Bachelor/Master Thesis: Al-Based Image Recognition tor Determining Local Row Velocity in a Rolling Bearing

The Institute for Machine Elements and System Development investigates the fundamental structural and tribological behavior of machine elements and represents this through experimentally validated model descriptions. These models are used to analyze and design the functional, loss, and noise behavior of complete technical systems, with a focus on drive technology. The developed models also serve as a basis for research and development of methods in Model-Based Systems Engineering as a key element of future industrial product development processes.

The analysis of oil flow in rolling bearings is crucial for optimizing lubrication, friction losses, and the service life of mechanical systems. Classical numerical methods such as CFO are precise but time-consuming in terms of modeling and computation. The aim of this project is to develop an Al-based approach to determine local flow velocities directly from video data. A high-speed camera will be used to record the oil flow in a rotating rolling bearing. Using modern image processing and artificial intelligence, distinctive points (e.g., oil particles or flow features) will be automatically detected and tracked. The movement of these points will be used to derive flow velocities. The results will be validated against our CFO simulations and, if available, experimental measurements.

Tasks:

• Literature review on deep learning and feature tracking in fluid flows
• Execution and analysis of high-speed recordings of oil flow
• Calculation of local velocities from image data
• Validation of results using existing CFD and experimental data

Requirement:

• Interest in Al, fluid mechanics, and image processing
• Experience with Python and deep learning frameworks is a plus
• Independent and structured work style

We offer:

• Flexible design of work focus
• Quick processing possibilities
• Intensive support
• Immediate start or by arrangement
• Very good working atmosphere

We look forward to your application by email:

Amirreza Niazmehr, M. Sc. RWTH
Institut für Maschinenelemente und Systementwicklung

Schinkelstraße 10, 52062 Aachen
amirreza.niazmehr@imse.rwth-aachen.de