top of page


Mr. John Smith

Job title



From lab to dyno to car: transfer learning for brake NVH Winter, Geier, Stender, Saigol, Thévenot, Dufrénoy, Chancelier, Hamdi, Deutzer, Hoffmann The transfer from highly instrumentalized lab environments to dynamometers and from there to the real-world driving scenario in a car has been, and still is, a major challenge for research and development. Often times, it is very hard to reproduce phenomena encountered in the vehicle on dynamometers. It can be even harder to learn from lab tests, such as pin-on-disk experiments, about the final tribological behavior of a brake system during operation. This contribution addresses the transfer of machine-learning based NVH prediction methods on data from lab to car in terms of potentials and limitations. Deep learning prediction models are generated for large NVH data sets acquired from pin-on-disk, quarter-car dynamometer and real-world driving experiments. Given the brake system loading sequences, the neural architectures are predicting the occurrence of noise and vibrations. Across the different experiments, different measurement dimensions are available, and different system behavior is observed. Our study analyses the overall performance of optimal prediction methods, as obtained through hyperparameter studies, and compares whether qualitative differences in the prediction tasks can be read from the respective neural prediction models. The results are utilized to identify key physical parameters, i.e. measurement channels, that are required for accurate predictions and substantial contribution to understanding the system’s NVH behavior. The final objective of this endeavor is to include data-driven NVH and particle emission prediction modules into a next-generation brake control strategy for electric vehicles for emission-reduced and energy-efficient braking. The systematic study across different system integration levels is therefore a fundamental building block for integrating machine learning-based intelligence into future brake systems.

Mr. Nathanael Winter, Research Assistant, Technical University Hamburg; Prof. Dr.-Ing. Merten Stender, Chair of Cyber-Physical Systems in Mechanical Engineering, TU Berlin; Ms. Charlotte Geier, Research Assistant, TUHH; Dr. Maël Thévenot, Post-Doc, Laboratoire de Mécanique, Multiphysique, Multiéchelle LAMCUBE; Prof. Philippe Dufrénoy, Professor, Laboratoire de Mécanique, Multiphysique, Multiéchelle LAMCUBE; Dipl.-Ing. Thierry Chancelier, NVH Manager, Hitachi Automotive Systems, Ltd.; Mr. Said Hamdi, NVH project coordination Eng., Hitachi Astemo; Mr. Marcel Deutzer, Calculation Engineer, Volkswagen AG; Prof. Dr. Norbert Hoffmann, Head of Institute Dynamics Group, TUHH; Mr. Ahmad Saigol, Research assistant, TUHH

From lab to dyno to car: transfer learning for brake NVH

EB2023-BSY-008 • Oral • EuroBrake 2023 • NVH methodology


Sign up or login to the ICC to download this item and access the entire FISITA library.

Upgrade your ICC subscription to access all Library items.

Congratulations! Your ICC subscription gives you complete access to the FISITA Library.


Retrieving info...

Available for purchase on the FISITA Store


bottom of page