A class for data science professionals and managers working with maintenance personnel.
The goal is to enable data scientists working on classification, prediction and information extraction to understand 1) how the algorithms and models they develop support decision making and 2) the context generating the data used, and 3) how decisions in maintenance are made.
Maintenance of assets is a significant cost input for the manufacturing, resources, defence and infrastructure sectors. Maintenance costs typically range between 20–60% of operational expenditure depending on industry and asset type. There has been significant effort in the last decade to move from reactive to preventative and predictive maintenance strategies propelled by developments in sensing, WiFi, cloud computing, and data analytics. However, generating value from analytics using these platforms has often proved challenging for a number of reasons (see list of useful references below). However one unexplored reason is that what happens in maintenance and the way maintenance management is structured and run inside organisations is not clear to non-maintenance people. This leads to poor decisions being made on what problems to focus on, with many apparently 'killer apps' being impossible to implement or sustain in the real world.
This 8 hour course was developed by Professor Melinda Hodkiewicz as part of a sabattical visit to Professor Hedi Karray at ENIT in Tarbes France in 2023. The material was delivered as part of unit NDENI-EC0112M4 - Innovation management.
- Identify steps in the maintenance work management process.
- Describe key elements of maintenance strategy development and assessment process.
- Understand where and how data, commonly used for maintenance models, is generated.
- Discuss when, and why, predictive maintenance is an appropriate strategy and when it is not.
- Understand opportunities for natural language and ontologies use on maintenance texts.
These materials are intended to assist data science professionals in gaining an understanding of core concepts in maintenance with the goal of making them aware of the business processes and engineering constraints within which their data pipeline(s) have to operate and their algorithms support.
- IoT-Based Prognostics and Systems Health Management for Industrial Applications[
- Where do we start? guidance for technology implementation in maintenance management for manufacturing
- 20 mins YouTube video on AI in maintenance - challenges to value delivery
- The future of maintenance for distributed fixed assets