Hormones for self-reconfigurable robots

Self-reconfigurable, or metamorphic, robots can change their individual and collective shape and size to meet operational demands. Since these robots are constructed from a set of autonomous and connectable modules (or agents), control of the robots and coordination among the modules are highly complex and challenging tasks. The difficulties stem from the fact that all locomotion, perception, and decision making must be distributed among a network of modules. This network has a dynamic topology, and each individual module has only limited resources in terms of computational power and local information about the topology in its neighborhood. To meet these challenges, this paper presents a distributed control mechanism inspired by the concept of hormones in biological systems. We view hormones as broadcast messages that trigger different actions in different modules, and exploit such to coordinate motions and perform reconfiguration in the context of limited communications and dynamic network topologies. The paper develops a primitive theory of hormone-based control, and reports the experimental results of applying such a control mechanism to our CONRO metamorphic robots, along with the results of simulations.