How AI is revolutionizing maintenance at LANXESS
If a car tire loses air, the on-board electronics in a modern car warn the driver immediately. Twenty years ago, it was usually only possible to react when the tire was flat.
Similarly, digital technologies have also changed production at LANXESS over the past two decades. A prime example of this is "Predictive Maintenance". State-of-the-art AI technologies are used here to increase plant availability and reduce costs.
AI complements the “trained ear”
In the past, only the trained ear of a long-serving employee could usually tell from the noises in the plant whether production was running smoothly or whether system failures were imminent. Today, the Production, Technology, Safety and Environment (PTSE) team in Leverkusen uses sophisticated algorithms and artificial intelligence to monitor the condition of critical machines and systems in LANXESS plants around the world.
"Our aim is to intervene in good time before problems can even arise and possibly cause consequential damage."
With the help of AI and sensors integrated into the systems, it is possible to predict with a very high degree of probability when and why a compressor, agitator or other system component could fail. The highlight: the AI not only recognizes gross anomalies such as the vibration of a pump. It also analyzes complex vibration and temperature profiles in their interaction with other process variables and can thus determine whether machines or systems deviate from their normal behavior and require intervention. These precise predictions help to identify potential failures hours, days and sometimes even weeks in advance.
Interplay of experience and technology
The AI models behind the predictive maintenance system are fed with both technical expertise and information on machines and processes.
"The specialists who program these AI models need the expertise of our experienced colleagues who know the machines and their typical behavioral patterns. Without this knowledge, we would not be able to make precise predictions."
LANXESS has therefore set up a team in Leverkusen, Germany, that works closely with experienced production employees, IT specialists, experts in process analysis technology and maintenance experts. In this way, the know-how of long-serving employees is combined with the power of data analysis and AI.
The future of maintenance
The results of the first few months are impressive: since the start of the global rollout of predictive maintenance activities, the team has already achieved savings in the seven-figure range. Although the implementation on other systems and machines will still take months, the investment and efforts have already paid off. But what's next?
rbtc"The aim is not only to use AI support to monitor individual machines, but also to optimize entire production processes. This is a key part of our Operation Excellence Streams. By using predictive maintenance and intelligent online measurement technology, we can manufacture the best possible products while minimizing the consumption of resources."
Stay up-to-date
Curious now? Subscribe to our Inside LANXESS newsletter and stay up-to-date.
