Fast and Compute-efficient Sampling-based Local Exploration Planning via Distribution Learning

TL;DR

This paper introduces a learning-based approach for exploration planning in robotics that improves efficiency and performance by modeling informative view distributions and information gain, outperforming classical methods.

Contribution

It presents a novel method to learn sampling distributions and information gain for exploration, enhancing efficiency and adaptability in varying environments.

Findings

Up to 28% improvement in exploration performance.

Learning gains offers better compute-performance trade-offs.

System generalizes well to different environments.

Abstract

Exploration is a fundamental problem in robotics. While sampling-based planners have shown high performance, they are oftentimes compute intensive and can exhibit high variance. To this end, we propose to directly learn the underlying distribution of informative views based on the spatial context in the robot's map. We further explore a variety of methods to also learn the information gain. We show in thorough experimental evaluation that our proposed system improves exploration performance by up to 28% over classical methods, and find that learning the gains in addition to the sampling distribution can provide favorable performance vs. compute trade-offs for compute-constrained systems. We demonstrate in simulation and on a low-cost mobile robot that our system generalizes well to varying environments.