describe the Urban Challenge and Boss, the vehicle that won the challenge. Boss is a modified Chevy Tahoe that fuses data from many sensors to interpret the ...
Plenary Talks The Real World Charles Morefield, DARPA Abstract: The mission of the Information Processing Techniques Office (IPTO) is to understand the world. From sensing to cognition, IPTO brings the future of computing to the warfighter. In a world where complexity and ambiguity reign, IPTO technologies lift the fog of war and enable timely, accurate decision-making. Biography: Dr. Charles Morefield is Director of the Information Processing Techniques Office (IPTO), which spans the information space from core computer science through applications. His job is to identify promising new threads of information technology, translating them into viable DARPA programs. Internally he is focused on IPTO program managers, helping them develop proposals for new programs, reviewing existing programs, and attracting and retaining a high quality program management staff. Externally, he builds relationships with the academic community, private industry, and government entities than DARPA to understand emerging needs, initiate programs, and transition IPTO products to end users. Prior to joining DARPA in 2007, Dr. Morefield was Vice President of BAE Systems Inc. National Security Solutions. At BAE he provided leadership for research in command and control, information fusion, social computing, and information assurance. His entire career has been spent either performing advanced research or managing brilliant people He has founded or served on the board of a number of research companies.Dr. Morefield holds a Ph.D. in Applied Mathematics & Engineering Sciences from University of California San Diego. He served for a decade on the US Air Force Science Advisory Board, and as LT(jg) in the US Navy. He is entitled to wear the USAF Decoration for Exceptional Civilian Service and the Vietnam Service Ribbon.
The Urban Challenge Chris Urmson, Carnegie Mellon University Abstract: The Urban Challenge was a robotic vehicle race through a simulated urban environment. Full size autonomous vehicles were required to complete a 60 mile course, while independently reasoning about other autonomous and human driven vehicles. The vehicles were required to safely handle intersections, multi-lane roads, parking lots and unusual situations. Teams from around the world attempted the competition with eleven qualifying for the final challenge. In this talk I describe the Urban Challenge and Boss, the vehicle that won the challenge. Boss is a modified Chevy Tahoe that fuses data from many sensors to interpret the world around it and drive safely. I will highlight how Boss incorporates radar and lidar data to track moving vehicles and how this information is used. I will also speculate on the future of autonomous vehicles and the critical open challenges. Biography: Chris Urmson is an assistant research professor at Carnegie Mellon University and a member of the technical staff at Google. He was the Director of Technology for Tartan Racing at Carnegie Mellon University, helping the team to win the 2007 DARPA Urban Challenge. Chris has also worked as a robotics research scientist with SAIC. Chris has developed numerous robotic navigation architectures and software systems currently in use by Carnegie Mellon University, NASA JPL and NASA Ames. He has made significant contributions to the development of over a half dozen robots, with an emphasis on software development and system integration. He earned his PhD in 2005 from
10
Carnegie Mellon University and his B.Sc. in Computer Engineering from the University of Manitoba in 1998. Chris has earned a variety of corporate and academic awards including being named a Siebel Scholar, and receiving technology innovation awards from Boeing Phantom Works and SAIC.
Likelihood Ratio Detection and Tracking Lawrence Stone, Metron, Inc. Abstract: Likelihood Ratio Detection and Tracking (LRDT) is an extension of Bayesian tracking that simultaneously estimates whether a target is present and its state if present. It is particularly suited to difficult detection and tracking situations involving low signal-to-noise ratios or high clutter density. LRDT extends the basic Bayes Markov single target recursion by adding a null state to represent the possibility of no target present. By converting to likelihood ratios, one obtains a recursion which is parallel to the Bayes Markov single target recursion and does not explicitly retain the null state as a separate state. This allows LRDT to use un-thresholded or below-threshold sensor responses and to provide a natural and correct method of incorporating information from multiple sensors and disparate sensor types. LRDT performs incoherent integration of sensor responses over time by cumulating measurement likelihood ratios over possible target paths (tracks). When the cumulative likelihood ratio exceeds a specified threshold, a detection is called and a state estimate is extracted. LRDT is a recursive, Bayesian, Track-Before-Detect (TBD) system that does not require explicit association of sensor responses to target tracks. This allows LRDT to consider a vastly larger set of possible target paths than TBD systems based on multiple-hypothesis tracking or other track-based techniques. We give examples of the application of LRDT to (1) the detection of periscopes by surface ship radar, (2) the detection of submarines by multi-static active sonar, and (3) the automatic detection and tracking of acoustic sources by passive acoustic arrays. The similarity of certain forms of the LRDT recursion to those of the PHD and Multitarget Intensity filters is also discussed. We close by considering some non-traditional extensions of LRDT beyond classical detection and tracking such as monitoring the maritime supply chain to detect suspicious behavior. Biography: Dr. Stone joined Metron in 1986. He became President and CEO of Metron in 2004. He is a member of National Academy of Engineering and a fellow of the Institute for Operations Research and Management Science.In 1975, the Operations Research Society of America awarded the Lanchester Prize to Dr. Stone's text, Theory of Optimal Search. He was codirector of the 1979 NATO Advance Research Institute on Search Theory and Applications in Faro, Portugal, and coeditor of the conference proceedings, "Search Theory and Applications." He has published numerous papers in search theory, taught the subject at the Naval Postgraduate School, and has participated in many search operations. He has also published papers in probability theory and optimization. In 1986, he produced the probability maps used by the Columbus America Discovery Group to locate the S.S. Central America which sank in 1857, taking millions of dollars of gold coins and bars to the ocean bottom one and one-half miles below. He is a coauthor of the 1999 book, Bayesian Multiple Target Tracking, and continues to work on a number of detection and tracking systems for the U. S. Navy and Coast Guard. Dr. Stone worked at Daniel H. Wagner, Associates from 1967 until 1986.
