Data-Driven Decision Support for Maintenance Prioritisation - Connecting Maintenance to Productivity

Doctoral thesis, 2018

Manufacturing companies are constantly seeking ways to be globally competitive. They are adopting smart and digital methods to compete, i.e. digitalisation. Digitalised manufacturing is seen as the next era in manufacturing that will ramp up the pace at which companies adopt digital technologies. In this era, production systems are expected to drastically improve their productivity, increase the level of automation, and achieve high resource efficiency. In order to fulfil these aims, managing maintenance becomes strategically important to manufacturing companies. However, maintenance management in manufacturing companies is generally fairly poor. Traditional maintenance approach tends to be reactive and focus solely on maximising the availability of machines. Especially, maintenance decision-making practices are reported to be made subjectively. Several researchers have reported low levels of resource efficiency, economic losses and productivity losses caused by poor maintenance. Therefore, there is a need for maintenance organisation to transform themselves in order to contribute towards increased productivity and cost-effectiveness.

The transformation should enable maintenance organisations to move from focusing on a component-level to achieving a systems perspective for solving maintenance problems. This is needed mainly in maintenance decision making in order to make fact-based decisions. This thesis contributes towards this transformation through studying the maintenance prioritisation decisions and their decision support systems. The research was conducted with the help of five empirical research studies using theoretical inputs, survey, interviews, experiments and case studies. The thesis contributes to,

(i) Identifying gaps between maintenance prioritisation practice and research.
(ii) Assessing real-time data for decision support and assessing maintenance prioritisation decisions for effectiveness.  
(iii) Developing a data-driven machine criticality assessment framework to guide prioritisation decisions. 

Furthermore, the framework also provides guidelines on how to approach the data needed, the analysis required and the type of decisions that can be made. An example from one of the studies showed a five percent increase in productivity when maintenance was prioritised based on machine criticality. The results obtained in the thesis will help maintenance organisations to transform from having a narrow focus (solving machine-level maintenance problems) to achieve a factory-focus (systems perspective), thus supporting an increase in productivity. By connecting maintenance to productivity, maintenance organisations can help manufacturing companies compete in global production.