Founded in 2022, Figure AI, a US-based company, has quickly become a notable player in AI-powered humanoid robots.
This article explores Figure AI’s technology, potential impact, and future within the robotics market. We will examine the company’s background, technology, applications, and prospects, considering the competitive landscape and challenges in installing humanoid robots at scale.
The intense early interest in Figure AI suggests a groundbreaking approach in the robotics sector, potentially addressing critical market needs with unique technological advantages.
From newbie to leader
Established in 2022 by serial entrepreneur Brett Adcock, Figure AI aims to “expand human capabilities through advanced AI”.
Adcock also founded Archer Aviation, an electric vertical take-off and landing aircraft company, and Vettery, an AI-based recruitment company – both of which he sold for small fortunes.
Adcock envisions autonomous humanoid robots solving labor economy challenges and becoming a “universal interface in the physical world”, assisting in homes and even in space exploration.
Figure AI has achieved rapid progress, introducing the Figure 01 prototype for manual labor in logistics in 2022. By May 2023, they secured $70 million in funding.
A significant $675 million venture capital raise in February 2024, with participation from Jeff Bezos, Microsoft, Nvidia, and OpenAI, valued the company at $2.6 billion.
A partnership with OpenAI to develop specialized AI models followed , which concluded in 2025. August 2024 saw the launch of the second-generation Figure 02, designed for industrial environments.
Demonstrating manufacturing ambitions, Figure AI unveiled BotQ in March 2025, a high-volume facility targeting 12,000 robots annually.
Plans include shipping 100,000 robots by 2029 , and a potential future funding round could raise an additional $1.5 billion, potentially increasing their valuation to $39.5 billion.
Brett Adcock’s history of successful ventures indicates strong leadership. The rapid increase in Figure AI’s valuation reflects investor confidence in their technology and future prospects.
The strategic shift to in-house AI development in 2025, marked by the end of the OpenAI partnership, was driven by a “major breakthrough” in their own AI capabilities, aiming for greater autonomy and a competitive edge.
Inside Figure AI’s humanoid robots
Figure AI has developed two main robot models. Figure 01, the initial prototype in 2022, targeted logistics and warehousing, standing about 5 feet 6 inches tall and weighing around 130 pounds.
It had a 5-hour battery life and used cameras, LiDAR, and tactile sensors. Demonstrations showed it performing tasks like dynamic walking and making coffee.
Figure 02, launched in August 2024, is an improved model for industrial use. It features integrated cabling, a torso battery, six RGB cameras, and an onboard vision language model. Its Nvidia RTX GPU-based modules offer three times the inference capability of Figure 01.
Equipped with microphones and speakers, it initially used OpenAI for speech-to-speech conversation. The redesigned hands have 16 degrees of freedom and can carry up to 25 kg. Figure 02 was tested at a BMW plant and has a 10-hour runtime with a speed of 1.2 meters per second.
A key advancement is the Helix VLA Model, introduced in February 2025. This generalist model unifies perception, language, and control, enabling high-rate continuous upper body control with 35 degrees of freedom. It can operate two robots simultaneously for collaborative tasks on unfamiliar objects.
Helix allows robots to pick up household items with natural language prompts , using a single neural network for various behaviors without task-specific tuning. Designed for commercial use, it runs on embedded GPUs.
Helix comprises System 2 (7 billion parameters) for reasoning and System 1 (80 million parameters) for fast control at 200Hz , trained on 500 hours of teleoperation data. It has shown effectiveness in logistics, exceeding human speeds in package handling.
Figure AI’s in-house AI breakthrough, embodied in Helix, led to the end of the OpenAI partnership. They are also using reinforcement learning (RL) to train their robots’ locomotion in high-fidelity simulations, achieving natural walking gaits that transfer to physical robots without fine-tuning.
The BotQ Initiative
Figure AI unveiled BotQ in March 2025, a high-volume manufacturing facility in California with an initial capacity of 12,000 robots per year. This facility marks a significant step towards commercial production.
Figure AI employs a vertically integrated manufacturing approach, bringing the build process in-house for quality control. This includes in-house assembly of key components like actuators, hands, and batteries, while leveraging outside vendors for piece parts. They have assembled a team of production line experts.
BotQ plans to use robots building robots, with Figure AI’s humanoids handling material within the facility. A robust software infrastructure was developed over six months, including MES, PLM, ERP, and WMS, with a custom MES for real-time production tracking and quality control.
