Outdoor navigation: time-critical motion planning for nonholonomic mobile robots.

Abstract

The question of timing in mobile robot navigation still remains an area of research not thoroughly investigated. In certain situations, a mobile robot may need not only to reach a desired location safely, but to arrive at that location at a specified time. Such a situation may have significant ramifications for applications to which a robot is tasked, for example patrolling large areas, delivering goods or coordinating multiple mobile robots. Thus, it is important for a mobile robot to be able to plan its trajectories and movements in order to navigate from initial location to a final destination whilst considering timing, orientation and velocity. Furthermore, it should also be able to detect and avoid any obstacles encountered in its path during navigating through the environment. The aim of this research is therefore to develop a time-critical motion planning algorithm, which includes planning the trajectory, position and orientation of a mobile robot, with obstacle avoidance capability for a single or multiple nonholonomic mobile robots. In addition, the mobile robot should be able to replan its original trajectories in order to ‘make up’ any loss of time caused by avoiding obstacles. An Ackermann car-like robot has been considered specifically during the development stage, with consideration given to the kinematic and dynamic constraints of nonholonomic mobile robot in general. The resultant algorithm is based on the geometric approach. In achieving the research objectives, this study is conducted in four stages. The first stage deals with the development of a new algorithm for time-critical motion planning in order to navigate safely in an environment, to reach the specified location at the specified time, with the required orientation, velocity and with the consideration of the kinematic and dynamic constraints of the mobile robot. In the second stage, the algorithm should have the capability to avoid any unknown static and dynamic obstacles when the mobile robot starts to move from its initial point. The algorithm should have the ability to replan its original trajectory to compensate for time loss due to avoiding obstacles. Prior to experimental works, the simulations will be carried out to ascertain the effectiveness of the algorithm. In the final stage, experimental works will be undertaken to validate the algorithms utilising an Ackermann car-like robot.