In this project, which started at the beginning of 2022, TITUS’ primary focus is on researching acoustic solutions for rail transport. These can be used, for example, to detect specific components or even damage. The aim is to investigate the feasibility of evaluating acoustic signals using AI.
In rail transport, a wide variety of sensor technologies are already used on the train itself or in the trackbed to collect data, such as error patterns, to ensure safe operation. Not only because previous approaches are partly susceptible to faults or maintenance-intensive, but also because it enables an extension of the data collection, it makes sense to supplement the existing systems with acoustic measuring systems along the rail.
Within the framework of the project, the requirements for the necessary hardware and software components are to be derived through a comprehensive analysis of the application scenarios and transferred into an implementation concept. This includes not only the measurement setup and implementation but also the architecture of the AI module. After recording and processing the data, the AI model is trained, tested and then evaluated and optimised.