Adapting Generalist Vehicle Models for High-Speed MPC Across Terrains

2026-07-14

AI Summary

This work addresses the challenge of achieving both precise platform-specific control and cross-terrain generalization in high-speed off-road navigation, where conventional vehicle models often fall short. The authors propose OptCar, a method that constructs dynamical context tokens from historical state-action sequences and fine-tunes a universal forward kinematic dynamics foundation model using minimal real-world data—just five minutes per terrain—augmented with system identification–driven synthetic trajectories. A history-conditioned dynamics adaptation module further enhances performance. Evaluated at speeds up to 6 m/s, OptCar reduces trajectory tracking error by approximately 55% over the AnyCar baseline on vegetated and muddy terrains, maintains superior accuracy under out-of-distribution conditions such as unseen trailer loads, and matches the performance of terrain-specialized models on road scenes using only one-sixth of the training data, thereby significantly improving cross-terrain adaptability.

Abstract

High-speed off-road autonomy requires precise closed-loop control for a target vehicle while remaining robust across changing terrains. Recent forward kinodynamic (FKD) prediction foundation models suggest a promising path, starting from a generalist model and specializing it to the target platform. However, effective specialization remains challenging, as it often requires substantial real-world data, and models adapted to one setting can still overfit to specific terrains or driving regimes. We present OptCar (Optimized Car), a recipe for bridging the gap from generalist to specialist FKD models that preserves cross-terrain generalization while optimizing performance for a specific vehicle. $\texttt{OptCar}$ introduces a history-conditioned dynamics adaptation module that encodes recent state-action observations into a dynamics context token, and then fine-tunes the generalist model using limited real-world data together with targeted synthetic rollouts from environment-specific system identification. In closed-loop model predictive control (MPC) experiments across three terrains and an out-of-distribution cart-pulling task, the largest gains appear at 6~m/s, the highest speed evaluated and the regime in which slip dominates tracking error. On vegetation and dirt, the most slip-diverse terrain, OptCar reduces 6~m/s trajectory tracking error by roughly 55% relative to a fine-tuned AnyCar baseline, and remains the most accurate even when an unseen cart payload changes the dynamics. With only 5 minutes of real data per terrain, OptCar is competitive on road with a specialist trained on 30 minutes of road data, and substantially outperforms it once the terrain changes.

Problem

• high-speed off-road autonomy
• cross-terrain generalization
• trajectory tracking
• model adaptation
• slip-dominated dynamics

Innovation

• dynamics adaptation
• foundation models
• model predictive control
• off-road autonomy
• synthetic rollouts