Compositional Resource Analysis of Parametric Box-based Autonomous Vehicle Systems

Orchestrating a flock of autonomous vehicles so that the flock as a whole can guarantee to perform its mission, while at the same time reacting to changing circumstances or mission requirements requires good understanding of its behaviour, especially the execution time and energy usage implications for each possible action that the flock could take. This paper describes a new analysis approach that determines the overall resource usage of an autonomous vehicle in terms of resource usage of its hardware and software components. Our analysis expresses resource usage information in terms of a number of control parameters. These parameters can be used by reconfiguration algorithms (such as those simulated by Waterfall Solutions’ Cerberus system) to control the behaviour of each member of a flock of autonomous vehicles in resource-constrained situations.