Motion Planning with Safety Constraints and High-Level Task Specifications
This work addresses the challenge of efficient sequential decision-making for robotics or autonomous systems, representing an incremental improvement over existing methods.
The paper tackles the problem of motion planning with safety constraints and high-level task specifications in unpredictable environments, achieving a significant reduction in computation time for optimal policy generation through a novel pruning step.
We present a method to solve planning problems involving sequential decision making in unpredictable environments while accomplishing a high level task specification expressed using the formalism of linear temporal logic. Our method improves the state of the art by introducing a pruning step that preserves correctness while significantly reducing the time needed to compute an optimal policy. Our theoretical contribution is coupled with simulations substantiating the value of the proposed method.