Join our expert team of engineers and researchers to work on the future of automotive user interfaces. You will take on responsibility in an exciting project from day 1 and be able to contribute your ideas while getting to know the Porsche spirit.
Motivation
For the next generation of automotive user interface (UI) systems, it will be essential to have a profound understanding of the user interface display and operating behavior. Information we can gain through sensors & external APIs are the enabler for an optimized user interface, hence a more performant driver experience. One factor is being able to predict the next UI interaction based on the driving context.
Objective
The master thesis aims for a recommender system that recognizes operating patterns for classified display & control sequences on a large naturalistic driving data set of vehicle bus information.
The exact structure of the thesis will be defined with the thesis student and the supervisor.
Tasks
- Research on the state of the art in the area of Deep Learning (LSTM, CNN architectures), Time Series Classification and Prediction, Representation Learning for detecting user interaction patterns, and Recommender Systems
- Exploratory and predictive ML model for structural user interaction pattern recognition
- Appropriate feature engineering and fusing high-dimensional multivariate time series data ranging from continuous, categorical and textual input variables
- Model evaluation regarding statistical efficiency, technological feasibility and prediction quality
- Analysis of applied algorithms under human-computer interaction aspects
Start: February/March 2023
Duration: 6 months