MEKAWIRE
Research

Papers

New robotics and physical-AI papers, with what each one means for the industry.

1 papers
arXiv
teleoperation and data collection for legged manipulationIntermediate

LocoPilot: Single-Operator Bilateral Teleoperation for Loco-Manipulation Policy Learning

Antonio Langella, Giovanni Minelli, Giovanni B. Dessy, Lorenzo Amatucci, Giulio Turrisi, Victor Barasuol, Claudio SeminiIstituto Italiano di Tecnologia (IIT), Dynamic Legged Systems Lab

A single operator can now drive both the legs and the arm of a Unitree Go2 quadruped fitted with an AgileX Piper arm while feeling contact forces estimated without a force sensor, with a noise floor of 1.27 N, according to a preprint from Italy's IIT. The 289 demonstrations it recorded fine-tuned the 5.5-billion-parameter MolmoAct2 model, which then ran the robot through the same 11-number command interface the operator used.

Topic pages

Papers come from arXiv robotics feeds. Where our summary is not written yet, you see the opening lines of the abstract.