Stochastic modeling, control, and verification of wild bodies

This paper presents strategies for controlling the distribution of large numbers of minimalist robots (ones containing no sensors or computers). The strategies are implemented by varying area, speed, gate length, or gate configuration in environments composed of regions connected by gates and modelled by Continuous Time Markov chains. We demonstrate the effectiveness and practical feasibility of our strategies through physical experiments and simulation. We use Continuous Stochastic Logic to verify high level properties of our system and to evaluate the accuracy of our model. Also, we prove that our model is accurate and that our algorithms are efficient with respect to the number of regions and number of bodies.