Remote Teleoperated Manipulation

In the next stage, users can engage in teleoperated manipulation tasks by remotely controlling Robotin robotic arms to perform household activities such as picking up objects, moving items, and organizing spaces.

Each session generates structured action trajectories and multi-modal sensor data, capturing the nuances of real-world manipulation. This data is essential for training embodied AI models, providing rich examples of human-like interaction in diverse household environments.

Participants earn additional token rewards, calibrated to task complexity and interaction diversity. This incentivizes repeated participation and encourages exploration of varied manipulation scenarios, enhancing dataset richness.

The teleoperation experience is designed to be interactive and gamified, combining utility and engagement to strengthen community involvement and create a sense of ownership in the network’s growth.