Tesla has informed employees that it will start gathering motion-capture data to train its Optimus humanoid robots at the Austin Gigafactory beginning in February. This follows more than a year of data collection and prototype training at its Fremont, California plant, where dedicated teams of several dozen trainers—separated from production workers—have recorded themselves organizing vehicle parts, operating conveyor belts and performing simple assembly tasks. Trainers wear helmets fitted with multi-angle cameras tethered to backpacks (with a lighter fanny-pack prototype also trialed) to capture high-fidelity video that feeds Tesla’s machine-learning pipelines. CEO Elon Musk recently indicated Optimus is already executing “simple tasks” in these facilities, with plans to advance to more complex industrial functions by year’s end and a public sale by late 2026. The move to Austin not only expands Tesla’s internal robotics testing ground but also underscores the company’s long-term strategy to offset labor shortages and boost factory efficiency through autonomous humanoid operators.