Learning Context-Aware Neural ODE Dynamics for Adaptive Robotic Control

2026-06-13

AI Summary

This work addresses the challenge of stabilizing robotic systems in dynamically uncertain environments, where performance is often degraded by contact transitions, aerodynamic effects, and external disturbances. The authors propose a context-aware neural ordinary differential equation (neural ODE) dynamics model that uniquely integrates context encoding with neural ODEs. Through a two-stage training procedure, the model enables online inference of environmental disturbances directly from state-action histories, adapting to spatiotemporally varying perturbations without requiring full prior knowledge. The resulting dynamics model is seamlessly incorporated into a model predictive control (MPC) framework. Experimental validation on a quadrotor simulator, a Sphero BOLT spherical robot, and a Fanuc robotic arm demonstrates that the approach significantly enhances robustness and adaptability to both time-varying and spatially varying disturbances.

Abstract

Robotic systems deployed in uncertain and dynamically changing environments often face variations in contact conditions, aerodynamic effects, and external disturbances that challenge reliable control. To remain effective under model-based control, these systems require dynamics models that can adapt to such changes, especially when direct access to complete environmental information is limited. To enable adaptability and facilitate integration with model predictive control, we propose a context-aware dynamics model based on neural ordinary differential equations, which infers environmental factors from state-action histories using a two-phase training procedure. We validate the approach across diverse robotic platforms, including a quadrotor in simulation, as well as a Sphero BOLT robot and a Fanuc manipulator in real-world experiments. The results demonstrate that our method effectively adapts to temporally and spatially varying environmental changes across different tasks. Videos are available at https://youtu.be/PY0sNyF2rqE , and the source code is available at https://github.com/syyu410-yu/context-aware-neural-ode-control.git .

Problem

• robotic control
• environmental uncertainty
• adaptive dynamics
• model-based control
• external disturbances

Innovation

• context-aware
• neural ODE
• adaptive control
• model predictive control
• robotic dynamics