analiza aiqbench
The paper's central contribution is a reinforcement learning setup: an airflow simulation extension for Isaac Lab that supports massively parallel training, used to train push-blow and reorientation policies that transfer zero-shot to a physical robot. For ML readers, the useful points are the sim-to-real transfer results (80% success over 20 trials; 2.6° mean heading error) and the reliance on experimentally identified airflow models, whose fidelity limits how far the simulated policies generalise. Robotics is the application, but the RL methodology and simulation tooling are what matter here.
BeyondContact: A Reinforcement Learning Framework for Airflow-Based Non-Contact Manipulation
arXiv:2611.03059v1 Announce Type: new Abstract: Contact-based robotic manipulation is constrained by end-effector reachability and the mechanical properties of objects. Airflow can exert forces without direct robot-object contact, enabling manipulation of objects beyond the reach or when direct contact is undesirable. However, the lack of suitable simulation has hindered the application of reinforcement learning to airflow-based manipulation, limiting the broader adoption of this mechanism compared with contact-based methods. To address this gap, we developed a robot- and task-agnostic airflow extension for Isaac Lab that supports massively parallel training. The extension uses experimentally identified airflow models to compute the forces and torques exerted on objects. Based on this extension, we train a hybrid push-blow (HPB) policy that combines contact-based pushing with airflow manipulation and a rectangular reorientation (RR) policy that controls a cuboid's heading using airflow alone. To enable physical execution, we design a lightweight, high-speed blower that coexists with a gripper. In zero-shot real-world deployment, HPB achieves an 80% success rate over 20 trials, while RR achieves a mean absolute heading error of 2.6 degrees across four target headings. The project materials are available at https://github.com/iit-DLSLab/BeyondContact.
Abstract:Contact-based robotic manipulation is constrained by end-effector reachability and the mechanical properties of objects. Airflow can exert forces without direct robot-object contact, enabling manipulation of objects beyond the reach or when direct contact is undesirable. However, the lack of suitable simulation has hindered the application of reinforcement learning to airflow-based manipulation, limiting the broader adoption of this mechanism compared with contact-based methods. To address this gap, we developed a robot- and task-agnostic airflow extension for Isaac Lab that supports massively parallel training. The extension uses experimentally identified airflow models to compute the forces and torques exerted on objects. Based on this extension, we train a hybrid push-blow (HPB) policy that combines contact-based pushing with airflow manipulation and a rectangular reorientation (RR) policy that controls a cuboid's heading using airflow alone. To enable physical execution, we design a lightweight, high-speed blower that coexists with a gripper. In zero-shot real-world deployment, HPB achieves an 80% success rate over 20 trials, while RR achieves a mean absolute heading error of 2.6 degrees across four target headings. The project materials are available at this https URL.
Peiyu Yang and Mohamad S. Moudallal contributed equally. This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible
Robotics (cs.RO); Systems and Control (eess.SY)
arXiv:2611.03059 [cs.RO]
(or arXiv:2611.03059v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2611.03059 arXiv-issued DOI via DataCite (pending registration)
Text integral sub licența deschisă a publicației · 237 cuvinte