Guest Editorial Introduction to the Special Issue on Swarm Robotics

The field of robotics has expanded tremendously over last several decades. Researchers' attention has shifted from its early focus on industrial manipulators and individual mobile robot toward today's more challenging topics such as multi-robot systems, mobile sensor networks, and swarm robotics. Swarm Robotics is an emerging area which studies novel approaches to coordinate a large number of relatively simple robots to achieve desired collective behaviors and objectives that would not have been possible for individual robots. Swarm robotics emphasizes scalability, local interaction among agents, and fault tolerance. Potential applications for swarm robotics include tasks that demand for distributed sensing such as odor source localization, surveillance coverage, search and rescue. This special issue presents a collection of 7 papers which investigate different research problems in swarm robotics such as coverage, deployment, distributed control, self-localization, odor source localization, diffusion, and interactive animation. I believe that this special issue, although not cover all the topics in swarm robotics, provides the most recent developments in these selected topics and will be an important source of information for researchers. The paper by Diana Spears et al. presents a physics-based approach inspired by particle motion in a gas for controlling robot swarms to maximize coverage. The algorithm is based on kinetic theory (KT) which provides analysis tools to control, predict and guarantee the swarm behavior within desirable bounds. The approach requires limited sensors and communication between robots and is exceptionally robust. Extensive simulation and comparison studies against the three alternative algorithms designed for coverage tasks (i.e., random controller, trained finite-state machine controller, and Ant algorithm) demonstrate the superiority of the KT approach in terms of sweep time and stealth. It also achieves the second best performance in spatial coverage next to the Ant algorithm. Since KT approach does not have the explicit pheromone trail requirements of Ant algorithm, it is better-suited than Ant algorithm for situations where stealth, and speed of coverage are required. The paper by Carlo Branca and Rafael Fierro presents hierarchical optimization strategies for deployment of mobile robots in task assignment problem. The algorithm is based on model predictive control (MPC) and mixed integer linear programming (MILP) techniques. MILP allows the encoding of logical rules, decisions, and constraints into the optimization problem, and has the capability to model requirements such as obstacle avoidance or inter-robot collision avoidance. To overcome the NP-hard drawback of MILP and achieve on-line processing of the optimization problem, the …