Optimal positioning of combined mobile platform-manipulator systems for material handling tasks

TLDR: This paper is concerned with a particular aspect of the resolution of this redundancy, which is its utilization to optimize the system's position and configuration during task commutations when changes occur in both task requirements and task constraints.

Abstract: Mobile manipulators are attracting significant interest in the industrial, military, and public service communities because of the potential they provide for increased efficiency in material handling and manipulation tasks. Corresponding interest has arisen in the robotics research community since the combination and coordination of the mobility of an autonomous platform with the robotic motion of a manipulator introduce complex analytical problems. One such problem arises from the particular kinematic redundancy which characterizes practical mobile manipulators. This paper is concerned with a particular aspect of the resolution of this redundancy, which is its utilization to optimize the system's position and configuration during task commutations when changes occur in both task requirements and task constraints. Basic optimization schemes are developed for cases when load and position constraints are applied at the end-effector. Various optimization criteria are investigated for task requirements including obstacle avoidance, maneuverability and several torque functions. The problem of optimally positioning the platform for execution of a manipulation task requiring a given reach is also treated. Emphasis is then placed on the multi-criteria optimization methods which are necessary to calculate the commutation configurations in sequences of tasks with varying requirements. Sample results are presented for a system involving a three-link planar manipulator on a mobile platform. The various optimization schemes are discussed and compared, and several directions (in particular the novel use of minimax optimization pioneered here for redundancy resolution) are outlined for further extensions of the methods to the general problem of motion planning and control of redundant robotic systems with combined mobility and manipulation capabilities.