Pervasive Diagnosis: Integration of Active Diagnosis into Production Plans

In model-based control, a planner uses a system description to create a plan that achieves production goals. The same model can be used by model-based diagnosis to indirectly infer the condition of components in a system from partially informative sensors. Existing work has demonstrated that diagnosis can be used to adapt the control of a system to changes in its components, however diagnosis must either make inferences from passive observations of production plans, or production must be halted to take specific diagnostic actions. In this paper, we observe that the declarative nature of model-based control allows the planner to achieve production goals in multiple ways. We show that this flexibility can be exploited by a novel paradigm we call pervasive diagnosis which produces diagnostic production plans that simultaneously achieve production goals while generating additional information about component conditions. We derive an efficient heuristic search for these diagnostic production plans and show through experiments on a model of an industrial digital printing press that the theoretical increase in information can be realized on practical real-time systems and used to obtain higher long-run productivity than a decoupled combination of planning and diagnosis.