Simo Käki

Working machinery in agriculture, forestry, mining, and construction operates in harsh, often remote conditions. Unexpected failures cause costly downtime and safety risks, and on-site maintenance is difficult. A single unplanned stoppage can cost thousands of euros per hour. Despite widespread use of planned maintenance, these failures still occur. There is a need for control methods that actively reduce failures during operation, not just better schedules.

In industry, fixed schedule and corrective maintenance remain common because they are simple to plan and contract, yet they do little to reduce unexpected failures. Condition-based maintenance is utilized but often stops at alarms or fault codes with limited feedback to the control system. In research, condition monitoring and fault diagnosis are widely studied, and predictive maintenance with RUL estimation is mature. However, life-extending control is mostly applied in battery systems, and there is a clear gap in linking health information of components into the control layer. This gap is even more pronounced for hydraulic components such as gear pumps in electro-hydraulic systems.

My research asks how condition monitoring can be integrated with control in electro-hydraulic systems so that the controller actively extends the remaining useful life of components during real-time operation. It also evaluates when it is more practical to design new health-aware controllers versus adding adaptation layers around existing industrial controllers.

The approach starts from practical degradation pattern identification of gear pumps. Experimental data are used to build and validate degradation models, which are then embedded in simulations of electro-hydraulic systems to represent aging pumps under realistic duty cycles. On this basis, two types of health-aware controllers are investigated. First, controllers are designed from scratch with explicit life-extending objectives. Second, a supervisory layer is developed around existing controllers already used in industry to achieve life-extending behavior with minimal changes to current practices. Extensive simulation studies approximate the feasibility and impact of these strategies, and practical studies on working machinery are conducted to validate the results in real operating conditions.

Expected benefits are longer component life, reduced unplanned downtime, and extended, better-scheduled maintenance intervals. These outcomes lower total cost of ownership, improve availability, and help operators move from reactive repair towards proactive, life-extending operation of electro-hydraulic machinery.