Publication
Dynamic Experience Replay
This paper presents a novel technique in reinforcement learning that largely improves the training efficiency, especially in contact-rich robotic assembly tasks.
Download publicationAbstract
Dynamic Experience Replay
Jieliang Luo, Hui Li
Conference on Robot Learning 2019
We present a novel technique called Dynamic Experience Replay (DER) that allows Reinforcement Learning (RL) algorithms to use experience replay samples not only from human demonstrations but also successful transitions generated by RL agents during training and therefore improve training efficiency. It can be combined with an arbitrary off-policy RL algorithm, such as DDPG or DQN, and their distributed versions. We build upon Ape-X DDPG and demonstrate our approach on robotic tight-fitting joint assembly tasks, based on force/torque and Cartesian pose observations. In particular, we run experiments on two different tasks: peg-in-hole and lap-joint. In each case, we compare different replay buffer structures and how DER affects them. Our ablation studies show that Dynamic Experience Replay is a crucial ingredient that either largely shortens the training time in these challenging environments or solves the tasks that the vanilla Ape-X DDPG cannot solve. We also show that our policies learned purely in simulation can be deployed successfully on the real robot.
Associated Autodesk Researchers
Jieliang (Rodger) Luo
Sr. Principal AI Research Scientist
Related Resources
2024
BrepGen: A B-rep Generative Diffusion Model with Structured Latent GeometryA generative AI model that can learn from and generate boundary…
2024
A hyperreduced reduced basis element method for reduced-order modeling of component-based nonlinear systemsThis method balances accuracy and computational speed through adaptive…
2024
Experiential Views: Towards Human Experience Evaluation of Designed Spaces using Vision-Language ModelsExploratory research on helping designers and architects anticipate…
2022
SkexGen: Autoregressive Generation of CAD Construction Sequences with Disentangled CodebooksWe present SkexGen, a novel autoregressive generative model for…
Get in touch
Something pique your interest? Get in touch if you’d like to learn more about Autodesk Research, our projects, people, and potential collaboration opportunities.
Contact us