A Multi-Robot Task Assignment Framework for Search and Rescue with Heterogeneous Teams

Hamid Osooli,Paul Robinette,Kshitij Jerath,S. Reza Ahmadzadeh

2023-09-22

Abstract:In post-disaster scenarios, efficient search and rescue operations involve collaborative efforts between robots and humans. Existing planning approaches focus on specific aspects but overlook crucial elements like information gathering, task assignment, and planning. Furthermore, previous methods considering robot capabilities and victim requirements suffer from time complexity due to repetitive planning steps. To overcome these challenges, we introduce a comprehensive framework__the Multi-Stage Multi-Robot Task Assignment. This framework integrates scouting, task assignment, and path-planning stages, optimizing task allocation based on robot capabilities, victim requirements, and past robot performance. Our iterative approach ensures objective fulfillment within problem constraints. Evaluation across four maps, comparing with a state-of-the-art baseline, demonstrates our algorithm's superiority with a remarkable 97 percent performance increase. Our code is open-sourced to enable result replication.

Robotics

What problem does this paper attempt to address?

This paper attempts to solve the problem of how to conduct search and rescue operations efficiently in post - disaster environments. Specifically, it focuses on the task assignment problem of multi - robot systems (including heterogeneous teams). Existing planning methods usually only focus on certain specific aspects and ignore key elements such as information collection, task assignment, and path planning. In addition, considering the capabilities of robots and the needs of victims, existing methods have a high time complexity due to repeated planning steps.

To solve these problems, the author proposes a comprehensive framework - Multi - Stage Multi - Robot Task Assignment (MSMRTA). This framework integrates the three stages of reconnaissance, task assignment, and path planning, and optimizes task assignment based on the capabilities of robots, the needs of victims, and the past performance of robots. The iterative method is used to ensure that the goals are achieved within the problem constraints.

### Main contributions:
1. **Comprehensive framework**: Introduced a multi - stage framework covering reconnaissance, task assignment, and path planning.
2. **Optimized assignment**: Optimized task assignment according to the capabilities of robots and the needs of victims.
3. **Efficient planning**: Reduced the time complexity of planning and improved overall efficiency.
4. **Empirical evaluation**: Verified the effectiveness of the algorithm through experiments on four different maps. Compared with the baseline algorithm (MRGA), MSMRTA shows a significant performance improvement (97% performance improvement).

### Specific problem descriptions:
- **Task assignment problem**: In search and rescue tasks in buildings, how to assign tasks to heterogeneous robot teams to meet the diverse needs of victims.
- **Information collection**: Obtain victim information (location, needs, etc.) in the environment through the reconnaissance stage.
- **Task assignment optimization**: Optimize task assignment based on the capabilities of robots and the needs of victims.
- **Path planning**: Use the A* algorithm for path planning to ensure that robots can reach the designated locations safely and effectively.

Through these improvements, this framework aims to improve the cooperation efficiency of multi - robot systems, so as to perform search and rescue tasks more effectively.