Automatic Identification of Remote Environments

A system that can process sensory information collected during telemanipulation tasks in order to automatically identify properties of the remote environment has many potential appli cations These include generating model based simulations for training operators in critical procedures and improving real time performance when time delays are large The research issues in volved in developing such an identi cation system are explored focusing on properties that can be identi ed from remote manip ulator motion and force data As a case study a simple block stacking task performed with a teleoperated two ngered planar hand is considered An algorithm is presented which automat ically segments the data collected during the task given only a general description of the temporal sequence of task events Using the segmented data the algorithm then successfully es timates the weight width height and coe cient of friction of the two blocks handled during the task This case study high lights the broader research issues which must be addressed in automatic property identi cation