Ant Lingbo Signs Middle East Cooperation Memorandum; Shutu Technology Closes Third Funding Round in a Month

The memorandum was signed by Dr. Ayman Ghoneim, assistant secretary-general of AFDE, and Yu Liang, commercialization head of Ant Lingbo Technology. The two sides will establish a regular cooperation mechanism to promote the deployment and localization of full-stack embodied intelligence and robotics solutions in the United Arab Emirates, the Gulf Cooperation Council (GCC) and the broader Middle East market. The agreement covers five main areas: regional deployment and commercialization of Ant Lingbo's embodied intelligence technology, models and robotics solutions; joint proof-of-concept and industry applications in agreed sectors including hotels and catering, retail, real estate, logistics, industrial operations, smart cities, healthcare and entertainment tours; localization for Arabic-language interaction, regional operating environments, local workflows, cultural needs and customer demands; exploration of real-scene embodied intelligence data and training collaboration in the Middle East, subject to laws, regulations and customer permission; and cooperation among government agencies, companies, research institutions, universities, investors and industry partners in China and the Middle East to accelerate deployment and ecosystem building.

Ant Lingbo says it builds native general-purpose brains for robots, with full-stack technology spanning spatial intelligence, dexterous manipulation and dynamic environment interaction. As of September, its related open-source models had received 31,800 stars on GitHub, placing it among the leading domestic embodied models, the company said. Ant Lingbo is also preparing to release a 'model cloud toolchain platform.' The company said leading model performance and a complete engineering toolchain system were key reasons AFDE selected it.

Discussing the path to industrialization, Ant Lingbo CEO Zhu Xing said, according to Leiphone.com, that 'generalization capability is the core economic variable determining the success or failure of embodied intelligence commercialization.' He added that Ant Lingbo insists on training an embodied foundation model for the physical world from scratch, aiming to let a 'general brain' seamlessly adapt to robot hardware of different configurations and scenarios.

Based on that technology base, Ant Lingbo's self-developed LingBot-VLA 2.0 foundation model supports 17 brands and 20 robot configurations during pretraining, an industry engineering sample of 'one brain, multiple machines,' according to the company. The solution has been deployed in pharmacies, logistics and industrial loading and unloading. In a 'robot smart pharmacy' built with Shanghai Guoda Pharmacy, a robot equipped with the Lingbo brain can perform all-weather precise medicine picking in an aisle only 80 centimeters wide, easing operating pressure during peak hours and at night.

The cooperation with AFDE, announced during the trade expo, marks international recognition of a Chinese self-developed embodied intelligence 'brain' and opens a new model for international cooperation, moving from hardware cooperation toward deeper work involving core technology models, engineering toolchains, localized scenario adaptation and industrial ecosystems, according to Leiphone.com.

In the funding announcement, Shutu Technology said the Pre-A round was co-led by Dinghui Baifu and Kunpeng Fund, a national-level big data fund, with participation from Guangzhou Industrial Investment and Yihe Capital and an over-subscribed follow-on investment from existing shareholder Trend Investment. The company said the proceeds will further support research and validation of tactile data representation training, push technology and product iteration of its SynaTac tactile product line, and accelerate business layout for robot contact strategy training and edge deployment.

Shutu Technology has built two human behavior learning engines, SynaData for world understanding and SynaTac for contact execution. SynaData converts human operation videos into structured, multimodal training data containing action trajectories, object poses and contact information through video elevation, human-object interaction reconstruction and cross-embodiment mapping. Its Syna-Ego, Syna-Exo and Syna-RR products serve more than 60% of embodied intelligence unicorns valued at RMB 10 billion and have received batch purchases from leading technology manufacturers, according to the company.

SynaTac focuses on unified structured representation of tactile data, including alignment of objects, actions and contact states and fusion of vision, touch and robot states. Shutu Technology said it uses demonstration, interaction and trial-and-error to train contact strategies, corrects actions with high-frequency feedback, and advances edge training and deployment of contact intelligence on dexterous hands and different robot bodies. The company also publicly introduced an embodied data and skills evaluation system covering data metric evaluation, trainability evaluation, and task and real-environment validation, with results feeding back into data production, representation optimization, model training and skill iteration. It said it will accumulate reusable task benchmarks, evaluation methods and testing processes with industry partners to support standards research.

Shutu Technology's CTO, Mu Wei, formerly served as vice president of R&D at Silicon Valley embodied intelligence pioneer Covariant AI and head of its China software operations, according to Leiphone.com. The company's core technical team is based on Covariant AI's algorithm team and has experience in algorithm development and scenario deployment. The company argues that moving robots from 'able to demonstrate' to 'deliverable' requires not only algorithmic breakthroughs but also continuous iteration of data, training and validation around on-site problems. It identifies two core gaps: insufficient world understanding at the brain level and unstable contact execution at the cerebellum level. Human behavior video elevation data helps robots understand tasks, while tactile representation modeling supports reliable execution, the company said.

Editor's Summary Ant Lingbo Technology and the Arab Digital Economy Alliance signed a framework cooperation memorandum covering deployment and localization of embodied intelligence and robotics in the UAE, GCC and the Middle East, while Shutu Technology raised nearly RMB 100 million in a Pre-A round, its third in a month. Both developments point to Chinese embodied AI companies expanding commercialization through international partnerships and specialized data and tactile infrastructure.