teams. An additional 20 teams competed remotely from around the world. The students came from a wide variety of backgrounds, and for many it was their first experience with AI programming. ''We don’t want kids to be left out of technology,'' said Brian White, president of WholeRen Group. Using WholeRen’s ''AI-In-A-Box,'' a toolkit developed with input from CMU alumni and faculty, each team designed and executed projects that answered a simple question: How can AI change our lives? Two 11-year-old boys envisioned a robot that could take orders at a restaurant and deliver the correct plates. One team had a 6-year-old member with a speech impediment and was inspired to program Cozmo to speak for their friend in response to stimuli. Several teams saw the potential for robots to mine resources in areas too dangerous for humans, like underwater caves or to rescue people from disaster zones. A pair of girls, Tassneem Khattab and Weeam Guetari, ninth and 10th graders, respectively, at Pittsburgh’s South Fayette High School, designed a scenario in which Cozmo could be shrunk down and used in micro-surgeries. ''I didn’t know about a lot of the things that Cozmo could do. I learned that technology is more advanced than I thought,'' Khattab said. ''We used facial recognition, landmark navigation, state machine programming, and voice generation,'' Guetari said. Khattab and Guetari, who said she plans to study robotics in college, won the Multimedia Excellence in Audio Design and the Senior Division Award. ''In the future, AI education will be as common as math skills or language skills, so we want students to, from a young age, learn what AI really means,'' said Andrew Chen, chief development officer of WholeRen Group. The main component of the AI-In-A-Box is the Cozmo Robot, the palm-sized creation of three CMU alumni who founded the technology company Anki after graduating from CMU’s Robotics Institute. Also in the box was programming software called Calypso, designed by CMU computer science professor Dave Touretzky, an Amazon Fire 7 tablet, an Xbox controller and an instruction book for students and teachers. Touretzky taught some of the week’s workshops and was one of the judges of the competition. His software Calypso ''lets Ph.D. level programming be done by 8-year-olds,'' he said. ''Conditional programming languages force students to work at a very concrete level — tell the robot, 'Move forward 50 mm. Turn 15 degrees.' It’s very tedious and it’s easy to make mistakes. Calypso works at a more abstract level. There are no coordinates. Instead, Calypso talks about objects — 'If you see a blue cube, move towards it.' So students are able to describe what they want Cozmo to do,'' Touretzky said. Because Cozmo is equipped with a camera, students are able to view what the robot ‘sees’ and give visual directives — another first in AI education. Sam Critelli, a lead software engineer for Anki and another participating judge, was especially impressed by a team of three 10-year-old girls from the South Fayette School District, who used Cozmo’s facial recognition capabilities to prompt the robot to bring them baked goods when they showed Cozmo a ''sad'' face. This team won the Intermediate Division Award for their project. ''I was blown away by how much they had learned in a couple of days. Young students have a raw intelligence that can yield amazing results when exercised,'' Critelli said. Other schools in the Pittsburgh area, including the Montour School District and Quigley Catholic School in Beaver County, have already incorporated the WAICY AI workshops into this year’s curriculum. Dominic Panucci, branch director of the Boys and Girls Club in Carnegie, said the afterschool program also is using the WAICY curriculum this year. ''We are looking to have these workshops at every club in Pittsburgh and then have a local competition between clubs,'' Panucci said. ''AI is a team-building activity,'' which is a focus of the Boys and Girls Club. CMU has a long history in artificial intelligence including the creation of the first AI computer program in 1956 and pioneering work in self-driving cars, facial recognition and natural language processing. CMU recently announced a new undergraduate degree in artificial intelligence — the first of its kind in the U.S. — providing students with in-depth knowledge of how to transform large amounts of data into actionable decisions.  

Filter News

News by Year