Industrial manufacturing companies worldwide are in a relentless pursuit of competitive advantages. One crucial area of focus is the optimization of manufacturing equipment, particularly milling machines, to improve quality, performance, and availability. In his doctoral dissertation, MSc (Tech) Teemu Mäkiaho developed a sophisticated physics-based simulation model that closely replicates the operational behaviour of peripheral milling machines.
“This model simulates various factors such as cutting forces, torque, and material removal rates, providing a robust foundation for diagnosing and predicting abnormal states of operation. Moreover, my research delves into data-driven methods to uncover significant correlations between machine excitations and wear phenomena,” says Mäkiaho.
A groundbreaking aspect of Mäkiaho’s doctoral research is the introduction of two novel hybrid methods: Fused Data Prediction Model (FDPM) and Simulation-Enhanced Anomaly Diagnostics (SEAD). FDPM combines simulation data with real sensor data to predict the remaining useful life (RUL) of milling machines, while SEAD generates high-quality synthetic data to improve anomaly detection using deep neural networks. According to Mäkiaho, these methods have shown significant potential in increasing system availability and reducing unplanned downtimes, offering a strategic advantage to industrial manufacturers.
Mäkiaho’s research is a profound contribution to the field of industrial manufacturing, providing practical tools and strategies for enhancing asset availability.
"The findings are particularly relevant in the context of rising material costs, as well as the ongoing drive for more efficient resource utilization. Reliable and robust hybrid models designed for industrial products create a foundation for interlinking multiple product availability prediction, fostering a more holistic view of predictive systems. When integrated with services related to hybrid product ecosystems, my approach can leverage richer data to deliver more customized outcomes for customers," he states.
By integrating advanced simulation models with cutting-edge machine learning techniques, Mäkiaho’s dissertation paves the way for smarter, more efficient maintenance practices in the manufacturing industry. His work not only addresses current challenges but also sets the stage for future innovations in life cycle engineering strategies.
Teemu Mäkiaho conducted his doctoral research at the Mechatronics Research Group (MRG) at Tampere University. Today, he offers his expertise in the field of industrial life cycle engineering through his company Makiatech Oy, contributing to both academic and industrial communities.
Public defence on Friday 13 September
The doctoral dissertation of M.Sc. (Tech) Teemu Mäkiaho in the field of engineering sciences titled Proposals for Availability Prediction Methods for Peripheral Milling Machines will be publicly examined at the Faculty of Engineering and Natural Sciences at Tampere University at 12:00 on Friday 13 September 2024. The public defence is held at Hervanta campus in auditorium K1702 of Konetalo building (Korkeakoulunkatu 6, Tampere).
Opponents are Professor Petri Kuosmanen from Aalto University and Professor Heikki Handroos from LUT University, Finland. The Custos will be Professor Kari T. Koskinen from the faculty of Engineering and Natural Sciences at Tampere University.
The doctoral dissertation is available online.
The public defence can be followed via remote connection. (link to be added)