Paper 2510.12403
Robot Learning: A Tutorial
- Published
- Oct 2025
- Research lab
- Independent
- Citations
- 5
- GitHub
- 563 stars
01 In brief
Summary
This tutorial by Capuano et al.
(University of Oxford, Hugging Face) provides a comprehensive overview of modern robot learning, tracing the shift from classical, model-based control to data-driven, learning-based paradigms.
It covers foundational concepts in classical robotics, including forward/inverse kinematics and their limitations, then introduces Reinforcement Learning (RL) with algorithms like SAC and HIL-SERL for real-world training.
The tutorial also explores Behavioral Cloning (BC) and generative models such as VAEs, Diffusion Models, and Flow Matching, applied in methods like ACT and Diffusion Policy.
It concludes with generalist Vision-Language-Action (VLA) models like π0 and SmolVLA, which leverage large-scale datasets and pre-trained backbones for cross-task and cross-embodiment generalization.
Practical code examples using the open-source lerobot library are provided throughout, emphasizing the importance of open data and accessible tools in advancing the field.
The tutorial is dated October 15, 2025, and is available on arXiv (2510.12403).
02 From the paper
Abstract
Robot learning is at an inflection point, driven by rapid advancements in machine learning and the growing availability of large-scale robotics data. This shift from classical, model-based methods to data-driven, learning-based paradigms is unlocking unprecedented capabilities in autonomous systems. This tutorial navigates the landscape of modern robot learning, charting a course from the foundational principles of Reinforcement Learning and Behavioral Cloning to generalist, language-conditioned models capable of operating across diverse tasks and even robot embodiments. This work is intended as a guide for researchers and practitioners, and our goal is to equip the reader with the conceptual understanding and practical tools necessary to contribute to developments in robot learning, with ready-to-use examples implemented in $\texttt{lerobot}$.