Sidestepping verification complexity with supervisory control

While the goal of verification is to check whether a model of the system under consideration has a desired property, supervisory control achieves correctness by adding a so-called supervisor that prevents the occurrence of incorrect behaviors to the original system. Supervisory control methods are appealing because they can be much more tractable than the corresponding verification problems. Here we first examine two supervisory control algorithms, one for enforcing mutual exclusion properties and the other for enforcing real-time deadlines on Petri net models of the controlled system. Next, we argue that use of supervisory control methods may lead to a simpler and more effective coding style for embedded software than current practices. Finally, we highlight research issues that must be addressed in order to permit widespread application of supervisory control methods.