%0 Thesis %A Jimenez Andrioli, Paulo Andres %D 2017 %T Enhanced design and experimental verification of a bio-inspired architecture for a team of cooperative robots %U https://bridges.monash.edu/articles/thesis/Enhanced_design_and_experimental_verification_of_a_bio-inspired_architecture_for_a_team_of_cooperative_robots/4657879 %R 10.4225/03/58a5294eef3c9 %K 1959.1/930509 %K thesis(doctorate) %K Autonomous robots %K monash:120779 %K Cooperative robots %K 2014 %K Restricted access %K Bio-Inspired architecture %K ethesis-20140428-143724 %K Area coverage problem %X This study proposes and establishes a novel multi-robot architecture based on concepts observed from cooperation in biological systems. The different components, the behaviours, and the communication methodologies of the proposed architecture are de-scribed and defined. Each robot regulates their responses to a different stimuli using a series of motivational states. Robots respond to requests once their motivation state is sufficiently high, which leads to a reduction in communication messages without significantly altering the outcome of the task. This allows team-mates to iteratively respond to each other's actions without directly influencing them. The position of a robot has always an error associated with the movement of the autonomous platform. These errors are measured from data derived from the experimentation and then calibrated with the novel laser interferometry based calibration methodology. Finally the proposed architecture is experimentally evaluated, thus testing the deliberative capabilities as well as the autonomous capabilities. The area coverage problem is chosen to test the architecture, and therefore four solutions to the problem were successfully established and experimentally tested. The Swarm Global Planner Methodology considerer the entire region when calculating the paths; the Divide and Conquer methodology divides the region into sub-regions, and subsequently distributes the workload among robots. The Biological Motivation Based methodology assigns tasks as they appear in the environment; the Controlled Distribution of Task methodology maintains a maximum number of robots working on each task. The experimental results showed the feasibility of the methodologies, the performance of the proposed architecture, and its versatility. %I Monash University