Heterogeneous Multi-robot Task Allocation for Long-Endurance Missions in Dynamic Scenarios

TL;DR

The paper tackles heterogeneous MRTA for long-endurance missions under dynamic conditions by introducing recharge-aware capabilities, task fragmentation/relays, and coalition-based synchronization. It provides a comprehensive MILP formulation that captures robot heterogeneity, recharges, fragmentation, relays, and time coordination, complemented by a problem-specific heuristic to enable real-time planning. A mission planning and execution architecture supports online replanning and plan repair, improving robustness to delays and failures. Experimental results in a realistic UAV inspection use case demonstrate improved makespan, reliability, and plan quality, while showing the heuristic scales to larger problems where the exact MILP becomes intractable. Overall, the work contributes a unified framework that advances practical, scalable planning for complex, battery-constrained multi-robot missions.

Abstract

We present a framework for Multi-Robot Task Allocation (MRTA) in heterogeneous teams performing long-endurance missions in dynamic scenarios. Given the limited battery of robots, especially for aerial vehicles, we allow for robot recharges and the possibility of fragmenting and/or relaying certain tasks. We also address tasks that must be performed by a coalition of robots in a coordinated manner. Given these features, we introduce a new class of heterogeneous MRTA problems which we analyze theoretically and optimally formulate as a Mixed-Integer Linear Program. We then contribute a heuristic algorithm to compute approximate solutions and integrate it into a mission planning and execution architecture capable of reacting to unexpected events by repairing or recomputing plans online. Our experimental results show the relevance of our newly formulated problem in a realistic use case for inspection with aerial robots. We assess the performance of our heuristic solver in comparison with other variants and with exact optimal solutions in small-scale scenarios. In addition, we evaluate the ability of our replanning framework to repair plans online.

Figures (13)

• Figure 1: Complexity of the problem according to task categorization. Tasks are classified according to their decomposability and coalition size flexibility; a higher degree of freedom implies a planning problem that is harder to solve.
• Figure 2: Example with 3 robots executing 3 consecutive multi-robot relayable tasks (with different color code). For each task, displacement time is depicted in blue, waiting time in yellow and execution time in green. Each task is divided into 3 fragments ($n_t^f=3$) executed by coalitions of 2 robots ($n_t^r=2$). Robot 3 executes all tasks and recharges in between, while robots 2 and 1 relay each other to accompany robot 3. Left, solution where robots do not coordinate task execution; right, solution including time coordination constraints.
• Figure 3: Time coordination example with 5 robots performing a multi-robot relayable task, which is divided into 3 fragments ($n_t^f=3$) executed by coalitions of 4 robots ($n_t^r=4$). Recharge tasks are not shown. Dashed black arrows indicate that the corresponding relay variable ($z$) is activated, and solid colored arrows indicate the activated synchronization variables ($y$), with a different color for each robot. A robot executing two consecutive fragments of the same task is modeled as a self relay.
• Figure 4: Full matrix pattern
• Figure 5: Corrected matrix pattern

Theorems & Definitions (2)

• Theorem 1
• proof