STOBER continues to improve predictive maintenance

The drive specialist has defined several expansion stages, continuously monitoring the drive train and deriving maintenance recommendations based on the analysis of the relevant process and machine data. Possibilities range from analytical calculation models to AI-supported processes. STOBER is following an integrated solution strategy. This requires no external sensors or accessories and is accessible for a wide range of controllers.

"How can we use predictive maintenance to make the transition from pure condition monitoring to real condition-based servicing for drive trains?" Tim Lang, head of system and test at STOBER, and his team asked themselves. Predictability is becoming increasingly important for users who want to achieve higher availability, lower maintenance costs and longer life cycles. "How likely is it that the geared motor will fail soon?" or "When is the ideal time to service or replace it?" To answer these questions, STOBER is pursuing a three-stage development plan, the second stage of which is currently being implemented.

First stage: model-based analysis

In the first expansion stage, users received a predictive maintenance solution that monitors the geared motor of a drive system. Its life performance is calculated using an analytical model and then output in the drive controller software via a life performance indicator, which is a value between 0 and 100 percent. At a threshold value of 90 percent or above, the software recommends replacing the geared motor and also makes this information available to a controller in the form of a readable parameter. This efficient solution for predictive maintenance requires neither external sensors nor additional wiring.

Second stage: active measurement

"In the second stage, the calculation model is supplemented by active measurement," explains Lang. For this purpose, STOBER is integrating an acceleration sensor into its system consisting of a gearbox, motor, cable and drive controller. External power and voltage sources are not required. This approach allows targeted monitoring of bearing seats, toothing and other drive components.

"Frequency analyses enable us to draw conclusions about impending damage based on a spectrum," said Lang. "Right now, we are still in the prototype phase."

STOBER is working closely with DR. JOHANNES HEIDENHAIN GmbH to develop a geared motor with an integrated acceleration sensor.

The goal in sight: a smart drive train

"Among other things, we have improved the analytical model and expanded the database to over 80,000 combinations of gearboxes and motors," says Lang.

With the new LoadMatrixAnalyzer, which will be available in the coming months, customers will be able to compare load matrices even more easily, generate standardised reports and create individual evaluations on request. Users benefit from easier operation and improved visualisation. In the future, STOBER will also provide function blocks and sample programs with which the data obtained – in particular the load matrices – can also be read out via EtherCAT or PROFINET.

For the LoadMatrixAnalyzer, the load matrices form a solid database for recording real load situations. This opens up a wide range of applications for condition-based servicing, from the detection of design and assembly influences to meaningful long-term analyses.

Lang explains: "Users assign a project name for each analysis of the load matrices. This allows individual matrices to be clearly named, described and provided with all relevant information for the analysis report. The LoadMatrixAnalyzer itself offers several central functions. These include the representation of load cases. Here, torque and speed at the gearbox output are visualised over time in a 3D diagram, making load situations visible at a glance."

In addition, the standard limits for the motor, gearbox and drive controller can be displayed individually or in combination. This makes it immediately apparent whether certain load ranges are within the permissible value ranges or whether deviations occur that need to be examined more closely and evaluated individually. 
Finally, the results can be documented directly in the Analyzer. This determines whether the load is a non-critical momentary load or whether there are potential risks for individual components.

The LoadMatrixAnalyzer can also be used to compare up to four load matrices with one another. This makes it clear how the speed and torque have changed over a certain period of time. "This shows us, for example, whether a gearbox has already run in or whether there are signs of stress and defects," says Lang. The analyses can then be exported as a standardised report in PDF format.

"For me, the tool is more than just an auxiliary program – it's a stand-alone piece of software. It can be updated, and we are continuously developing it further," says Lang.

In the next steps, STOBER will expand its predictive maintenance solutions with AI support. The aim is a smart drive train that can both recognise its own condition and provide relevant field data in real time. 

www.stober.co.uk