Smart predictive maintenance

A photo of a person holding a tablet with graphs up to an assembly robot.

We develop integrated predictive tools for operation and maintenance of machinery, fleet and processes. These tools guide operators to make corrective action proactively to prevent unnecessary shutdowns. 

Key facts

Operation and maintenance data-analytics services for decision support 

Identifying critical circumstances and notifying about the impending anomalies 

Combining tribological root interactions, knowledge and data-driven analytics and  hybrid modelling into predictive maintenance 

Global industrial business requires innovations that are economically more effective and environmentally less consuming. But how to increase the efficiency and availability, ensure high production quality, and reduce losses of industrial processes and machineries?  

VTT’s mission in predictive maintenance is to generate solutions that increase overall equipment effectiveness (OEE) by analysing data and producing a solid basis for smart decision-making. We extract reliable indicators and patterns from data to predict issues before they happen. This helps avoid unscheduled maintenance, which maximises productivity and production quality. 

With real-time understanding of processes, machineries, structures and materials our customers can optimise their products’ performance and overall lifetime. This is essential for supporting longer lifecycles and circular economy. 

Predictive analytics for decision-making

For smart predictive maintenance, VTT develops data analysis techniques and tools to support proactive decision-making. Predictive maintenance analytic tools help the operator to keep the processes running optimally. They promptly alert the operator when the systems are starting to deviate from the planned, which would lead either to non-optimal performance or to premature failures.  

With predictive maintenance you can optimise your machines and processes  through the life cycle. 

The analytic tools guide the operators to make corrective action to get the system back on the track, and to prevent unnecessary shutdowns. Predictive analytics also offers time to plan and schedule maintenance actions. In addition, collected information on how the systems are used allow identification of best practices and variances in operative use both at a single process level or fleet level. 

Digital services for operation and maintenance analytics

Our operation and maintenance analytics offer carefully developed tools combining predictive maintenance analytics and knowledge-driven performance analytics with data-driven modelling. To understand the root causes of failures, we integrate latest scientific expertise from a range of fields.  

We understand possible failure scenarios and the dynamic interactions of processes and mechanisms. With our sensing expertise we can create real-time targeted data and use big data/machine learning analytics to turn the data into better decision-making. This creates advanced condition monitoring, diagnostics and prognostics solutions in lifecycle management.  

Our methods and tools have proven valuable addressing the operation and maintenance (O&M) of our customers’ industrial process, machinery, structural and material challenges. They offer clear transition from reactive to proactive O&M.  

For example, an online computational model of pulp quality has facilitated material efficiency to be improved by up to 2-3%. At a mid-size pulp mill, this corresponds to annual savings of ca. 700 truckloads of wood, implying a considerable improvement in both competitiveness and sustainability. In this work, VTT was responsible for developing the core computational model. 

Digitalised predictive maintenance tools and services can be utilised independently in customer applications as stand-alone solutions. They can also be integrated into up-to-date digital industrial platforms and frameworks with plug and play, self-configuration properties. The actual computation can occur at the edge, gateway and cloud level, based on the customer need. 

Share