Use case
Embodied-AI data engineers and robot-learning researchers preparing manipulation-policy training sets need to collect human teleoperation demonstration trajectories, reshape them into episode formats readable by their training stack (e.g. elizaOS), and keep provenance records for the data.
Building custom teleoperation hardware plus recording scripts, or buying/downloading public datasets and hand-writing format-conversion scripts to align with the training framework.
Building a teleoperation rig and recording scripts is costly and fragmented; trajectory formats differ across teams and require extra conversion scripts; provenance is hard to prove to partners or reviewers, so collection effort is hard to reuse.
xOcto's call
Demand is evidenced
The trend is that the bottleneck in robot training data is shifting from models to who collects it and how provenance is proven, and browser teleoperation lowers collection to a computer plus a web page. A wedge is to pick one concrete manipulation setting, such as warehouse sorting, lab pipetting, or home tidying, and package collection, cleaning, and provenance into a data service sold per usable hour to robotics teams that cannot afford their own collection line.
Reason to use it
Why users would choose it
Inference: versus building a collection line, it moves teleoperation into the browser and directly emits episodes plus a trajectory_db converter and on-chain provenance, removing the hardware-build and format-conversion steps, so small robotics teams or researchers lacking collection capacity would choose it when they need to quickly expand demonstration data; public material gives no collection scale, data quality, or reuse evidence.
Where the easy answer breaks down
The tension worth following
An English validation note will follow from the public evidence.
If this is your job
Worth trying. Inference: versus building a collection line, it moves teleoperation into the browser and directly emits episodes plus a trajectory_db converter and on-chain provenance, removing the hardware-build and format-conversion steps, so small robotics teams or researchers lacking collection capacity would choose it when they need to quickly expand demonstration data; public material gives no collection scale, data quality, or reuse evidence.
Entry and what to borrow
The trend is that the bottleneck in robot training data is shifting from models to who collects it and how provenance is proven, and browser teleoperation lowers collection to a computer plus a web page. A wedge is to pick one concrete manipulation setting, such as warehouse sorting, lab pipetting, or home tidying, and package collection, cleaning, and provenance into a data service sold per usable hour to robotics teams that cannot afford their own collection line.