Amortized Variational Inference for Road Friction Estimation

Preprint version, final version will be published in IEEE Xplore.

Abstract

Road friction estimation concerns inference of the coefficient between the tire and road surface to facilitate active safety features. Current state-of-the-art methods lack generalization capability to cope with different tire characteristics and models are restricted when using Bayesian inference in estimation while recent supervised learning methods lack uncertainty prediction on estimates. This paper introduces variational inference to approximate intractable posterior of friction estimate and learns amortized variational inference model from tire measurement data to facilitate probabilistic estimation while sustaining the flexibility of tire models. As a by-product, a probabilistic tire model can be learned jointly with friction estimator model. Experiments on simulated and field test data show that the learned friction estimator provides accurate estimates with robust uncertainty measure in a wide range of tire excitation levels. Meanwhile, the learned tire model reflects well-studied tire characteristics from field test data.

Code

We will soon publish simulation data protocol in CarMaker and datasets. Please follow up this link.

BibTEX

(Preliminary) S. Chen, S.Sihao, L.S. Muppirisetty, Y. Karayiannidis, M. Björkman, “Amortized Variational Inference for Road Friction Estimation,” in IEEE Intelligent Vehicles Symposium, IV 2020.