China begins testing humanoid robots at the busy China–Vietnam border, signaling a bold expansion of AI-powered automation in security operations.
![]() |
| AI-powered humanoid robots begin border patrol trials in southern China, highlighting Beijing’s push to lead the global robotics industry. Image: CH |
Beijing, China — February 14, 2026:
At one of the busiest crossings between China and Vietnam, a new kind of border guard is reporting for duty — not in uniform, but in metal and code.
China has begun testing a humanoid robot named Walker S2 at the Fangchenggang border checkpoint in Guangxi province. Developed by Shenzhen-based Ubtech Robotics, the AI-powered machine is part of a broader experiment that could redefine how border security and crowd management are handled in the world’s most populous nation.
Fangchenggang sees thousands of freight trucks, buses and tourists moving between China and Vietnam every day. Managing that volume requires constant coordination — a task now being partially assigned to humanoid automation.
Walker S2’s current responsibilities are operational rather than enforcement-driven. The robot guides passengers in queues, directs vehicles, answers routine questions and patrols corridors and waiting areas. If abnormal crowd formations or blocked pathways are detected, it can immediately notify human supervisors.
The deployment reflects China’s push to digitize and automate public infrastructure, particularly in high-traffic transport hubs.
Unlike fixed surveillance systems or wheeled patrol robots, Walker S2 is designed as a full-fledged humanoid platform. Equipped with advanced cameras, depth sensors and force-feedback systems in its joints, it can maintain balance in crowded environments and avoid collisions.
Its AI system allows it to interpret surroundings and make limited autonomous decisions in complex scenarios. In a notable feature aimed at continuous operation, the robot can replace its own battery pack when power runs low — reducing downtime and minimizing human intervention.
Such autonomy is critical in border environments that demand round-the-clock vigilance.
The Fangchenggang trial could pave the way for broader deployment. Authorities have indicated that, after evaluation, similar robots could be stationed at airports, seaports and major railway stations.
Walker S2 is already being used internationally for industrial inspections in steel and aluminum plants, where high temperatures and hazardous conditions pose risks to human workers. Expanding into public security roles represents a significant shift in scope.
China has set a goal of establishing a global innovation system in the humanoid robotics industry by 2025. As part of that ambition, a 3,000-square-meter robotics training center has opened in Beijing to accelerate development and commercialization.
The border deployment serves both practical and symbolic purposes. Practically, humanoid robots can handle repetitive guidance tasks, operate continuously without fatigue and reduce staffing pressures. Symbolically, they signal China’s determination to lead in advanced robotics and AI integration.
However, questions remain about reliability in unpredictable human environments and how such systems would perform during emergencies or security incidents. For now, the robots function as assistants rather than replacements for human officers.
Still, their presence at a major international checkpoint suggests a future where automation becomes embedded in everyday governance.
If Walker S2’s trial proves successful, Fangchenggang may mark the beginning of a broader shift — from human-only border patrols to hybrid security systems powered by artificial intelligence.
