China is putting a futuristic spin on one of the most ordinary sights in city life: the traffic officer. Humanoid and wheeled policing robots are now appearing at busy intersections, tourist areas and commercial districts, where they can detect selected violations, direct traffic, answer questions and issue verbal warnings.
But the “robocop” label makes the technology sound more powerful than it really is. These machines cannot arrest people, detain them, issue fines or make enforcement decisions. Instead, China is testing a human-machine model in which robots take over repetitive traffic duties while human officers remain in charge when situations get serious.
Cosmic Duo: two familiar public figures known for their space commentary have an unlikely friendship worth knowing about. For more, read our post on Neil deGrasse Tyson and William Shatner’s Cosmic Bromance Explained.
China’s Robot Traffic Police Are Getting More Advanced—But Humans Still Call the Shots
One of the most visible experiments is happening in Hangzhou, Zhejiang province, where SUPCON Information has deployed a 15-robot squad using its T2 traffic-management robot. The machine stands 1.88 meters, or about 6 feet 2 inches, tall and weighs 98 kilograms, or 216 pounds. Despite its humanoid appearance, it moves on a four-wheel chassis rather than walking on two legs.
Its design is deliberately police-like, complete with a black traffic-police cap and fluorescent yellow-green reflective vest. Cameras, radar and onboard computing allow it to identify violations such as riders without helmets, e-bikes carrying passengers, vehicles stopping beyond marked lines and pedestrians crossing against signals.
The T2’s seven-jointed arms can reproduce standardized traffic-police gestures in synchronization with traffic lights. It can also answer questions about directions, parking and traffic rules using large-language-model technology and a touchscreen, while an emergency function can connect users with local police.
SUPCON says the 15 robots began operating in May 2026 around important intersections near the West Lake Scenic Area and major commercial districts. The company claims they have issued more than 170,000 warnings, helped reduce monthly helmetless-riding and stop-line violations by more than 40%, and achieved recognition accuracy above 95%. Those figures have not been independently verified. Hangzhou traffic police did not respond to Reuters’ request for comment.
The robots reportedly operate autonomously from roughly 7 a.m. to 6 p.m., moving between preset positions without remote control, while a centralized platform allows traffic-police teams to monitor and dispatch them. SUPCON says its machines have been tested for thousands of hours in tourist areas, school zones, commercial districts, rush-hour conditions and poor weather, although strong sunlight, rain and temperature extremes can affect cameras and radar.
“It doesn’t get tired.”
That is how Cao Hui, head of SUPCON’s intelligent unmanned systems innovation centre, summed up one of the technology’s biggest selling points. Cao said the goal is not simply to punish violations after they happen, but to issue timely reminders before bad habits become established.
Alien Files Released: a recent order called for the release of long-withheld government records on an unusual subject. To know more, read our story on Trump Orders Release of Alien UFO Files in 2026.
The earlier May deployment also produced another set of figures. By late June, the first 15-machine squad had accumulated 647.7 hours of operational service, delivered more than 25,000 traffic-violation warnings and provided public consultation services more than 2,000 times. During active periods, the robots averaged one reminder every one minute and 43 seconds.
Hangzhou traffic officer Chen Sanchuan called the machines a new partner capable of taking over routine duties.
“The robots’ verbal reminders are objective and direct, so people are more willing to accept them.”
The system uses three warnings. If a violation continues after three reminders, information is archived and sent to an early-warning center under the Hangzhou Public Security Bureau’s traffic-management detachment, where human personnel decide what action should follow.
SUPCON says nearly 50 robots are operating across roughly eight cities in three Chinese provinces and expects that number to approach 200 by the end of 2026. Again, those figures are company claims.
And Hangzhou is only part of the experiment.
In Wuhu, Anhui province, AiMOGA Robotics, backed by automaker Chery, deployed its Intelligent Police Unit R001, also known as ZhiJing R001 or the Wuyou series, around January 10, 2026, China People’s Police Day. It has operated at busy intersections including Zhongjiang Avenue and Chizhu Mountain Road.
