TY - GEN
T1 - Analysis of Randomization Effects on Sim2Real Transfer in Reinforcement Learning for Robotic Manipulation Tasks
AU - Josifovski, Josip
AU - Malmir, Mohammadhossein
AU - Klarmann, Noah
AU - Zagar, Bare Luka
AU - Navarro-Guerrero, Nicolas
AU - Knoll, Alois
N1 - Publisher Copyright:
© 2022 IEEE.
PY - 2022
Y1 - 2022
N2 - Randomization is currently a widely used approach in Sim2Real transfer for data-driven learning algorithms in robotics. Still, most Sim2Real studies report results for a specific randomization technique and often on a highly customized robotic system, making it difficult to evaluate different randomization approaches systematically. To address this problem, we define an easy-to-reproduce experimental setup for a robotic reach-and-balance manipulator task, which can serve as a benchmark for comparison. We compare four randomization strategies with three randomized parameters both in simulation and on a real robot. Our results show that more randomization helps in Sim2Real transfer, yet it can also harm the ability of the algorithm to find a good policy in simulation. Fully randomized simulations and fine-tuning show differentiated results and translate better to the real robot than the other approaches tested.
AB - Randomization is currently a widely used approach in Sim2Real transfer for data-driven learning algorithms in robotics. Still, most Sim2Real studies report results for a specific randomization technique and often on a highly customized robotic system, making it difficult to evaluate different randomization approaches systematically. To address this problem, we define an easy-to-reproduce experimental setup for a robotic reach-and-balance manipulator task, which can serve as a benchmark for comparison. We compare four randomization strategies with three randomized parameters both in simulation and on a real robot. Our results show that more randomization helps in Sim2Real transfer, yet it can also harm the ability of the algorithm to find a good policy in simulation. Fully randomized simulations and fine-tuning show differentiated results and translate better to the real robot than the other approaches tested.
UR - http://www.scopus.com/inward/record.url?scp=85146318787&partnerID=8YFLogxK
U2 - 10.1109/IROS47612.2022.9981951
DO - 10.1109/IROS47612.2022.9981951
M3 - Conference contribution
AN - SCOPUS:85146318787
T3 - IEEE International Conference on Intelligent Robots and Systems
SP - 10193
EP - 10200
BT - IEEE/RSJ International Conference on Intelligent Robots and Systems, IROS 2022
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems, IROS 2022
Y2 - 23 October 2022 through 27 October 2022
ER -