Continual Robot Policy Learning via Variational Neural Dynamics

• URL: http://arxiv.org/abs/2606.27353v1
• Date: Thu, 25 Jun 2026 17:56:04 GMT
• Title: Continual Robot Policy Learning via Variational Neural Dynamics
• Abstract summary: Most learning-based controllers are trained once and deployed as if learning were complete.<n>This prevents the robot from using deployment experience to further improve task performance.<n>We propose a continual learning framework that uses real-world experience to improve robot policies under hidden and recurring dynamics.
• Score: 20.3323041710497
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Robots deployed in the real world rarely operate under a single fixed dynamics model: wind changes, payloads vary, batteries drain, contacts shift, and hardware wears. Yet most learning-based controllers are trained once and deployed as if learning were complete. This prevents the robot from using deployment experience to further improve task performance. In this work, we propose a continual learning framework that uses real-world experience to improve robot policies under hidden and recurring dynamics. Our method learns a condition-aware dynamics model from real state-action trajectories by combining an analytical physics prior with a neural residual for unmodeled effects. A recurrent encoder infers the current hidden condition from recent interaction, and this estimate conditions both the residual model and the policy. Policy learning is performed via differentiable simulation using diverse learned dynamics sampled from the latent model. At deployment, these sampled conditions are replaced by conditions inferred online from recent real interaction, allowing the policy to recover recurring dynamics by recognition rather than residual re-fitting. Through extensive simulation studies and real-world experiments, we demonstrate that the framework improves policy performance under diverse unobserved disturbances. On real quadrotor trajectory tracking under changing wind, the policy recovers from recurring disturbances in roughly 1s, about 5x faster than online residual re-fitting. It also reduces large-disturbance hover and tracking errors by 65.7% and 53.3% over the state-of-the-art online adaptation approaches

Related papers

• Learning Context-Aware Neural ODE Dynamics for Adaptive Robotic Control [55.01272784312405]
  We propose a context-aware dynamics model based on neural ordinary differential equations.<n>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.
  arXiv  Detail & Related papers  (2026-06-13T20:55:46Z)
• Physics-Informed Policy Optimization via Analytic Dynamics Regularization [5.537513058221965]
  Reinforcement learning (RL) has achieved strong performance in robotic control.<n>State-of-the-art policy learning methods, such as actor-critic methods, still suffer from high sample complexity.<n>We introduce a novel physics-informed RL framework, called PIPER, that seamlessly integrates physical constraints directly into neural policy optimization.
  arXiv  Detail & Related papers  (2026-03-15T16:20:45Z)
• Offline Robotic World Model: Learning Robotic Policies without a Physics Simulator [50.191655141020505]
  Reinforcement Learning (RL) has demonstrated impressive capabilities in robotic control but remains challenging due to high sample complexity, safety concerns, and the sim-to-real gap.<n>We introduce Offline Robotic World Model (RWM-O), a model-based approach that explicitly estimates uncertainty to improve policy learning without reliance on a physics simulator.
  arXiv  Detail & Related papers  (2025-04-23T12:58:15Z)
• Mitigating Covariate Shift in Imitation Learning for Autonomous Vehicles Using Latent Space Generative World Models [60.87795376541144]
  A world model is a neural network capable of predicting an agent's next state given past states and actions.<n>During end-to-end training, our policy learns how to recover from errors by aligning with states observed in human demonstrations.<n>We present qualitative and quantitative results, demonstrating significant improvements upon prior state of the art in closed-loop testing.
  arXiv  Detail & Related papers  (2024-09-25T06:48:25Z)
• DiAReL: Reinforcement Learning with Disturbance Awareness for Robust Sim2Real Policy Transfer in Robot Control [0.0]
  Delayed Markov decision processes fulfill the Markov property by augmenting the state space of agents with a finite time window of recently committed actions. We introduce a disturbance-augmented Markov decision process in delayed settings as a novel representation to incorporate disturbance estimation in training on-policy reinforcement learning algorithms.
  arXiv  Detail & Related papers  (2023-06-15T10:11:38Z)
• Residual Physics Learning and System Identification for Sim-to-real Transfer of Policies on Buoyancy Assisted Legged Robots [14.760426243769308]
  In this work, we demonstrate robust sim-to-real transfer of control policies on the BALLU robots via system identification. Rather than relying on standard supervised learning formulations, we utilize deep reinforcement learning to train an external force policy. We analyze the improved simulation fidelity by comparing the simulation trajectories against the real-world ones.
  arXiv  Detail & Related papers  (2023-03-16T18:49:05Z)
• Active Learning of Discrete-Time Dynamics for Uncertainty-Aware Model Predictive Control [46.81433026280051]
  We present a self-supervised learning approach that actively models the dynamics of nonlinear robotic systems. Our approach showcases high resilience and generalization capabilities by consistently adapting to unseen flight conditions.
  arXiv  Detail & Related papers  (2022-10-23T00:45:05Z)
• Real-to-Sim: Predicting Residual Errors of Robotic Systems with Sparse Data using a Learning-based Unscented Kalman Filter [65.93205328894608]
  We learn the residual errors between a dynamic and/or simulator model and the real robot. We show that with the learned residual errors, we can further close the reality gap between dynamic models, simulations, and actual hardware.
  arXiv  Detail & Related papers  (2022-09-07T15:15:12Z)
• Learning Reactive and Predictive Differentiable Controllers for Switching Linear Dynamical Models [7.653542219337937]
  We present a framework for learning composite dynamical behaviors from expert demonstrations. We learn a switching linear dynamical model with contacts encoded in switching conditions as a close approximation of our system dynamics. We then use discrete-time LQR as the differentiable policy class for data-efficient learning of control to develop a control strategy.
  arXiv  Detail & Related papers  (2021-03-26T04:40:24Z)
• Neural Dynamic Policies for End-to-End Sensorimotor Learning [51.24542903398335]
  The current dominant paradigm in sensorimotor control, whether imitation or reinforcement learning, is to train policies directly in raw action spaces. We propose Neural Dynamic Policies (NDPs) that make predictions in trajectory distribution space. NDPs outperform the prior state-of-the-art in terms of either efficiency or performance across several robotic control tasks.
  arXiv  Detail & Related papers  (2020-12-04T18:59:32Z)
• Never Stop Learning: The Effectiveness of Fine-Tuning in Robotic Reinforcement Learning [109.77163932886413]
  We show how to adapt vision-based robotic manipulation policies to new variations by fine-tuning via off-policy reinforcement learning. This adaptation uses less than 0.2% of the data necessary to learn the task from scratch. We find that our approach of adapting pre-trained policies leads to substantial performance gains over the course of fine-tuning.
  arXiv  Detail & Related papers  (2020-04-21T17:57:04Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.