Date of this Version
Land Use, Infrastructure and Transportation, Urban Data Systems
We present a general framework for generating trajectories and actuator forces that will take a robot system from an initial configuration to a goal configuration in the presence of obstacles observed with noisy sensors. The central idea is to find the motion plan that optimizes a performance criterion dictated by specific task requirements. The approach is motivated by studies of human voluntary manipulation tasks that suggest that human motions can be described as solutions of certain optimization problems.
Date Posted: 13 July 2006