Counterexample-guided simulation framework for formal verification of flexible automation systems

This paper proposes a framework for formal verification of industrial automation software in an intuitive way. The IEC 61499 function block architecture is assumed to be the input language, and the Intelligent Mechatronic Components (IMC) architecture is assumed as an underlying design pattern for the applications, which implies autonomous control logic in each IMC and their compositions to systems in a plug-and-play way. Then the system is automatically verified using model checking and the counter examples for the failed model checking properties are played back step-by-step and state-by-state in the simulation model that most industrial automation control systems would have built as the basis for initial testing. Net Condition Event Systems formalism (a modular extension of Petri net) is used to model the decentralized control logic and discrete-state dynamics of the plant. The model is then subjected to model checking using the ViVe/SESA tool chain. The method's application is illustrated using a simple pick and place manipulator. A closed loop model of Plant and Controller is used. Controller is extensively verified for safety, liveliness and functional properties of the robot. We then show how a counter example for deadlock detected by the model checker is played back in the simulation model for visualizing how exactly the system deadlocked.