Modular reconfigurable robots can deliberately assemble into diverse configurations to adapt to complex environments, making accurate configuration identification essential for automation. This paper aims to develop and experimentally validate a distributed configuration identification framework for the Reconfigurable Flight Array (RFA), a modular reconfigurable aerial robot composed of identical flight unit modules.
To achieve distributed configuration identification of the RFA, this study integrates hardware design, information routing and distributed consensus-based computation. Each flight unit module is equipped with a dedicated communication circuit that enables local data exchange with directly connected neighbors. An information routing strategy is developed to allow each module to obtain the relative coordinates of its adjacent modules within its own coordinate frame. And a consensus protocol is implemented to enabling all modules to collaboratively reconstruct the complete RFA configuration with local information.
Experimental results verify that the proposed framework enables each RFA module to autonomously and accurately determine its relative position within the overall structure. Flight tests further demonstrate that RFAs with correctly identified configurations achieve significantly improved tracking performance.
The proposed method enables fully decentralized configuration identification, greatly enhancing system scalability and autonomy. The demonstrated improvements highlight the method’s practical applicability and its potential to advance the intelligent autonomy and cooperative capabilities of modular robot systems.
