Self-Reconfiguration in Response to Faults in Modular Aerial Systems

We present a self-reconfiguration technique by which a modular flying platform can mitigate the impact of rotor failures. In this technique, the system adapts its configuration in response to rotor failures to be able to continue its mission while efficiently utilizing resources. A mixed integer linear program determines an optimal module-to-position allocation in the structure based on rotor faults and desired trajectories. We further propose an efficient dynamic programming algorithm that minimizes the number of disassembly and reassembly steps needed for reconfiguration. Evaluation results show that our technique can substantially increase the robustness of the system while utilizing resources efficiently, and that it can scale well with the number of modules.

I. Introduction

Termites, bees, and ants are prominent examples of many small units working together to accomplish large undertakings (e.g., a termite mound). Recently, there has been a trend towards replicating such swarm-like behavior in robotic systems to solve complex collective tasks such as exploration [20], construction [13], and transportation [4] [8]. For example, robot swarms have been used to create large structures that are impossible for a single agent to accomplish [1], [6], [18].