Robot Native World Action Model Released! First Spatio-Temporal Unified Architecture, Produced by Fudan University Team
A team from Fudan University has launched STI-WM, a spatiotemporally integrated world model designed for robots. This model aims to overcome limitations in current vision-language-action models by incorporating physics-based constraints and end-to-end integration for more robust real-world robot control. The company behind the model, Moshine Intelligence, has secured significant funding and is partnering with industry leaders to deploy its technology across various robotic applications. AI
IMPACT This new world model could accelerate the development and deployment of more capable and autonomous robots in real-world scenarios.