Tesla shifts Optimus training toward vision-only human task video

Business Insider reported that Tesla changed its Optimus training approach toward a more vision-led method, using video of workers performing everyday manipulation tasks.

The report describes a camera-heavy data workflow, which aligns with the need for stable, repeatable, operator-view footage in robot learning programs.

For T-Flash DualCam, the relevance is clear: robotics teams need compact capture hardware that can document human intent, hands, objects, and surrounding context at scale.