This paper reports on and discusses a set of user experiments using the TREC 2003 Web interactive track protocol. The focus is on comparing humans and machine algorithms in terms of performance in a topic distillation task. We also investigated the effect of the search results layout in supporting the users' effort.We have demonstrated that machines can perform nearly as well as people on the topic distillation task. Given a system tailored to the task there is significant performance improvement and finally, given a presentation that supports the task, there is strong user satisfaction.

In industrial problems both the task and environment are often very complex. Local planners are not sufficient and, since tasks often involve changing the environment, global planners are limited by time constraints. This paper adopts the notion that pre-processing of the known environment, in order to generate a roadmap, can significantly improve on-line path planning tasks. The authors present results obtained from an instance of the Ariadne's clew algorithm for path planning problems within the steam generator of a nuclear power plant. An, algorithm is then described which performs further pre-processing on the known environment so that, when the environment changes during on-line planning, a valid, although not necessarily connected, roadmap always exists. This is the first step in a planner that could be used in evolving industrial environments.

Roadmaps, generated by pre-processing of the known environment, have been a popular approach to robot motion planning. Generally designed for static environments, roadmaps are often invalidated by any anomaly between the real world and the model of the environment. We investigate the effect on roadmaps when sensor information indicates such an error and describe an algorithm which, by performing further pre-processing on the known environment, generates a mapping between points in the workspace and the road map. Used during on-line planning the mapping enables a valid, although not necessarily connected, roadmap to be maintained if such an error is discovered. We then present a divide and conquer algorithm which uses guided random motion to attempt to reconnect paths affected by the error. This is the first step towards a global planner that could be used in evolving environments.

By dividing the path planning problem into two subproblems (local and global planning), we obtain a planner for redundant manipulators that is fast, effective and versatile. Local planning is achieved within a potential field paradigm, modelling a rigid linked manipulator with slightly flexible links. The planner can either be used within advanced tele-operated systems, or as path refinement within an autonomous planner. We use a geometric construction which we call the skeleton to provide a concise connection between the global and local planners. The global planner acquires critical geometric information from the skeleton and end-effector paths are extracted and used to direct the local planner as a first step to a geometric reasoning planner.>