Developing a bimanual teleoperation device to control to robot manipulator arms for robot learning.

Overview

This project will require someone with a strong mechatronics background with the ability to quickly prototype mechanical designs, electrical designs as well as software to control the two. You will get the chance to work with two Franka Panda robot arms and develop a novel teleoperation frame work that will enbable data collection for training complex imitation learning behaviours. You will be exposed to a cutting edge line if research at the intersection of AI and Robotics.

Research engagement

Mechanical design, electrical design, software development ,

Research activities

You will be working closely with your supervisor as well as other PhD students

Outcomes

The design and development of a fully working teleoperation system that can control 2 robot arms simultaneously, a data collection pipeline and an open source repository which will be released after the project.

Skills and experience

CAD Design (Fusion360, Solidworks), Python programming, Raspberry PI/Arduino experience,

Start date

1 November, 2024

End date

19 February, 2025

Location

S Block Level 11 (Robotics Lab)

Additional information

https://tonyzhaozh.github.io/aloha/

https://diffusion-policy.cs.columbia.edu/

Keywords

• imitation learning
• robot learning

Contact

Dr. Krishan Rana, ranak@qut.edu.au