Balancing exploration and exploitation in motion planning

Computationally efficient motion planning must avoid exhaustive exploration of configuration space. We argue that this can be accomplished most effectively by carefully balancing exploration and exploitation. Exploration seeks to understand configuration space, irrespective of the planning problem, while exploitation acts to solve the problem given the available information obtained by exploration. We present an exploring/exploiting tree (EET) planner that balances its exploration and exploitation behavior. The planner acquires workspace information and subsequently uses this information for exploitation in configuration space. If exploitation fails in difficult regions, the planner gradually shifts its behavior towards exploration. We present experimental results demonstrating that adaptive balancing of exploration and exploitation leads to significant performance improvements compared to other state-of-the-art sampling-based planners.

I. Introduction

Sampling-based motion planners routinely solve complex, high-dimensional planning problems. This may seem surprising, considering that the general motion planning problem [1] is PSPACE-hard [2]. However, the configuration spaces of many practical problems contain considerable structure that may help in solving a planning problem. In addition, not all parts of configuration space have to be explored to solve a particular motion planning problem. Today's sampling-based planners leverage these properties of practical motion planning problems to achieve computational efficiency.