天娱数科Behavision 平台亮相具身智能大会

Core Insights - The speech by Wu Bangyi at the "2025 Embodied Intelligence Global Hard Technology Development Conference" highlighted the core challenges in the humanoid robotics industry, specifically the "four shortages" that hinder development: lack of a universal algorithm development platform, lack of a universal 3D data platform, lack of standard certification platforms, and lack of scene testing and talent training platforms [1] Group 1: Industry Challenges - The humanoid robotics industry is facing significant challenges, including repeated R&D investments due to the absence of a universal algorithm development platform [1] - The lack of a universal 3D data platform restricts the efficient training of embodied intelligence [1] - Poor product compatibility and ecological collaboration are caused by the absence of standard certification platforms [1] - The lack of scene testing and talent training platforms further obstructs the transition of technology from laboratory settings to practical applications [1] Group 2: Company Solutions - To address these industry challenges, the company has developed the Behavision spatial intelligence MaaS platform, which operates on the logic of "world perception - spatiotemporal reasoning - generalized execution" [1] - The company has established a leading high-quality 3D and multimodal data system, accumulating 1.5 million 3D data points and 650,000 multimodal data points, including 200,000 specialized data points for robot task training [2] - The company’s 3D hinge data has been recognized as a typical case in the "Beijing Industry High-Quality Data Set" [2] Group 3: Technological Advancements - The company emphasizes that the implementation of physical AI begins with a deep understanding of the physical world, requiring machines to comprehend geometric structures, material hardness, and motion laws [2] - The Behavision platform integrates sensing, cognitive decision-making, and motor execution into a unified support system, addressing the "brain-body coordination" bottleneck in humanoid robots [3] - The platform successfully bridges the "data gap" in physical AI through the deep application of 3D hinge data, which covers appearance, structure, physical properties, and semantics [3] Group 4: Industry Impact - The construction of the Behavision system creates a complete industrial closed loop of "data - algorithm - platform - scene," facilitating the transition of embodied intelligence from laboratory settings to various industries [3] - This development path, centered on high-quality data and collaborative platforms, not only builds technological and ecological advantages for the company but also supports the implementation of the "Artificial Intelligence+" strategy [3]