Figure AI releases Figure 02 autonomy update: fully autonomous warehouse picking without human teleoperation
In one sentence Figure AI's Figure 02 humanoid robot now performs warehouse picking tasks fully autonomously using an LLM backend for natural language task assignment, achieving 95% task success rate in a BMW pilot with 100 deployed units.
Picture a warehouse where robots work entirely on their own, like experienced employees who do not need to be guided step by step. Until recently, even the most advanced industrial robots still needed a human operator watching remotely — a bit like flying a drone — especially for complex tasks. Figure AI has now made a significant leap: their humanoid robot Figure 02 can carry out warehouse picking tasks completely on its own, with no remote human intervention required.
The robot understands orders given in plain language. You can tell it "pick the red box from shelf B3 and bring it to the packing station" and it knows exactly what to do. This is made possible by a system built on large language models — the same kind of technology behind smart chatbots — but wired into a physical robot body that can move, grasp, and carry objects.
In a pilot program at BMW facilities, Figure 02 achieved a 95% success rate on assigned tasks. That is not perfect, but it is a remarkable result for a robot operating in a real, unpredictable physical environment. BMW has deployed 100 of these robots in active operation, meaning this is not a lab experiment but a genuine industrial-scale rollout.
This update marks a shift in how we think about automation: robots are no longer just pre-programmed machines following fixed sequences. They are becoming adaptive agents that can interpret instructions and handle variability. For logistics and manufacturing, it means that advanced automation is becoming genuinely practical even for complex, dynamic operations.
Companies
Figure AI
Tools
Figure 02
Tags
Sources