Figure AI's humanoid robots have moved beyond controlled experiments into genuine factory production, completing an 11-month deployment at a BMW plant in South Carolina where they processed parts for over 30,000 vehicles. The Figure 02 robot worked 10-hour shifts Monday through Friday, accumulating 1,250 hours of runtime and handling more than 90,000 parts. This is not a simulation or a vendor demonstration; it is a documented production record on a commercial vehicle program running under real shift conditions and real time pressure.

What Made This Factory Deployment Different From Previous Robot Trials?

The significance of the Spartanburg record lies in the specificity and transparency of the results. Every competing plant evaluating a humanoid pilot now has a real baseline, not a vendor's simulation. The BMW plant builds X-series and sports models for global markets, making it one of the highest-volume, highest-value automotive plants in the United States. The robots handled two distinct tasks during the deployment, each revealing different capabilities.

In the first phase, the Figure 02 robot supported the precise removal and positioning of sheet-metal parts for the welding process, a task demanding high speed and accuracy. Welding insertion is repetitive and positionally predictable, which is why traditional industrial robots have long excelled at it. However, the successor model, Figure 03, moved into a far more complex domain: logistics sequencing.

Why Is Sorting Loose Parts Harder to Automate Than Welding?

Components initially arrive in larger containers, completely unsorted. The Figure 03 robot reaches into the bin, identifies each component, and places it into a sequencing trolley in the precise order the assembly line needs. No human hands touch the parts during this step. A mis-sequenced trolley does not just slow one worker; it can stop a moving assembly line, which costs tens of thousands of dollars per minute at full production rate. This is why sequencing was done by hand until now.

The robot pulls the build sequence from the plant's production management system in real time, requiring a feedback loop that traditional factory robots never needed. The hardest part of robotic manipulation has never been moving an arm. It has been knowing what the fingers are touching. Traditional factory robots work with pre-positioned parts on fixed jigs, so they never need to feel anything. A humanoid reaching into a bin of loose, unsorted components faces a completely different problem: it must identify the object, estimate its orientation, plan a safe grip, and adjust in real time if that grip starts to slip.

How Does Figure 03 Actually Feel and Manipulate Objects?

The new robot handles this challenge through redesigned hands equipped with tactile sensors and integrated palm cameras. Palm-mounted cameras give the robot a close-up view of the object just before the fingers close, feeding a vision model that estimates shape and surface. Tactile sensors installed in the fingertips allow the robot to detect forces as light as three grams of pressure, roughly the weight of a paperclip. The robot's fingers are softer than traditional robotic end-effectors to allow a more stable grasp of objects with varied shapes and sizes.

This feedback loop, see, close, feel, correct, is what makes bin-picking of loose automotive components possible without a dedicated feeding system. Automotive parts in a sequencing trolley are often finished surfaces, painted panels, trim pieces, and parts with tight cosmetic tolerances. The tactile feedback loop and compliant fingertip material keep grip force within a safe range for the part and for the worker nearby.

What New Features Did Figure 03 Add Beyond Better Hands?

Figure 03 also introduces wireless charging, upgraded speech-to-speech audio, and soft safety components over its predecessor. Wireless charging lets the robot top up during natural pauses in the workflow rather than going offline for a fixed charge cycle. The upgraded audio system lets a line worker give a verbal instruction without touching a screen. Both features were absent from the predecessor and introduced as part of a set of new capabilities designed for expanded applications.

The soft safety components are not cosmetic. Traditional industrial robots operate inside metal cages because a rigid arm moving at speed will seriously injure anyone it strikes. A humanoid designed to work beside people on an open floor has to absorb accidental contact without transferring that force to a human body. The soft outer shell distributes impact energy across a larger surface area and gives the collision-detection system a fraction of a second more time to halt motion. Combined with sensors that detect a human in the robot's path, that physical property allows it to share an open logistics hall with workers.

Steps to Evaluate Humanoid Robots for Your Manufacturing Facility

• Assess Task Complexity: Determine whether your production tasks are repetitive and positionally predictable, like welding, or require real-time adaptation and object recognition, like bin-picking and sequencing.
• Review Safety Requirements: Evaluate whether your facility can accommodate traditional caged industrial robots or whether you need collaborative robots with soft safety components that can work alongside human workers.
• Examine Data Availability: Confirm that your production management system can provide real-time build sequences and that your facility has the infrastructure to support low-latency communication between the robot's vision model, grip planner, and production schedule.
• Plan for Integration Time: Recognize that real production data from pilot deployments will inform hardware and software improvements, so budget for iterative refinement rather than expecting immediate perfection.

On a factory floor, latency is measured in milliseconds, so the robot's vision model, grip planner, and production-schedule connection all have to operate without a cloud round-trip. This physical AI infrastructure matters as much as the hardware itself.

"The 11-month deployment of Figure 02 proved that humanoids are no longer lab experiments, they can be a valuable asset in establishing a flexible, reliable manufacturing workforce," stated Brett Adcock, founder and CEO of Figure AI.

Brett Adcock, Founder and CEO at Figure AI

BMW Manufacturing confirmed the collaboration demonstrated that humanoid robots can safely perform precise, repeatable work steps under real production conditions. Ulrich Wieland, vice president of production control and logistics at BMW Manufacturing, described Spartanburg as "the birthplace of humanoid robotics in BMW Manufacturing's operational day-to-day activities" and confirmed the plant is moving forward with Figure 03 for a sequencing use case in logistics.

The lessons learned from 11 months of real production data fed directly into Figure 03's redesign. The successor carries rebuilt hands and improved manipulation systems, not as theoretical improvements, but as fixes for documented production findings. Analyst projections suggest humanoid robots could begin addressing a meaningful share of manufacturing labor demand before the end of this decade, a modest start, but one representing large numbers of positions.