Figure ramps up humanoid robot manufacturing at unprecedented speed - Robotics & Automation News

Figure AI now claims to be crossing that threshold.

In a recent update, the company revealed that its BotQ manufacturing facility has increased production output of its Figure 03 humanoid robot from one unit per day to one unit per hour – a 24-fold increase in throughput achieved in less than 120 days.

The company says it has already produced more than 350 third-generation humanoid robots and manufactured over 9,000 actuators across more than 10 different SKUs.

While production figures remain small compared with the automotive industry, they are significant within the still-emerging humanoid robotics sector, where most companies remain focused primarily on R&D and limited pilot deployments. ### Figure’s broader AI ambitions

Alongside manufacturing updates, Figure also revealed new developments involving Helix, the company’s humanoid AI system.

According to Figure, its “System 0” whole-body controller now integrates visual perception data from onboard cameras with proprioceptive state information, enabling robots to navigate stairs and uneven terrain more autonomously.

The company says the system is trained end-to-end using reinforcement learning in simulation before being transferred directly to physical robots without real-world fine-tuning.

This “sim-to-real” transfer problem has long been considered one of the central technical bottlenecks in robotics. The company says its battery production line has achieved a 99.3 percent first-pass yield, while overall robot first-pass yield rates now exceed 80 percent and continue improving weekly.

Although the figures cannot be independently verified, they illustrate the growing industrialization of humanoid robot manufacturing.

Humanoid robots become fleet-managed systems

Another important aspect of Figure’s update concerns operational infrastructure.

The company says it has developed:

internal fleet management systems; over-the-air software update infrastructure; field service management tools; failure diagnostics systems; and “fallback ladders” that allow robots to recover gracefully from non-critical faults.

This may prove just as important as the robots themselves.