Path quality measures for sensor-based motion planning

Abstract

A mobile robot requires the perception of its local environment for position estimation. Traditional planning methodologies build plans by considering the obstacles position in the environment, and assume that the sensorial system is able to localize the robot. Not all sensorial systems are able to maintain the robot accurately localized at each instant and require a perception planning process to localize the robot at certain points. Recent works of different researchers, tend to overcome the problem of simultaneous perception-action planning. To plan simultaneously perceptual localization activities and movements it is required to determine the robot’s position uncertainty evolution and how does this uncertainty affect the perceptual problem in order to localize the mobile robot. The proposed algorithm provides a methodology to evaluate the uncertainty in an indoor semi-structured environment, which can be used to plan the robot’s movements and localization perception actions, and achieve an integrated plan able to reach the goal. The planning includes the position uncertainty analysis and the probability of success calculation as a path quality measurement. The resultant module has been integrated successfully in a more complex navigation system. Experiments demonstrate the effectiveness of the proposed method in real world applications.