MORRFx and Its Framework: Multi-objective Sampling Based Path Planning for Unpredictable Environments

We present MORRFx, an asymptotically optimal sampling based motion planning algorithm for fast and multi-objective planning in unpredictable dynamic environments. Often in robotics applications, the representation of the environment can change. As a result, over time, the information used to compute the solutions can evolve. The algorithm we propose takes these modifications into account to produce an approximation of the Pareto optimal set of solutions, that are valid at any time, and from which the decision-maker can select a trajectory according to his/her preferences. MORRFx (Multi-Objective Rapidly exploring Random Forest x) is a combination of two algorithms: RRTx (Asymptotically Optimal Single-Query Sampling-Based Motion Planning with Quick Replanning) [1] and MORRF* (Sampling-Based Multi-Objective Motion Planning) [2]. In this paper, we present a minimal, multi-objective framework in an efficient implementation. We also use simulations to show the efficiency and capabilities of MORRFx.