Asynchronous motors are built for demanding industrial environments. But even proven technology can fail, and when a motor sits at the heart of a critical process, an unexpected fault can have consequences across an entire operation. 

The challenge is identifying the warning signs early enough to act. 

At CWIEME Berlin 2026, Prof. Stjepan Tvorić, Head of the Electrical Engineering Department at Zagreb University of Applied Sciences, explored how predictive monitoring can turn motor data into earlier fault detection and more informed maintenance decisions. 

Moving maintenance closer to the motor 

Predictive monitoring changes the way engineers approach motor health. 

Instead of relying solely on scheduled maintenance or responding after a failure, monitoring systems continuously collect information about operating conditions. Changes in vibration, temperature and electrical quantities can then provide indications that a motor is moving away from its expected behaviour. 

This is particularly relevant for bearings and stator windings. The session identified bearings as a major source of motor faults, highlighting why mechanical and electrical parameters need to be considered together rather than in isolation.

The potential benefit is straightforward: detect developing problems sooner, plan interventions more effectively and reduce the risk of unplanned downtime. 

Building a clearer picture from real time data 

Turning that principle into a practical industrial system requires more than installing sensors. 

The research presented at CWIEME Berlin uses a dedicated test bench to recreate motor operating conditions and simulate faults in a controlled environment. Sensors collect electrical measurements including voltage, current and power alongside mechanical vibration and shaft displacement. 

Beckhoff TwinCAT software is then used to visualise the information in real time. The platform incorporates early warning functions, stores historical data and supports baseline anomaly detection. FFT analysis can also be applied to collected data for diagnostic purposes. 

This combination is important because a single measurement rarely provides the complete answer. 

By establishing a baseline for healthy motor operation and monitoring how different parameters change over time, engineers can build a stronger picture of developing faults and determine when further investigation is required. 

Industrial monitoring needs to be practical 

More data does not automatically result in better maintenance. 

For predictive monitoring to deliver value at industrial scale, diagnostic systems need to be practical enough to deploy across large motor populations. The session highlighted limitations associated with some existing techniques, including air gap monitoring, and the need for simpler approaches that can scale across industrial applications. 

There is also an important distinction between detecting an anomaly and understanding its significance. 

Thresholds need to reflect real operating conditions. Historical information needs to be useful rather than simply accumulated. Monitoring also needs to provide engineers with information they can act on. 

This makes the development of reliable reference baselines and fault simulation particularly valuable. Controlled testing allows researchers to understand how specific faults appear within sensor data before those methods are applied in live industrial environments. 

Connecting research with industrial requirements 

The work also demonstrates the value of closer collaboration between academia and industry. 

Zagreb University of Applied Sciences is using its diagnostic platform not only to develop monitoring techniques, but also to give engineering students practical experience with sensors, data acquisition and fault analysis. 

That connection between research, skills and industrial application reflects a wider priority for the motor sector. 

CWIEME Berlin provides a platform for these communities to exchange knowledge across motor design, manufacturing and digital technologies. As monitoring capabilities develop, those connections will become increasingly important. 

Predictive maintenance ultimately depends on turning information into action. For asynchronous motors, better use of real time data could give engineers the visibility they need to identify faults earlier, plan maintenance with greater confidence and keep critical equipment operating for longer.