To achieve high production rates, Figure AI redesigned its robots, leading to the Figure 03 model focused on manufacturability and reduced part count. They transitioned from CNC machining to tooled processes like injection molding, significantly reducing manufacturing time and costs.
Dedicated safety and reliability teams at BotQ conduct accelerated lifecycle tests to ensure product reliability. Figure AI is also implementing targeted automation on its production line for quality and speed, such as automated grease dispensing and battery cell testing.
Real-world applications for humanoids
Figure AI is targeting diverse applications, starting with manufacturing. A commercial agreement with BMW Manufacturing in January 2024 aimed to deploy robots in their automotive plants.
Testing at BMW’s Spartanburg plant involved Figure 02 inserting sheet metal parts. While initial claims about the scale of deployment were debated , the collaboration continues for data collection and further training.
The logistics and warehousing sector is another key target, with reported discussions with UPS for potential integration of humanoid robots. Tasks considered include sorting parcels and handling packages. Figure AI demonstrated its Helix VLA model performing logistics tasks efficiently.
Looking ahead, Figure AI aims for domestic tasks and home applications , with plans for “alpha testing” in residential settings in 2025.
Demonstrations of the Helix AI show its ability to understand commands and manipulate household objects, including collaborative tasks like storing groceries. Helix exhibits strong object generalization. Figure AI also envisions robots assisting the elderly and in space exploration.
Figure AI’s competition
Figure AI faces competition from several companies in humanoid robotics. Tesla’s Optimus aims for affordable, general-purpose robots with high production goals.
Agility Robotics, backed by Amazon, focuses on logistics and warehousing with their Digit robot already in use. Apptronik, partnering with Google DeepMind, develops the Apollo robot for logistics, manufacturing, and retail.
Sanctuary AI focuses on general-purpose industrial robots mimicking human movement. Boston Dynamics, owned by Hyundai, is known for its agile Atlas robot and has commercially released Stretch for warehouse tasks.
1X Technologies (Norway), backed by OpenAI, develops humanoid robots for home use, with their NEO robot in pilot programs. Unitree Robotics (China) develops smaller, more affordable humanoids like the H1.
Figure AI’s advantages include significant funding, a focus on general-purpose robots, rapid development, the breakthrough Helix VLA model, ambitious manufacturing plans, early partnerships, and a strategic shift to in-house AI development.
Challenges include intense competition, scaling production, ensuring safety and reliability, gaining commercial adoption, addressing ethical concerns, and managing expectations.
A future full of humanoids
Figure AI aims to ship 100,000 robots by 2029, with its BotQ facility having an initial capacity of 12,000 per year.
The potential impact on the labor economy is significant, with Figure AI aiming to address labor shortages in manufacturing, logistics, warehousing, and retail. Their robots could automate manual labor tasks representing a large portion of global GDP , and perform undesirable or unsafe jobs.
Figure AI also has long-term goals for domestic life, envisioning robots performing household chores like dishwashing and laundry. Future applications could include elder care and space exploration.
The global robotics market is projected to reach $286.8 billion by 2032, with humanoid robotics having the potential to become a multi-trillion-dollar industry.
Challenges and considerations
The controversy around the BMW partnership highlights the complexities of real-world deployment. Technical challenges remain in achieving the dexterity and adaptability needed for complex environments, especially homes , along with ensuring safety and efficiency.
Intuitive human-robot interfaces are also crucial for home applications , as is managing the vast data required for AI training and operation.
Ethical and societal considerations include potential job displacement and the need for public acceptance and regulations. Transparency and realistic expectations are essential, as seen with the BMW situation.
Strong potential
Figure AI has rapidly become a significant player in humanoid robotics, driven by its advanced Figure 02 robot, the innovative Helix VLA model, and ambitious manufacturing plans with BotQ.
Strategic partnerships and real-world testing with companies like BMW and potential collaborations like with UPS underscore their commitment to industrial transformation and future entry into the consumer market.
Figure AI shows strong potential in the development of general-purpose humanoid robots, particularly with its in-house Helix AI. Their production goals indicate a belief in substantial market demand to solve labor shortages.
While industrial applications are the immediate focus, their long-term vision includes integrating robots into homes.
The future success of Figure AI and the broader humanoid robotics field depends on continued technological advancements, effective scaling of production, and careful consideration of ethical and societal implications.
Overcoming technical hurdles, ensuring safety and reliability, gaining commercial acceptance, and maintaining public trust will be critical. While the future of work and home life may indeed be increasingly “humanoid,” Figure AI is at the forefront of making this vision a reality.
