Adapt hierarchical component-based systems dynamically – ADAPT

Smart cyber-physical systems operate in uncertain environments and constantly adapt to changing internal or external conditions. In the specific context of programmable matter, systems are composed of a multitude of small similar modular robots, such as Blinky Blocks under development, which organize and collaborate to achieve a common global goal.

We aim at a formalism to support correct-by-construction development of such systems through 1) hierarchical modeling to manage organic complexity and 2) natural integration of adaptive control to guide reconfiguration and to optimize their operation depending on both cyber (e.g. system configurations, interactions) and physical (e.g. energy consumption, environmental noise) run-time conditions. Our hierarchical models with mixed discrete and continuous-time dynamics will be used to design control mechanisms and to validate them before deployment on existing modular robots through simulation in DR-BIP and VisibleSim.