C*: A Coverage Path Planning Algorithm for Unknown Environments using Rapidly Covering Graphs

2025-05-20

AI Summary

Real-time complete coverage path planning in unknown environments remains challenging due to inefficiencies in map representation, trajectory sparsity, and coverage gaps. Method: This paper proposes C*, a sampling-based algorithm built upon the Rapid Coverage Graph (RCG), which enables incremental mapping without grid decomposition and online dynamic pruning for non-myopic, sparse trajectory generation. It innovatively integrates local Traveling Salesman Problem (TSP) optimization to detect and immediately fill obstacle-surrounded voids. Contribution/Results: C* is theoretically proven to guarantee complete coverage. Extensive simulations and real-robot experiments demonstrate that C* significantly reduces coverage time, turning frequency, path length, and overlap ratio, while eliminating coverage gaps entirely. Moreover, it successfully adapts to both energy-constrained single-robot and multi-robot cooperative scenarios.

Technology Category

• Intelligent Robots: Localization, Mapping, and Navigation
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web

Abstract

The paper presents a novel sample-based algorithm, called C*, for real-time coverage path planning (CPP) of unknown environments. The C* algorithm is built upon the concept of Rapidly Covering Graph (RCGs). The RCG is constructed incrementally via progressive sampling during robot navigation, which eliminates the need for cellular decomposition of the search space. The RCG has a sparse-graph structure formed by efficient sampling and pruning techniques, which produces non-myopic waypoints of the coverage trajectory. While C* produces the desired back and forth coverage pattern, it adapts to the TSP-based locally optimal coverage of small uncovered regions, called coverage holes, that are surrounded by obstacles and covered regions. Thus, C* proactively detects and covers the coverage holes in situ, which reduces the coverage time by preventing the longer return trajectories from distant regions to cover such holes later. The algorithmic simplicity and low computational complexity of C* makes it easy to implement and suitable for real-time onboard applications. It is analytically proven that C* provides complete coverage of unknown environments. The performance of C* is validated by 1) extensive high-fidelity simulations and 2) real laboratory experiments using autonomous robots. A comparative evaluation with seven existing CPP methods demonstrate that C* yields significant performance improvements in terms of coverage time, number of turns, trajectory length and overlap ratio, while preventing the formation of coverage holes. Finally, C* is evaluated on two different applications of CPP using 1) energy-constrained robots and 2) multi-robot teams.

Problem

• Real-time coverage path planning for unknown environments
• Efficiently detects and covers small uncovered regions
• Reduces coverage time and prevents coverage holes

Innovation

• Uses Rapidly Covering Graphs for path planning
• Incrementally constructs sparse-graph structure
• Proactively detects and covers coverage holes