Data-driven integrated production scheduling and predictive maintenance for batch processes

Project ID: 2228cd1279 (You will need this ID for your application)

Under Offer

Research Theme: Manufacturing The Future

UCL Lead department: Chemical Engineering

Department Website

Lead Supervisor: David Bogle

Project Summary:

Poor maintenance planning is known to affect operational efficiency and safety of chemical processes, due to disruptions in scheduling due to unplanned downtime and deterioration of end-product qualities among others. At the same time data-driven decision-making has received considerable attention from the process systems engineering domain, mostly focusing on surrogate modelling and the development of dynamic digital twins. Contemporary processes have access to a diverse spectrum of data that can be leveraged to provide added value to their business units by enhancing operational decisions. Predictive maintenance is popular set of methods that employ machine learning techniques in conjunction with real-time sensor data and provide estimates on when maintenance may be needed. Nevertheless, systematic investigations on the integration of production planning, scheduling and predictive maintenance remain scare, if any, and this is the motivation behind this project.

The scope of the project focuses on the development of a systematic framework for real-time integration of tactical planning, scheduling, and predictive maintenance. Following previous findings from my group, when considering simultaneously different layers of decision-making we can exploit underlying interdependencies and enhance operations (Charitopoulos et. al, 2019). Thus, through this project a systematic way to simultaneously optimise production and maintenance with the goal to reduce downtime and improve efficiency is proposed.