Mandatory Internship Data Analysis & ML Model Research for Automotive Systems

This internship is the foundational first phase of an innovative project aimed at developing a machine learning-based "Load Profile Generator" for vehicle E/E powernet simulations. Your mission is to analyze our rich datasets of vehicle operational data and conduct the critical research needed to select the optimal ML architecture for this task.
The results of your internship will directly inform and enable a subsequent Master's thesis project focused on model implementation and integration.

• During your internship you will dive deep into large time-series datasets from vehicle measurements to understand underlying patterns, distributions, and characteristics of vehicle power consumption.
• Furthermore you will develop as well as implement robust scripts and workflows for cleaning, transforming, and preparing the raw data into a structured format suitable for ML model training.
• You will identify and create meaningful features from the time-series data that will enhance the performance of a future generative model.
• Gain experience in conducting a comprehensive literature review and comparative analysis of state-of-the-art machine learning models for synthetic time-series generation (e.g., Generative Adversarial Networks (GANs), Variational Autoencoders (VAEs), RNNs, Transformers).
• Lastly you will conclude your internship by preparing a detailed report and presentation that summarizes your data findings and provides a well-reasoned recommendation for the most promising ML model architecture and data strategy to be pursued in the next phase.