Metin, who has a great passion for the world of technology and science, enjoys sharing his knowledge with the readers of DonanımHaber. Figure AI, a US-based robotics company, has significantly increased its production rate for the human-like robot model Figure 03 at the BotQ production facility from one unit per day to one unit per hour in just four months. This development clearly indicates the company’s transition from the prototype stage to mass production.
The production increase was made possible through the company’s specialized software infrastructure and a production line consisting of more than 150 interconnected workstations. Figure has announced that it has delivered over 350 robots to date and aims to reach an annual production capacity of 12,000 robots, with plans to further increase this capacity. The company also revealed its commitment to producing 55 human-like robots this week alone.
In order to enhance production efficiency, Figure tightened quality standards in the supply chain and added more than 50 interim checkpoints in the production process. As a result of this approach, the success rate on the final production line exceeded 80% on the first attempt, particularly achieving remarkable results in critical components such as a battery efficiency of 99.3% and the production of over 9,000 actuators. Furthermore, the company has completed shipments of over 500 robots, each undergoing comprehensive post-production testing.
With the company’s robot fleet expanding, more scaled data is being collected for the artificial intelligence model Helix used in the robots, contributing directly to the development of their autonomous capabilities. Figure has also developed remote update (OTA), service, and fleet management systems to manage its robot fleet efficiently, allowing real-time monitoring, updates, and feedback integration into the development processes.
In addition to advancements on the production side, Figure announced a significant update to its artificial intelligence model. The new version of the Helix System 0 (S0) model enables robots to perceive their surroundings and move accordingly by detecting visual data through stereo cameras, creating a three-dimensional map of the environment. This update allows robots to both “sense” and “see” their surroundings, enhancing their ability to navigate in complex environments. The update also enables learned behaviors to be directly transferred to the real world without additional calibration.
In conclusion, Figure 03 robots have evolved to handle tasks such as climbing stairs, moving steadily on different surfaces, and maintaining stable performance under varying light conditions.
About the Author
