Formal models for cognitive systems

Abstract

A robotic system is an integration of various systems such as perception, navigation, planner, controller, etc. The adaptation of the robotic system to the dynamic environments is embedded in the functionality of the constituent systems. This severely limits the configuration space of such systems. The modeling of the variability in the solution space can expand this design space, help finding the best possible solution, and perform run-time adaptation of the system. In this paper, a formal specification using primitive compositional elements is proposed to analyze the solution space and to streamline the decision making using non-functional properties of the system. The relevancy of the model is demonstrated using a case study on a vehicle tracking problem.