RIS: A Framework for Motion Planning Among Highly Dynamic Obstacles

Abstract

We present here a framework to integrate into a motion planning method the interaction zones of a moving robot with its future surroundings, the reachable interaction sets. It can handle highly dynamic scenarios when combined with path planning methods optimized for quasi-static environments. As a demonstrator, it is integrated here with an artificial potential field reactive method and with a Bézier curve path planning. Experimental evaluations show that this approach significantly improves dynamic path planning methods, especially when the speeds of the obstacles are higher than the one of the robot. The presented approach is used together with a global planning approach in order to handle complex static environments in presence of fast-moving obstacles. When the ego vehicle is not holonomic the presented approach is able to take dynamic constraints into account, which improve the prediction accuracy.

Review of Decision-Making and Planning Approaches in Automated Driving

“…First, the RIS constrains free space by considering the risk of collision among the obstacles and the ego-vehicle; and secondly, a path is computed by the interpolation of Bézier curves surrounding the RIS and allowing the robot to move up to its destination. [37] 3) Simultaneous path and speed planning approaches…”

“…An MPC-based approach for motion planning in overtake scenarios was presented in [24]. The authors combined reachable sets, to iteratively generate reference targets based on the current maneuver, together Interpolating curves [33], [89], [34], [35], [36], [37] Interpolation of quintic Bézier curves for trajectory planning in urban environments; interpolation of consecutive quartic Bézier curves optimizing consecutive curves for generation smooth trajectories; B-spline curves for the interpolation of centerline of reference lane for candidate paths generation. Cubic polynomial splines for speed profile generation; Clothoids based path tentacles in dynamic environments for obstacle avoidance; assembling of circular arcs, line segments and clothoids for collision avoidance; RIS combined with Bézier curves for path generation in semiconstrained highly dynamic environments [33], [34], [35], [36], [37] [35]…”

“…[33], [34], [35], [36], [37] [35], [36], [37] Graphsearch based [49], [38] State lattices based discretization combined with splines for the paths set generation for fast real-time planning in semi-structured environments; A* graph-search combined with RRT for the navigation of an AV through an unmapped road scenario [38] [38] [38] [49], [38] Samplingbased [35], [48], [39], [16] Evidential occupancy grid for modeling the environment and allowing to represent the uncertainty, combined with clothoid tentacles; navigation through clothoid tentacles selection with a high-level maneuver planner for the obstacle avoidance application; occupancy grid discretization with optimization based path generation; sampling-based motion planner with tactical maneuver discovery reasoning [39], [16] [35] [35], [48] [35], [48], [39], [16] Optimizationbased [40], [41], [50], [42] Path and speed profile generation by iterative optimization, combining dynamic programming with quadratic splines; convex optimization for optimal speed profile generation for static and dynamic environments; speed profile generation based on optimal control; MPC-based simultaneous path and speed planning for obstacle avoidance scenarios [40], [41], [42] [42]…”