Keynote Address 1: Rescue Robotics: A Grand Challenge

Abstract: Research in rescue robotics should be considered a worthy challenge in part because they are an excellent example of a technology of benefit to society. In an urban disaster, such as a building collapse, 80% of the survivors are found on the surface of the disaster, with only 20% from within the rubble. While 80% of the survivors are from the surface, the majority of the victims may be deep with the collapsed structure. Thus, many trapped people may die due to the limitations of traditional search and rescue techniques in quickly searching more than 6 meters into the pile. In these situations, ground robots can enter through small voids and penetrate through fire or oxygen depleted areas to find survivable voids. In a large, geographically distributed disaster, such as an earthquake, hurricane, tsunami, or wildfire, information on the extent of the disaster and state of the transportation infrastructure is needed within the first twelve hours to optimize the allocation of response and recovery assets. Here, unmanned aerial vehicles can provide “information on demand” to responders, rather than scheduling and coordinating manned aircraft. Rescue robots are more than a worthy challenge, they pose the multi-disciplinary challenge of making truly acceptable systems. In the past 5 years, ground, aerial, and water-based unmanned systems have been inserted into urban disaster response and underground mine rescue, but these systems have not been adopted by rescue agencies. I believe that this delay is the result of two distinct challenges: one technological, one about the nature of the diffusion of innovation. Since 1995, my research group has investigated rescue robotics, striving to identify the key technological barriers. Through the Center for Robot-Assisted Search and Rescue, I have participated in five disasters. At the World Trade Center 9/11 Disaster (2001), military proto-type man-packable ground robots were used for the first time for urban search and rescue. The robots performed better than expected. However, improvements to robot types failed at the La Conchita, California, mudslides (2005), uncovering several mechanical weaknesses that had largely gone undetected in the relatively benign conditions used to test human and canine responders. Hurricane Charley (2004) showed that ground robots are not useful for large scale, residential disasters. At Hurricane Katrina (2005), we introduced two types of micro aerial vehicles that were tasked to survey areas responders could not reach due to flooding and downed trees. We conducted extensive post-Katrina structural inspection with rotary-wing aerial vehicles, creating new protocols that are being adopted by the FAA. Later during that season, we introduced unmanned surface vehicles at Hurricane Wilma to inspect damage to sea walls, bridges, and waterways as well as trace environment hazards, such as leaking fuel lines, etc.

These experiences have identified major gaps between current robot capabilities and mobility, wireless communications, sensing (especially computer vision), artificial intelligence, power, and human-robot interaction needs. The research demands within each of these distinct areas as well as the larger systems engineering issues make rescue robotics a formidable research topic.

While the gaps in functionality and reliability are certainly a contributor to the lack of adoption, our work in pro-actively creating awareness training and responder-oriented exercises indicate the larger socio-organizational culture of emergency response also is a factor. Therefore rescue robotics can be also cast as a diffusion of innovation grand challenge: how can technologists design rescue robots to facilitate adoption? We are exploring models of technology diffusion which shed insight into the adoption process, and attempting to use these models to effect a shift in thinking about control, automation, robotics, and computer vision which will hopefully led to more rapid adoption.

Biography: Robin Roberson Murphy received a B.M.E. in mechanical engineering, a M.S. and Ph.D in computer science (minor: Computer Integrated Manufacturing Systems) in 1980, 1989, and 1992, respectively, from Georgia Tech, where she was a Rockwell International Doctoral Fellow. She is a Professor in the Computer Science and Engineering Department at the University of South Florida with a joint appointment in Cognitive and Neural Sciences in the Department of Psychology. From 1992 to 1998, she was an assistant professor in the Department of Mathematical and Computer Sciences at the Colorado School of Mines. Prof. Murphy joined USF in 1998, and in Jan., 2002, she became Director of the Center for Robot-Assisted Search and Rescue. In March 2003, she helped start the industry/university cooperative research center on Safety Security Rescue (SSR-RC) with the University of Minnesota and is the overall director. She leads the CRASAR rescue robot response team, the only such team in the world, and is a technical search specialist with Florida Task Force 3.

Though best known for her foundational work in rescue robotics, Dr. Murphy has concentrated her basic research on sensor fusion, distributed sensing, and fault tolerant perception for mobile robots. These efforts are/have been funded by DoE (RIM), DARPA (including the Tactical Mobile Robots, Unmanned Ground Combat Vehicle, PerceptOR and Synergistic Cyber-Forces programs), ONR, NASA, NSF and industry, and have led to over 70 publications in the field, including the textbook Introduction to AI Robotics (MIT Press). AI and Mobile Robots: Case Studies of Successful Systems (MIT Press), which she co-edited with David Kortenkamp and Pete Bonasso was cited by Michael Critchon as source material for his best seller Prey. Since 1995, she has focused on Urban Search and Rescue (USAR) as the test domain for her research, leading to her participation in the first known use of robots for urban search and rescue at the WTC disaster. Her USAR robotics work has earned a NIUSR Eagle award, and she serves on the NIUSR Executive Board. She has also won a USF Outstanding Faculty Research Achievement Award (2003), and the Honor Society of Phi Kappa Phi, USF Chapter Artist and Scholar of the Year Award (2004). Dr. Murphy will be one of seven "Grand Challenges" speaker at the "Robots and Thought" Carnegie Mellon Robotics Institute 25th Celebration in Oct, 2004. She was profiled in the June 14, 2004, issue of TIME Magazine as an innovator in Artificial Intelligence.

Dr. Murphy is active in the robotics, industrial, and military communities. She co-chaired the DARPA/NSF Study on Human-Robot Interaction (with Dr. Erika Rogers, Cal Poly), with segments appearing in the May/June issue of IEEE Transactions on Systems, Man and Cybernetics on HRI). She is an associate editor for IEEE Intelligent Systems, was an IEEE Computer Society Distinguished Visitor Program speaker, and was the first woman to serve on the Executive Committee of the IEEE Robotics and Automation Society (her post was RAS Society Secretary, and now serves as Parlimentarian and member of the Administrative Committee). Dr. Murphy was a member of the 1998-9 Defense Science Study Group and is currently a member of the US Air Force Scientific Advisory Board and the DARPA ISAT. She recently served on DoD Air Platforms FY2004 Technology Area Review and Assessment and on the National Research Council Committee on Army Unmanned Vehicle Technology. She is a member of the Board of Directors for Continential Divide Robotics, which provides GPS and intelligent agent software for tracking parolees. Prior to graduate work, Dr. Murphy worked in the process control industry as a software project engineer.