Anytime Replanning of Robot Coverage Paths for Partially Unknown Environments

TL;DR

This work proposes an anytime coverage replanning approach named OARP-Replan that performs near-optimal replans to an interrupted coverage path within a given time budget by solving linear relaxations of integer linear programs to identify sections of the interrupted path that can be optimally replanned within the time budget.

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.

Figures (18)

• Figure 1: Replanning coverage paths in partially unknown environments. (a) An initial coverage path for the known base environment, (b) the robot (blue triangle) covers part of the environment (gray region) and observes a new obstacle (red box) interrupting the initial path (red path), and (c) a near-optimal replanned path (green path) computed online using the proposed approach (OARP-Replan).
• Figure 2: The effect of an obstacle (black box) on replanning a coverage path consisting of ranks (black lines) and transition paths (green lines). The ranks of the initial path (left) when interrupted can be replanned in many ways, two of which are shown here. Example 1 has fewer new ranks (red lines in dotted box) than example 2 and is preferable for shorter planning time. Example 2 has fewer total ranks and is preferable for shorter drive time.
• Figure 3: A run of OARP-Replan with one obstacle (dotted grey box). The visible part of the obstacle (pink dots) is used to identify blocked grid cells (grey) and obstacle encounters (blue dots) along the current path. For each encounter $i \leq k$, the planner attempts to replan the remaining path (red path) within the time $\tau_i$ to reach the encounter (black path). To compute $\tau_i$, previous encounters are replanned using GD replan (green dotted lines).
• Figure 4: Integral orthogonal polygon (IOP) of an environment where each grid cell is either oriented horizontally (orange) or vertically (purple). The left, right, top, and bottom rank endpoints are also shown (see legend). Each cell $c_i$ (zoomed-in cell) has four endpoint indicator variables. If the cell is an endpoint (e.g. $c_i$ is a right endpoint), the corresponding variable is active ($y_R^i = 1$).
• Figure 5: The directed graphs $G_L, G_R, G_T,$ and $G_B$ for the example IOP.

Theorems & Definitions (7)

• Lemma 1
• proof
• Proposition 1
• proof
• Remark 1: Linear Relaxation
• Lemma 2
• proof