Anytime Replanning of Robot Coverage Paths for Partially Unknown Environments (2311.17837v2)

Abstract: In this paper, we propose a method to replan coverage paths for a robot operating in an environment with initially unknown static obstacles. Existing coverage approaches reduce coverage time by covering along the minimum number of coverage lines (straight-line paths). However, recomputing such paths online can be computationally expensive resulting in robot stoppages that increase coverage time. A naive alternative is greedy detour replanning, i.e., replanning with minimum deviation from the initial path, which is efficient to compute but may result in unnecessary detours. In this work, we propose an anytime coverage replanning approach named OARP-Replan that performs near-optimal replans to an interrupted coverage path within a given time budget. We do this by solving linear relaxations of integer linear programs (ILPs) to identify sections of the interrupted path that can be optimally replanned within the time budget. We validate OARP-Replan in simulation and perform comparisons against a greedy detour replanner and other state-of-the-art coverage planners. We also demonstrate OARP-Replan in experiments using an industrial-level autonomous robot.

• The paper introduces OARP-Replan, an anytime replanning algorithm that efficiently adapts robot coverage paths when unexpected obstacles are encountered.
• The paper employs a two-step MILP approach to replan coverage ranks and connect them, reducing unnecessary turns and computational overhead.
• Simulations demonstrate that OARP-Replan outperforms greedy and offline methods, ensuring robust performance in dynamic, real-world settings.

Overview

In robotics, ensuring efficient coverage of an area is a crucial task, particularly in environments that are not fully known in advance. Robots often face challenges such as avoiding newly detected obstacles while trying to cover an area completely. This paper introduces an innovative method to address the problem of replanning coverage paths when robots encounter unexpected obstacles in their environment.

Replanning Coverage Paths

The paper focuses on developing an anytime replanning approach for optimizing coverage paths called "OARP-Replan." When a robot engaged in a coverage task discovers an obstruction, conventional methods may either compute a completely new coverage path or take a naive detour, which can be computationally expensive or inefficient. OARP-Replan smartly modifies the current path to accommodate newly detected obstacles, focusing on near-optimal solutions.

To minimize the number of turns—which is often associated with efficiency—OARP-Replan refashions paths by adjusting straight-line sections, also known as coverage ranks. It breaks down the replanning process into two steps using relaxed mixed-integer linear programs (MILPs). The first step involves replanning the coverage ranks within a certain time frame. The second step creates a cost-effective route that connects these ranks, effectively determining the robot's new path through the area.

Methodological Contributions

OARP-Replan’s unique contributions are threefold:

1. A MILP that quickly replans coverage ranks, keeping in mind that the resultant path should be computed within the robot's available time budget.
2. Introduction of an anytime coverage replanning algorithm named OARP-Replan that leverages MILPs to perform near-optimal replans under the given time constraints.
3. The approaches are validated through simulated trials in environments modelled after real-world conditions, demonstrating their superiority over existing state-of-the-art methods, including scenarios where all obstacles are known beforehand.

Simulations and Performance

Simulations are conducted to evaluate the performance of OARP-Replan. These simulations are set up using maps based on real-world environments, to which unknown obstacles have been added. The proposed approach shows better performance than both a greedy detour replanner and a modified version of an existing state-of-the-art offline coverage approach under various conditions, revealing significant improvements in the time required to cover the area. Additionally, a comparative analysis with an offline planner is presented, revealing that OARP-Replan can outperform even with full knowledge of the environment, owing to its efficiency in path optimization.

Conclusion and Implications

The proposed OARP-Replan demonstrates a promising new direction for online coverage path planning in robotics. It manages to provide efficient coverage planning in dynamic and partially known environments by adaptively replanning to navigate unexpected obstacles. This approach could be particularly useful in applications such as automated cleaning, warehouse management, and agricultural robotics where environments change frequently, and disruption-free operation is essential. The ability to adapt in real-time without significantly increasing the total path length or coverage time suggests that such intelligent replanning could be incorporated into a wide range of robotic systems. The next steps could involve scaling this replanning method to more complex and dynamic environments, improving its applicability to a broader spectrum of real-world scenarios.