Ant Lingbo Showcases Cross-Body General Robot Brain at 2026 Bund Conference
At the 2026 Bund Conference tech exhibition, Ant Lingbo Technology drew media attention not with dancing or flipping robots, but with practical demonstrations.
These included a robot precisely picking a box of medicine from a shelf and delivering it to a user, continuous sorting in logistics, and loading/unloading on industrial production lines.
At the Bund Conference, embodied intelligence is moving into the physical world. This is what makes the exhibition area worth watching.
In past years, the most eye-catching robots were those that ran, jumped, and flipped. But when robots enter pharmacies, warehouses, and factories, the challenge changes: they must not only perform an action but also understand the environment, interpret tasks, predict changes, and work continuously in unstructured settings.
Lingbo aims to solve precisely this problem.
A representative case at the exhibition is a "smart pharmacy" recreated on site. A robot receives an order, moves quickly to a shelf, extends its arm, precisely picks a box of medicine, and delivers it to the pickup window. This "robot smart pharmacy," built by Ant Lingbo and Shanghai Guoda Pharmacy, is already deployed in Guoda retail stores.
On-site staff told reporters that real pharmacies are more complex: similar-sized medicine boxes, dense packaging, narrow aisles as small as 80 cm, and visual interference from transparent packaging, reflections, and random placement. In such environments, the robot with Lingbo's brain can autonomously receive orders, identify target medicines, complete grasping, picking, and delivery, and assist pharmacists with sorting during night shifts. More importantly, stores require no modifications.
The robotics industry has long faced the "specialized machine" challenge: each warehouse task requires rewriting control code, and changing an arm configuration means redoing algorithms. Ant Lingbo's solution is an "embodied native general brain." At the expo, logistics and industrial scenarios use different robot bodies from Leju and Tihu, but all are powered by Lingbo's self-developed LingBot-VLA 2.0 model.
Currently, LingBot-VLA 2.0 has been pre-trained to adapt to 17 mainstream robot brands and 20 robot configurations, covering single-arm, dual-arm, bipedal, and wheeled robots. At the Lingbo booth, 机器人大讲堂 observed two significant changes in the embodied AI industry. First, technologically, the robot brain is achieving generalization, supporting 17 configurations in pre-training. Second, industrially, robot services in real scenarios are scaling up. The pharmacy scenario is not a demo but a real service rolling out in Shanghai Guoda stores.
From this perspective, what Lingbo showcases at the Bund Conference is not just a few working robots. It is validating that once robots have a brain capable of crossing bodies, tasks, and continuous learning, embodied AI begins to move from the lab into the real physical world.
At the forum "Beyond the Base: Scenario Breakthrough and Collaborative Evolution of Embodied Intelligence," Ant Lingbo will engage in deep dialogue with top scientists and CEOs of billion-dollar embodied companies, focusing on "embodied ecosystem co-building" and "real scenario implementation," exploring industry trends and technology breakthroughs.
At the main forum, Ant Lingbo's Chief Scientist Shen Yujun will participate in the roundtable "Physical Intelligence: Divergence and Choices" with young scholars and leading entrepreneurs to discuss the development path of embodied intelligence. Shen will share how Lingbo uses its embodied native brain to break through real scenarios, addressing the generational leap from "competing on body" to "competing on brain."
During the conference, Lingbo will also jointly launch the Ant Lingbo Embodied Large Model Challenge with ModelScope and Alibaba Cloud Tianchi, opening technology and scenario needs to developers to accelerate the embodied algorithm ecosystem.
At the 2026 Bund Conference, Ant Lingbo fully demonstrated the cross-body, multi-scenario generalization capabilities and industrial deployment results of its Embodied Native Robot General Brain 2.0, providing a direct example for the industry to enter real scenarios and serve industrial value.