Research Article
Federated Learning for Lane-Change Prediction
@INPROCEEDINGS{10.1007/978-3-031-86370-7_19, author={Lilit Yenokyan and William Lindskog-Muenzing and Christian Prehofer and Matthias Schubert}, title={Federated Learning for Lane-Change Prediction}, proceedings={Intelligent Transport Systems. 8th International Conference, INTSYS 2024, Pisa, Italy, December 5--6, 2024, Revised Selected Papers}, proceedings_a={INTSYS}, year={2025}, month={4}, keywords={Federated Learning Lane-Change Prediction Automotive Privacy}, doi={10.1007/978-3-031-86370-7_19} }- Lilit Yenokyan
William Lindskog-Muenzing
Christian Prehofer
Matthias Schubert
Year: 2025
Federated Learning for Lane-Change Prediction
INTSYS
Springer
DOI: 10.1007/978-3-031-86370-7_19
Abstract
In this paper, we present a data-driven approach for predicting lane changes of vehicles in a highway scenario based on observing position, speed, and movements of surrounding vehicles. To train a prediction model, we employ federated learning (FL) with various locations acting as clients. The study employs Long Short-Term Memory (LSTM) networks that utilize 1 s of historical data to forecast lane changes over a 1, 3 and 5-s prediction horizon. We show that personalized FL performs well for a distributed setup without data sharing. The findings demonstrate FL’s potential in automotive safety applications, nearly matching centralized performance while significantly improving data security and privacy across distributed locations. This study supports using federated learning as a viable and robust solution for privacy-preserving predictive tasks in dynamic environments.
