9320419 Khatib The general thrust of this effort is to develop mobile manipulation systems with ever-increasing degrees of autonomy. The equipment, two mobile robots with manipulators, will be used for the research projects outlined below: Robotic Material Handling Assistants: Autonomous transport of material through complex domains is a task that might be considered beyond the current capability of robotic systems. Construction areas are a prime example of such an environment. However, a set of cooperating robotic "assistants" with human guidance could accomplish just such a task, increasing productivity and decreasing the risk of injury. The goal in this project is to develop such a set of mobile robotic assistants. Initially, these robots would be guided in their actions by a human "master." As the project progresses, ever-increasing levels of autonomy would be exhibited, increasing the dexterity of the system. Visual Sensing for Mobile Robots: The project is aimed to develop real-time visual sensing capabilities for mobile robots. This goal offers a considerable challenge, but it is a crucial step that must be taken to bring mobile robots into real-world unstructured task environments. This work starts by implementing and evaluating a variety of vision algorithms using two experimental setups: a stereo platform and motion analysis stage. The results of the research will be both a systematic comparison, and possibly combination, of existing algorithms, as well as a benchmark of calibrated image data to be used at Stanford and by others in the field. Cooperation among Autonomous Mobile Age nts: Several aspects of cooperation will be investigated, including communication modality, cooperation strategy, type of information or service exchanged, and degree of responsibility exchanged. The focus will be on agents that have advanced cognitive and physical capabilities and considerable autonomy. Their interactions will be discretionary and opportunistic, reflecting an explicit (but usually incomplete) understanding of their own capabilities, goals, and plans, along with those of their cooperative partners. Research on Autonomous Agents: In this project, the goal is to provide robotic systems with a high degree of run-time "conditionality" which are capable of adapting to complex and changing environments. *** ` $ AUTOEXEC33 Xt V LM VW @$ CONFIG BAK V CONFIG OLD nW WIN31 X 9320419 Khatib The general thrust of this effort is to develop mobile manipulation systems with ever-increasing degrees of y { 2 4 $ $ $ $ G M M Times Symbol " Helvetica Chicago Times New Roman & Arial 5 Courier New R ZapfDingbats Palatino Greek GenMath MathMeteor MT Extra &&& " b eFj 8 Khatib/Stanford Mark Purvis Mark Purvis
