teardex

Robots/Parts/Training

First-person human video

Video filmed from a person's own point of view, as if through their eyes, while they go about everyday tasks. Figure uses it to teach its Helix AI how tasks are done, without needing a robot to be shown each one.

1 of 10
robots in the database
Not named
made by
Training
brain layer
Not public
price

Key numbers

Figure Helix
taught to move about homes using only first-person human video, with no robot demonstrations
Recorded in
real homes owned by Brookfield, recorded as people went about their day
Brookfield's homes
more than 100,000 residential units

Source: Figure, Project Go-Big: Internet-Scale Humanoid Pretraining and Direct Human-to-Robot Transfer (figure.ai)

Strengths and trade-offs

Our analysis

Strengths

Far cheaper and quicker to collect than recordings made with a robot.

Covers a huge variety of homes, objects and tasks.

Shows tasks from the same point of view as a robot's head cameras.

Can grow to internet scale.

Trade-offs

Human hands and bodies differ from a robot's, so the lessons do not carry over perfectly.

Video shows what happened but not how hard things were gripped or pushed.

Filming inside people's homes raises privacy questions.

Usually still needs some robot data to fine-tune each task.

Why robot makers choose it

Our read

Robot demonstrations cost an hour of a person's time for every hour of data. People filming their own daily lives produce far more examples, far more cheaply, which is why Figure is betting on it to teach Helix at large scale.

This is our reasoning from the public facts. No maker has confirmed it.

Robots using it

Numbers and listed prices on this page come from the sources shown. Anything marked as our analysis, our read or our estimate is our own view and has not been confirmed by any company. Last checked 14 Oct 2026. How we verify

First-person human video | Teardex