Mälarenergi uses AI for its machine monitoring

At Mälarenergi in Västerås, there is a constant development of products, services, and energy and infrastructure solutions. Within the group and its subsidiaries, approximately 700 employees work towards the common goal of contributing to a resource-efficient society. There are many examples of how Mälarenergi has been at the forefront of introducing both new services and new technology . In 1992, for example, the group was the first in Sweden to offer district cooling to properties.

Mälarenergi is also at the forefront of adopting new technology that facilitates monitoring and maintenance in a machine-intensive business. One example of this is how in 2023 MLT was commissioned to equip more than a hundred machines with the pureMEMS wireless vibration sensor from MLT’s sister company pureSignal of Sweden AB.

In such a large and socially critical facility, it is important to keep track of the machines.

More than 350 sensors have been installed, which has meant a big change for Mälarenergi’s vibration analysts. pureMEMS in combination with cloud services, AI and a user-friendly app has made an otherwise time-consuming task easier. Historically, the condition of the machines has been checked during manual rounds. Checking a large number of machines manually takes time, and it is only at the time of the measurement that you are notified of problems and deviations. Not having continuous monitoring always means a risk of disruptions and, in the worst case, breakdowns in the periods between measurements.

AI early detected a deviation that could have led to a costly breakdown

During the first year of operation, Mälarenergi has been able to see that pureMEMS and AI add great value. Through analyses of time signals in the MultiViz AI tool, developed by the Gothenburg company Viking Analytics AB, a deviation on a machine with an almost new bearing was detected at an early stage. The early detection most likely prevented a very expensive breakdown. The ability to detect damage and deviations in good time with the help of AI means that pureMEMS provides a new dimension to the concept of proactive maintenance.

Robert Hjort, who is a group manager at Mälarenergi, describes a big difference compared to how it was before with manual measurements. Robert’s group has just over fifteen employees, where four of the maintenance technicians are responsible for vibration analysis.

pureMEMS sends an email when a machine shows abnormal behavior and AI provides a recommendation on what might be wrong.

– The manual vibration measurements that we previously did once a week gave “actual values” at the time of the measurement. We did not know what the status was between the measurements, and here the wireless sensors and continuous monitoring have meant a huge change, says Robert.

– Each machine has several sensors installed and with several hundred measurement points, there are an incredible number of measurement values ​​that are captured and analyzed using AI. My vibration analysts have always been good at capturing errors, but with the help of AI we can do it even earlier. AI technology can also add up trends in a completely different way than is possible manually, he adds.

Previous vibration measurement routines is turned upside down

Tommy Persson is a maintenance engineer in his group and has worked with vibration measurements and analyses since 1998. pureMEMS wireless technology and AI as a tool have brought about such a big change in daily work that Tommy believes it has turned previous routines upside down.

– I can only say that it works better than I thought. Quite early after the installation, we felt compelled to double-check whether the AI ​​analyses and the values ​​presented were correct. They did, and AI has now become a very important tool for us. One of the biggest advantages is that the technology picks out alarms in very good time. The fact that AI continuously works on, analyses, and sorts different events to present patterns and trends, means that we save a lot of time.

AI is a self-learning technology that gives Tommy and his colleagues a clearer picture of incoming alarms without being overwhelmed by information. pureMEMS sends an email when a machine shows abnormal behavior and AI gives a recommendation on what might be wrong.

– It is fascinating to see how the AI ​​technology learns and adjusts. In the beginning, there was a lot of information presented. Gradually, the amount of information has been scaled down slightly. I usually start the day by looking through the information presented in the easy-to-use app together with a colleague. Usually, there are quite a few alarms that the AI ​​has analyzed from all the measurement data from several hundred vibration sensors. Another big advantage is that it is easy to scroll down to different measurement points when we want to check something a little more carefully.

Good dialogue is important

Mälarnergis’ employees have a high level of measurement expertise and many years of experience in vibration analysis. They have therefore chosen not to use the monitoring service offered by MLT, but Robert and his technicians still believe that continuous dialogue is important.

– We have received good support from MLT with regular meetings on a weekly basis where we can discuss and get help if there is something we are unsure about in the trend and deviation information. That support is important to us, says Robert.

Rapidly growing interest in pureSignal’s wireless concept pureFAMILY

The benefits of remotely monitored condition monitoring, AI analyses and very high quality of measurement data have become clear to Mälarenergi. Another energy company that has recognized the benefits of pureMEMS is Jönköpings Energi, where MLT also helped install the technology. Anton Groth at MLT is convinced that the benefits of wireless vibration measurement with pureMEMS will soon spread to other industrial sectors.

– There are many other heavy industries where manual condition monitoring of production equipment is both time-consuming for maintenance organizations and an uncertainty factor because there is no continuous monitoring. We see how the demands for more proactive maintenance are increasing as the degree of automation increases in many industries. This places new demands on aids that make it easier to detect errors and the risk of production disruptions and breakdowns early. The fact that pureMEMS is easy to install without requiring any integration with existing IT systems makes it easy to install, without requiring integration with existing IT systems, which makes it possible to quickly take the step from cumbersome manual routines to smart AI monitoring.

pureSignal – the company behind the wireless sensors

MLT’s sister company pureSignal of Sweden has developed and launched three products under the pureFAMILY brand since its inception in 2022. In addition to pureMEMS and the pureALUBE lubrication system, the purePULSE speed sensor is also a family member. Christian Sahlin is sales manager at pureSignal of Sweden and sees a large future market both in the Nordic region and internationally.

– We have built up a very strong concept in a short time and it is very gratifying to see the interest that the wireless technology is attracting. Mälarenergi has talked about the advantages of pureMEMS and it is exciting that they are also conducting tests with pureALUBE. When we show the possibilities of linking vibration measurement, lubrication and even speed control in a complete solution, many people realize the great advantages. The fact that we have integrated AI technology that handles the analysis work without human intervention also contributes to the strong and sought-after offering, concludes Christian.

Visit Mälarenergi for more info about them: https://www.malarenergi.se/