In 2020, there were 11 million electric vehicles worldwide, and their number is growing exponentially. With the increasing popularity of battery electric vehicles (BEV), sustainable repair of high-voltage batteries is becoming important. Especially because working with high-voltage batteries involves safety risks for service technicians. For example, damage or incorrect installation can lead to thermal instability and a risk of fire hazards.

The repair and maintenance processes for high-voltage batteries in electric commercial vehicles are complex, and the underlying technologies are relatively new. As a result, battery modules are often replaced rather than repaired, which is costly and unsustainable. In addition, expertise in this area is concentrated among a small number of specialists. To make this knowledge available to service technicians and provide assistance during the repair process, this poster presents an AI-supported, prototypical knowledge management system. This system was developed to collect, structure, and provide information for the repair and maintenance of electric commercial vehicles. For this purpose AI components for semantic data processing and natural language interaction are integrated. Using a large language model (LLM) in combination with text embedding and LangChain, structured information can be extracted from unstructured technical documents such as repair manuals and handbooks. This enables the knowledge management system to act as an assistant for service technicians in the field by analysing data, providing an initial assessment, or filtering out the important information. The resulting system aims to facilitate knowledge extraction and management and support sustainable and cost-effective maintenance of commercial vehicles.