Real-time adaptive robot motion planning
Jing Xiao, John Vannoy · 2007
In order for a robot to perform a task, its motion is usually planned beforehand. This assumes that nothing in the environment changes as the robot performs the task. In the real world, there are uncertainties and changes in a robot's environment that cannot be known a priori. Offline robot motion planning is not enough and will not work in a highly unpredictable and dynamic environment. This dissertation introduces a novel and general paradigm for real-time adaptive motion planning (RAMP) for high-degrees-of-freedom (DOF) robots such as mobile manipulators or other high-DOF robots. The RAMP approach enables simultaneous path and trajectory planning and simultaneous planning and execution of motion in real-time. It effectively deals with drastic changes in the environment through global planning of diverse trajectories and through further preservation of diversity if needed. It facilitates real-time optimization of trajectories under various optimization criteria, such as minimizing energy and time and maximizing manipulability. It also accommodates partially specified task goals of robots easily. The approach exploits loose-coupling of locomotion and manipulation of a mobile manipulator, taking advantage of the redundancy to best achieve obstacle avoidance and various optimization objectives. The RAMP paradigm is suitable not only for planning motion of a single robot in an environment of unknown dynamics, but also for planning the motion of a team of robots in the same environment. It enables distributed, real-time planning of motions for a team of robots to collaborate spontaneously in the same environment for a common task objective. It can also plan motions for two or more robots in tight coordination to manipulate a common object in a dynamically unknown environment.