Testing of AI-based systems such as autonomous vehicles is challenging due to many situations and scenarios. Brute force is expensive and has gaps, as we see in practice. We thus use synthetic data for an AI-driven testing. This data covers real-world scenarios to train autonomous systems in a simulation-based environment. The training success is evaluated in a data loop and enhanced to close blind spots and unknown knowns. This thesis targets to integrate a requirements and test engine to an automated test system.
The goal of the thesis is to integrate existing parts of the system. A fully running system shall be implemented. The integration comprises verification and validation checks for the existing parts. Professional tools such as DOORS shall be used for industry-scale AI-based testing of autonomous systems.
Knowledge in Python
Industry-scale software engineering and tools
Work in a self-independent way
Passionate about clean and good quality code
Capable of integrating your work with other parts of the system
Christof Ebert
♦
In high-speed deep drawing (over 400 strokes per minute), undesired waviness known as earing forms at the part rim. Within a research project, this shape deviation is to be captured optically in the production line and compensated by data-driven models. Before any metrology hardware is integrated into the series tool, the complete measurement chain is to be designed virtually: from physically based light transport simulation of the production scene through sensor models to reconstruction of the rim contour. The thesis is part of a DFG transfer project at the intersection of digital twins, computer vision, and machine learning.
Based on a CAD scene , a virtual measurement chain for optical inline acquisition of the rim contour is to be developed. This includes transferring the scene into a physically based rendering environment, modelling the candidate sensor concepts (laser light-section, camera system) as parameterised sensor models, segmenting and reconstructing the contour from synthetic sensor data, and quantifying the reconstruction error against the known reference geometry. Building on this, two measurement positions (entry into the transport system, free-falling part) are to be compared systematically regarding achievable contour coverage, orientation reference, and resolution, and a recommendation for the sensor specification is to be derived. Optionally, the influence of disturbance effects (contamination, vibration, illumination drift) can be investigated.
Solid Python skills; fundamentals of image processing or computer vision; interest in simulation and metrology. Experience with rendering tools (e.g., Blender/BlenderProc, Mitsuba) or machine learning is helpful but not required. Independent, structured way of working.
Sebastian Baum
♦