Robo.ai subsidiary secures first commercial order for 30,000 hours of embodied AI robot training data

DUBAI, UAE, Feb. 12, 2026 - Robo.ai Inc. said its subsidiary has secured its first commercial order for Embodied AI robot training data. Under the agreement, the subsidiary will provide 30,000 hours of data to U.S.-based DaBoss.AI Inc. over a 12-month delivery period. The scope includes high-value multi-modal raw data such as RGB-D vision, 6-DoF motion trajectories, and force/tactile feedback. The data is designed for training Level 4 and above Embodied AI models. Revenue from the transaction will be consolidated into Robo.ai's financial statements through the holding joint venture under U.S. GAAP. This contract marks the substantive launch of the Embodied AI Data Port strategy. Aiden Zhu, co-founder and U.S. CEO of DaBoss, said the order marks the successful transition of data acquisition from heavy-asset operations to an elastic cloud service model. Benjamin Zhai, CEO of Robo.ai, said the order demonstrates urgent global market demand for high-quality physical common sense data.