The efficacy and efficiency of mobile robots operating in real-world environments are challenged when compared to operation in (semi-)protected lab settings, because of the increasing difficulty to avoid collisions. This dissertation provides an algorithmic framework for safe and reliable motion planning and control of a specific class of mobile robots, those can safely operate while withstanding physical collisions with the environment. The dissertation has two main focus directions: a control approach integrating collision exploitation at run-time, as well as a motion planning algorithm to determine how to utilize those collisions in unknown (or partially-known) environments.In the first focus direction, we first develop a state-feedback closed-loop control approach based on optimal steering that features a collision switching strategy. Collisions are found beneficial in terms of increasing task success probability when steering a robot from an initial to a target spatial distribution. Then we propose a new deformation recovery and replanning strategy to handle collisions that may occur at run-time. Contrary to collision avoidance methods that generate trajectories only in conservative local space or require collision checking that often comes with high computational cost, our method directly generates (local) trajectories with imposing only waypoint constraints. If a collision occurs, our method then estimates the post-impact state and computes from there an intermediate waypoint to recover from the collision.
In the second focus direction, we first introduce a new sampling-based online planning algorithm that can explicitly handle the risk of colliding with the environment and can switch between collision avoidance and collision exploitation. This way, the planner can make deliberate decisions to collide with the environment if such collision is expected to help the robot make progress toward its goal. Then we propose a search-based planning algorithm to determine how to best utilize potential collisions to improve certain metrics, such as control energy and computational time.
Taken together, we derive a hierarchical framework generated by combining the global seach-based planning algorithm and the local deformation recovery and replanning strategy. Results demonstrate our method's efficacy for collision-inclusive motion planning and control in unknown environments with isolated obstacles for a class of impact-resilient robots operating in 2D.