The conference Robotics: Science and Systems was held at the University of Washington in Seattle, from June 28 to July 1, 2009. More than 300 international researchers attended this single-track conference to learn about the most exciting robotics research and most advanced robotic systems. The program committee selected 39 papers out of 154 submissions. The program also included invited talks. The plenary presentations were complemented by workshops.
Drew McDermott Research on planning for robots is in such a state of flux that there is disagreement about what planning is and whether it is necessary. We can take planning to be the optimization and debugging of a robot's program by reasoning about possible courses of execution. It is necessary to the extent that fragments of robot programs are combined at run time. There are several strands of research in the field; I survey six: (1) attempts to avoid planning; (2) the design of flexible plan notations; (3) theories of time-constrained planning; (4) planning by projecting and repairing faulty plans; (5) motion planning; and (6) the learning of optimal behaviors from reinforcements. More research is needed on formal semantics for robot plans.
This new conference series promotes multidisciplinary research on tools and methodologies for efficiently capturing knowledge from a variety of sources and creating representations that can be (or eventually can be) useful for reasoning. The conference attracted researchers from diverse areas of AI, including knowledge representation, knowledge acquisition, intelligent user interfaces, problem solving and reasoning, planning, agents, text extraction, and machine learning. Knowledge acquisition has been a challenging area of research in AI, with its roots in early work to develop expert systems. Driven by the modern internet culture and knowledge-based industries, the study of knowledge capture has a renewed importance. Although there has been considerable work over the years in the area, activities have been distributed across several distinct research communities.
Modeling various aspects of language--syntax, semantics, pragmatics, and discourse, among others--by the use of constrained formal-computational systems, just adequate for such modeling, has proved to be an effective research strategy, leading to deep understanding of these aspects, with implications for both machine processing and human processing. This approach enables one to distinguish between the universal and stipulative constraints.
The welcome was given by University of Pittsburgh President Wesley Posvar. The conference cochairmen, Stellan Ohlsson and Jeff Bonar, also gave brief welcomes to the participants. The relatively small size of the conference, about 425 participants, was undoubtedly in part responsible for the congenial ambiance of the meeting. In addition to the opportunity to reunite with old friends, it was easy to establish new relationships with nearly everyone at the conference. With so many attendees from abroad (The Netherlands, Japan, Canada, West Germany, England, Sweden, France, and Hong Kong were all represented by speakers), the international flavor of the conference was well established.
Thsi article is a slightly modified version of an invited address that was given at the Eighth IEEE Conference on Artificial Intelligence for Applications in Monterey, California, on 2 March 1992. It describes the lessons learned in developing and implementing the Artificial Intelligence Research and Development Program at the National Aeronautics and Space Administration (NASA). In so doing, the article provides a historical perspective of the program in terms of the stages it went through as it matured. These stages are similar to the "ages of artificial intelligence" that Pat Winston described a year before the NASA program was initiated. The final section of the article attempts to generalize some of the lessons learned during the first seven years of the NASA AI program into AI program management heuristics.
It was motivated by two observations regarding modeling in general and work in qualitative physics in particular. First, all modelbased reasoning is only as good as the model used (Davis and Hamscher 1988). Second, no single model is adequate or appropriate for a wide range of tasks (Weld 1989). A model of a real-world system is but an abstraction of some aspects of the system. To formulate a model of a physical system for a given task, we inevitably take certain perspectives of the system to capture proper scenarios by deciding what to describe and what to ignore (Hobbs 1985).
Sensor dependency is an affliction that affects an alarming number of robots, and the problem is spreading. In some situations, sensor use is advisable, perhaps even unavoidable. However, there is an important difference between sensor use and sensor abuse. This article lists some of the telltale signs of sensor dependency and reveals the tricks of the trade used on unwitting roboticists by wily sensor pushers.