What is Figure AI's strategy for commercializing general-purpose robots?
Figure AI CEO Brett Adcock outlined the company's commercialization roadmap during a Bloomberg interview, focusing on industrial deployment of their Figure-02 humanoid platform within 18 months. The $2.6 billion-valued startup is prioritizing warehouse automation and manufacturing applications, leveraging their 30-DOF backdrivable design and multimodal AI systems trained on proprietary datasets. Adcock emphasized that Figure's approach centers on whole-body control algorithms that enable zero-shot generalization across tasks, distinguishing their platform from task-specific automation solutions.
The interview comes as Figure AI has deployed pilot programs with BMW and other manufacturing partners, generating initial revenue streams while refining their sim-to-real transfer capabilities. Adcock noted that their vision-language-action (VLA) models, developed in partnership with OpenAI, enable the Figure-02 to perform complex manipulation tasks without explicit programming for each scenario. This technical foundation supports their goal of achieving unit economics that compete with human labor costs in target applications by 2026.
Figure AI's Technical Differentiation Strategy
Adcock detailed Figure's core technical advantages during the Bloomberg segment, emphasizing their integrated approach to humanoid development. Unlike competitors pursuing modular designs, Figure AI has developed proprietary actuators using custom harmonic drives optimized for backdrivability. This design choice enables the Figure-02 to safely operate in human-centric environments while maintaining the torque density required for industrial tasks.
The company's AI stack represents a significant departure from traditional robotics approaches. Rather than relying on pre-programmed motion primitives, Figure's VLA models process visual and language inputs simultaneously, generating motor commands through end-to-end learning. This architecture has demonstrated success in pilot deployments, where Figure-02 units adapted to new warehouse layouts without additional programming.
Adcock highlighted that Figure's dataset advantage stems from their manufacturing partnerships, which provide access to real-world industrial environments. This contrasts with competitors who rely primarily on simulation data, potentially limiting their sim-to-real performance in complex scenarios.
Market Entry and Revenue Projections
The Figure AI CEO outlined an aggressive commercialization timeline during the interview, targeting initial revenue of $50 million by Q4 2025. This projection assumes successful deployment of 200-300 Figure-02 units across existing pilot programs, with lease rates of $3,000-5,000 per month per unit.
Adcock acknowledged the challenge of scaling production while maintaining quality standards. Figure AI's manufacturing strategy involves partnering with established robotics suppliers rather than building entirely in-house capabilities. This approach aims to achieve production volumes of 1,000 units annually by late 2025, with capacity scaling to 10,000 units by 2027.
The company faces competition from Tesla's Optimus program, Boston Dynamics' Atlas successor, and well-funded startups like Agility Robotics. However, Adcock expressed confidence that Figure's focus on general-purpose capabilities and industrial-grade reliability provides a differentiated market position.
Investment and Partnership Ecosystem
Figure AI's recent funding rounds have attracted strategic investors including Microsoft, NVIDIA, and Amazon's Industrial Innovation Fund. Adcock noted that these partnerships extend beyond capital, providing access to cloud computing resources, semiconductor technology, and potential deployment opportunities.
The Bloomberg interview revealed new details about Figure's partnership with OpenAI, which extends beyond their initial VLA model collaboration. The companies are jointly developing multimodal training methodologies that leverage OpenAI's language models and Figure's robotics datasets. This partnership could accelerate Figure's ability to deploy robots capable of following complex verbal instructions.
Adcock emphasized that Figure AI's investor base provides strategic value beyond funding, with partners offering manufacturing expertise, supply chain access, and customer introductions. This ecosystem approach aims to accelerate market adoption compared to startups pursuing purely venture-backed strategies.
Key Takeaways
- Figure AI targets commercial deployment of 200-300 Figure-02 units by Q4 2025, generating $50 million in initial revenue
- The company's VLA models enable zero-shot generalization across industrial tasks without task-specific programming
- Figure's proprietary harmonic drive actuators provide backdrivability advantages over competitors using traditional servo systems
- Strategic partnerships with OpenAI, Microsoft, and NVIDIA extend beyond funding to provide technical and market access
- Manufacturing partnerships enable production scaling to 10,000 units annually by 2027 without massive capital investment
Frequently Asked Questions
How does Figure AI's approach differ from Tesla's Optimus robot? Figure AI focuses on industrial applications with custom backdrivable actuators and VLA models trained on manufacturing data, while Tesla's Optimus targets consumer markets with automotive-derived components and neural network architectures optimized for Tesla's existing AI infrastructure.
What makes Figure-02's 30-DOF design significant for general-purpose robotics? The 30-degree-of-freedom configuration enables whole-body control that mimics human movement patterns, allowing the robot to adapt to varied tasks and environments without requiring specific programming for each application scenario.
When will Figure AI robots be commercially available for purchase? Figure AI currently operates on a lease model with initial deployments in 2025. Adcock indicated that direct sales may become available by 2026-2027 as production scales and unit costs decrease to competitive levels.
How does Figure AI's partnership with OpenAI impact their robot capabilities? The OpenAI collaboration enables Figure's robots to process natural language commands and visual inputs simultaneously through multimodal AI models, allowing for more intuitive human-robot interaction compared to traditional programmed automation systems.
What are the main technical challenges Figure AI still needs to solve? Primary challenges include achieving reliable sim-to-real transfer for complex manipulation tasks, scaling production while maintaining quality standards, and developing safety systems for human-robot collaboration in industrial environments.