Core Insights - Figure AI has launched its next-generation robot, Figure 03, capable of performing various household tasks such as serving tea, doing laundry, and folding clothes [1] Group 1: Product Features - Figure 03 features a redesigned sensor suite and hand system, enhancing its adaptability for a wider range of applications [2] - The visual system of Figure 03 includes a new camera architecture that doubles the frame rate, reduces latency to one-fourth, and expands the single-camera field of view by 60%, providing higher density and more stable perception data [2] - The hand system has been improved with softer, more adaptable materials, increasing contact area for better stability in grasping various objects, thus enhancing dexterity and tactile perception [2] Group 2: AI and Decision-Making - The Helix AI system enables Figure 03 to make autonomous decisions by integrating perception, language understanding, and learning control, allowing it to process natural language commands and execute complex actions without pre-set scripts [3] - The hardware upgrades of Figure 03 are designed to support the Helix system, creating a more precise sensory apparatus and flexible execution limbs [3] Group 3: Commercialization and Production - Figure 03 is designed for high-volume manufacturing, with a dedicated factory, BotQ, initially planned to produce 12,000 units annually, aiming for a cumulative production of 100,000 units over four years [4] - The company has implemented a vertical integration strategy for key components, significantly reducing unit manufacturing costs while ensuring quality and delivery standards [4] - The launch of Figure 03 is expected to accelerate commercialization and may serve as a model for domestic robot manufacturers [4] Group 4: Investment Opportunities - The release of Figure 03 indicates progress in commercialization, with potential for the robot to transition from the technology exploration phase to the commercial growth phase [4] - Investment opportunities may arise in sectors related to robotic operating systems and large-scale robot models [4]