End-to-end process monitoring: Challenges and framework for case study design

Abstract

The theory and practice of process monitoring are diverging. Practical process monitoring requires fault detection, identification, diagnosis, and the implementation of process recovery actions. Algorithmic design and automation of all these steps are required if the future promise of more autonomous plants is to be realised. In contrast, theoretical research in data-driven process monitoring is overwhelmingly focused on fault detection. This paper discusses the current challenges of data-driven process monitoring research and presents a conceptual framework for improved experimentation of end-to-end process monitoring approaches. An end-to-end process monitoring solution is defined as the complete set of automated algorithms that is able to execute (in real-time) fault detection, fault identification, fault diagnosis, as well as process recovery intervention advisories. The major contribution of this framework is in the increased relevance added to the process monitoring problem to be solved, particularly in the extent of autonomy that is required by any proposed process monitoring solution.