One brain for all: China builds unified AI model to handle complex multi-task robotics
ShengShu Technology has unveiled Motubrain, a unified AI model that acts as a general-purpose brain for robots, combining perception, reasoning, prediction, and action. The model has shown strong benchmark performance and is being used by robotics firms in various environments.
ShengShu Technology has introduced Motubrain, a unified AI model designed to serve as a general-purpose brain for robots. Motubrain integrates video, language, and action to enable robots to complete complex tasks in real-world settings. The model has achieved a 63.77 score on WorldArena and averaged 96.0 across 50 tasks on RoboTwin 2.0. It uses a three-stream Mixture-of-Transformers architecture and is trained on a mix of unlabelled video, simulation data, and multi-robot task recordings. Motubrain can execute multi-step tasks involving up to 10 atomic actions and has demonstrated adaptability in real-world tests. The company is partnering with firms like Astribot and SimpleAI to expand deployment, backed by a $293 million Series B led by Alibaba Cloud.
This content was automatically generated and/or translated by AI. It may contain inaccuracies. Please refer to the original sources for verification.