- Partnership aims to turn embodied AI research into large-scale commercial applications
- New accelerator will focus on data, computing power and real-world deployment scenarios
Alibaba DAMO Academy (阿里巴巴达摩院) has partnered with China’s newly launched national pilot base for embodied AI applications in Hangzhou, as the company pushes to accelerate the commercialization of robotics technologies beyond laboratory research.
The agreement, signed on May 16 during the Hangzhou Embodied Intelligence Innovation and Development Conference, will see DAMO Academy jointly build an embodied AI innovation accelerator with the National Pilot Base for Embodied Intelligence Applications, which was launched on the same day.
The initiative aims to help robotics move from research prototypes into large-scale real-world deployment.
Embodied AI — which combines AI with physical machines capable of interacting with the real world — has emerged as one of China’s fastest-growing AI sectors following the rise of large language models.
“This collaboration is an important step in bringing DAMO Academy’s frontier technologies into industrial deployment,” said Zhao Deli, head of DAMO Academy’s embodied AI laboratory. “We hope to combine DAMO’s technical capabilities with the base’s scenario validation and industrial transformation resources so more robotics technologies can move out of the lab and into everyday life.”
Under the agreement, the two sides will jointly develop infrastructure spanning data centers, AI model services and computing platforms, building an integrated system covering data collection, training, inference, deployment and operational feedback.
The accelerator will also focus on deploying embodied AI technologies across consumer-facing sectors including restaurants, education, eldercare and logistics, while supporting developer ecosystems and broader industry collaboration.
DAMO Academy has expanded its investment in embodied AI in recent years, focusing on deployable and scalable robotics systems.
During the 2025 World Robot Conference, the institute open-sourced what it described as the “three core components” of embodied intelligence, including a robotics foundation model that enables robots to interpret instructions and execute actions, as well as standardized protocols connecting AI models with different robotic hardware systems.
Earlier this year, DAMO Academy also open-sourced its embodied AI foundation model, RynnBrain, which the company said gives robots spatial reasoning and spatiotemporal memory capabilities for the first time.
