Artificial intelligence is revolutionizing the maintenance process of industrial trucks

After each maintenance appointment, our service technicians manually prepare a service report on the problems that have occurred on the vehicles. Since these reports are often written under high time pressure, they can contain many abbreviations or colloquial expressions, which can make them extremely difficult for others to read. In itself, this is not a problem. The service reports are analyzed to evaluate the frequency and number of specific faults or problems that occur with the vehicles. Problems that occur repeatedly are detected and assigned to the same so-called "master problem".

It takes a lot of time and effort not only to translate all service reports, but also to correct and categorize them. This is how the idea for the MERLIN project came about, because "data is the new fuel of companies," as Bert Hörhold, Manager Customer Quality Excellence and Standards Development, puts it. In collaboration with KION's Analytics & AI team, the decision was made to use AI to make such tasks faster and more efficient. This is made possible by recent advances in deep learning, a subfield of machine learning (ML) in which artificially generated neural networks learn to recognize patterns and relationships in very large amounts of data. The resulting ML models are then able to recognize images, understand entire texts, and support decision-making.

To store the data and train the ML model, the team uses the cloud resources of the KION Analytics Platform (KAP), which is operated by KION Group IT. In the first phase, the AI was trained with the help of a glossary to distinguish KION-specific abbreviations from spelling errors in the service reports and to either write them out completely or correct them and then translate the documentation into English. The AI also orients itself on the truck model, series or spare part mentioned to improve the translation and correction of the documentation. This provides a consistent data basis for further classification and analysis. Until now, the translations have been done by an external agency. This was not only a major cost factor that has now become superfluous, but also took much longer. Translations used to come once a week, but now they can be downloaded twice a day.

In a second step, the AI is trained to search for patterns and automatically classify the service reports into the so-called "master problems". The MERLIN classification is currently being uploaded as a proposal with a percentage confidence level for further manual review. "Currently, the final decision is still made manually. In the long term, however, the plan is to transfer decision-making to the machine," says Bert Hörhold. In the future, defects that occur in our vehicles can be classified much more efficiently. Overall, the tool can be used to achieve annual savings in the 6-digit range. In addition, the use of MERLIN also relieves the iQ teams (iQ = improved quality) by reducing the effort required for classification. This leaves more time to focus on the core business: solving customer problems and providing recommendations based on the analysis. This is a great example of how AI can free highly skilled employees from manual, repetitive tasks so they can focus on more important tasks.

With the data lake as the basis and self-service query engines such as Dremio, data from different systems and sources can now be easily consolidated and pulled. Bert Hörhold is enthusiastic about the possibilities: "Where previously different data sets were stored in isolation in different areas, we can now bring everything together in the data lake. We are creating a treasure trove of data for the company." Looking ahead, the data could support predictive maintenance and other value-adding use cases.

"MERLIN is a great example of how AI can make KION's processes more cost-efficient, faster and better," says Philipp Pfitzmann, Director Analytics & AI. "As the Analytics & AI team, our mission is to pragmatically translate the full spectrum of machine learning into value-adding use cases for KION ITS EMEA, and in Bert we have found a great partner with a great use case. The combination of generative AI and a custom-built classification model has shown excellent results and points to the enormous potential for further AI use cases in the company."