Ten months ago, Figure introduced its first robot, Figure 01. While it was just another robot in a crowded market, it started gaining attention in January 2024 when BMW began using it in one of its factories. Interest surged in February when the company secured funding from AI giants OpenAI, Nvidia, Microsoft, and Jeff Bezos, fueling expectations for the robot’s future.

Figure 02 did not disappoint. With significant advancements in both hardware and software, the CEO Brett Adcock boldly claim it to be “the world’s most advanced humanoid robot.”

Figure 02 introduces numerous upgrades over its predecessor. Most notably, the hardware includes a dexterous hand with 16 degrees of freedom, allowing its five fingers to perform 16 different movements, enabling it to grasp, manipulate, and even rotate objects with precision. The robot’s computational power has tripled, and it now boasts a second Nvidia GPU to accelerate AI tasks.

Figure 02’s perception and understanding of its surroundings have also improved through the use of 6 RGB cameras and AI models that convert visual data into actionable information. The battery capacity has been doubled to 2.25 kWh, extending its operating time to approximately 7 hours.

On the software side, Figure leverages its partnership with OpenAI to enable the robot to converse with humans using the company’s latest speech-to-speech models. While this model operates in the cloud, a visual language model (VLM) that allows the robot to understand and interact with its environment runs on the robot itself and utilizes Nvidia chips. It was trained on Nvidia’s Isaac Sim simulation platform.

The company has showcased the new robot performing various tasks, and BMW has already begun using it in its factories.

These advancements position the Figure 02 robot as a formidable rival to Optimus (Tesla’s Robot). Although a definitive comparison is challenging without hands-on experience, I am highly optimistic about Figure 02’s potential to outperform Optimus.