Bachelor and Master Theses

To apply for conducting this thesis, please contact the thesis supervisor(s).
Title: Evaluation of unimodal and multi-modal Gen AI for root cause analysis of product defects
Subject: Computer science, Embedded systems, Robotics, Software engineering, Computer network engineering
Level: Basic, Advanced
Description:

Finding the root cause of product defects can consume considerable effort of software and system engineers, where information from multiple sources must be reconciled or consolidated to decide on the best path forward that provides a solution to the identified defect. This thesis will investigate the application of Gen AI models on this problem domain where documentation from several sources will be made available, such as product logs, product documentation, historical defects, and potentially, source code.  The thesis will investigate the application of Gen AI models on this corpus of data and evaluate the importance of one vs. several sources of information to reach an accurate root cause analysis.
Start date: 2026-01-01
End date: 2026-07-01
Prerequisites:

Good programming skills, good understanding of industrial software engineering principles, good understanding of software engineering tools such as configuration management tools, source code management tools, and document management tools, good understanding of current GenAI tools and how to configure them
IDT supervisors: Wasif Afzal
Examiner: Eduard Paul Enoiu
Comments:

This thesis is a MDU thesis, but documentation will be provided by our industrial partner, Alstom.

Company contact:

This thesis is a MDU thesis, but documentation will be provided by our industrial partner, Alstom.