In the Autonomous Performance Management in Digital Manufacturing (AURORA) project, the dualistic use of entropy for coordinating a multi-robot system was studied. To maintain a high system entropy, a coordinator will assign tasks so as to maintain an even distribution of agent capabilities in the remaining fleet. If this yields two or more candidates, the agent is selected which minimizes the task entropy, corresponding to performing the task more accurately.
The video below shows a real and simulated Autonomous Mobile Robots (AMRs) working in a shop floor machine tending scenario to demonstrate these entropy-based task allocation strategies. Three types of machines (blue, magenta, black) are present, while the AMRs have different machine tending capabilities.