Robust and Scalable Swarming of Quadruped Robots for Chemical, Biological, Radiological, and Nuclear Source Localisation

In this paper, a robust and scalable method for the localisation of Chemical, Biological, Radiological, and Nuclear (CBRN) sources is introduced using a swarm of quadruped robots. A cornerstone of this work is the development of an advanced simulation environment built on the Robot Operating System 2 (ROS2) and compatible with Gazebo. This environment features a sophisticated physics engine and supports a wide array of sensors, offering unprecedented scalability for a swarm of quadruped robot simulations. Following this, we present a novel swarming algorithm designed to exploit quadruped robots' unique kinematics. This algorithm is an enhanced extension of traditional Particle Swarm Optimisation (PSO) and incorporates features such as obstacle avoidance, separation policies, velocity clamping and boundary constraints to optimise the CBRN source localisation process. To rigorously assess the robustness and efficacy of the proposed method, extensive experiments are conducted in three different environmental setups. The results reveal a remarkable convergence and success rate in CBRN source localisation, thereby substantiating the effectiveness of the proposed method.