R001 wears a police uniform, reflective vest and white cap. It connects with traffic signals, reproduces standard hand signals and uses high-definition cameras, intelligent voice systems and large-model algorithms to detect selected violations involving pedestrians and non-motorized vehicles. It can also navigate independently to designated locations, identify illegal parking and monitor roads.
Its sensor package reportedly includes millimeter-wave radar, LiDAR, ultrasonic sensors and a 360-degree camera.
“a new colleague capable of assisting us effectively.”
Wuhu traffic police officer Jiang Zihao used those words to describe R001, noting that it could work around the clock and reduce workloads during peak periods or extreme weather. One example of its warnings tells cyclists to use the non-motorized lane.
The robotic traffic-police trend is spreading. Shanghai tested “Xiao Hu,” or “Little Tiger,” at an Huangpu District intersection in July 2025, where it used voice commands such as “Red light, please stop” alongside hand signals. In May 2026, AgiBot’s Lingxi X2 humanoid robot worked with human officers and drones in Shanghai’s Zhangjiang AI Innovation Town on urban-management support, including legal education, policy explanations and merchant Q&A, while humans retained legal responsibility.
Shenzhen has tested humanoid robots, including EngineAI models, at intersections such as Wuhe Avenue and Yabao Road. They can synchronize gestures and issue whistle or gesture warnings to helmetless e-bike riders and vehicles crossing stop lines. Possible future uses include assistance with drunk-driving inspections and preliminary accident handling.
Robots have also appeared in Chengdu, Kashgar and Ordos. During the May 2026 holiday, Kashgar and Ordos used robots for traffic direction and related public-safety duties. Global Times reported that two Ordos robot traffic-police units handled traffic direction, traffic-safety education and intelligent patrols. Hangzhou had already tested a traffic-management robot in December 2025.
Milestone Update: a recent rocket launch brought a mix of business and engineering news along with it. For more on that, read our article on SpaceX Starship Flight 13 Launch: IPO and Booster Milestones.
So, what can these machines actually do?
| Function | What the robots handle |
| Violation detection | Selected pedestrian, e-bike and vehicle violations |
| Traffic control | Standardized hand signals and traffic guidance |
| Warnings | Immediate verbal reminders |
| Public help | Directions, parking and traffic-rule questions |
| Emergency support | Connections with local police |
| Data logging | Persistent violations forwarded to human authorities |
| Movement | Navigation between designated positions |
| Endurance | Long repetitive shifts without human fatigue |
The crucial detail is what they cannot do. These are not autonomous police officers. They have no arrest powers, weapons or authority to issue fines. Their role is observation, warning, guidance and information collection. Human officers handle enforcement, accidents, emergencies, serious violations and decisions requiring judgment.
That division is arguably the entire point.
Traffic officers can spend hours dealing with heat, rain, exhaust fumes, noise and crowds. Robots can maintain a visible presence at repetitive posts without experiencing physical fatigue in the same way. SUPCON says its machines are designed for roughly eight-to-nine-hour working periods, although charging and operational requirements mean an individual robot is not literally active around the clock.
There are still obvious limitations. Weather and sunlight can affect sensors, robots require human support and delicate physical tasks can remain difficult. More importantly, some of the headline performance numbers have not been independently verified.
The wider experiment also raises questions beyond traffic. China is increasingly testing embodied AI in delivery, sorting, security, public services and urban assistance. AiMOGA has identified possible overseas interest, particularly in places such as the Middle East and Southeast Asia, although exporting these systems would require local law-enforcement integration, certification and compliance with data-protection rules because the robots depend heavily on cameras and sensors.
For now, China’s traffic “robocops” are best understood as tireless automated assistants—not replacements for police officers. They can watch, warn, guide, answer and report. When something genuinely serious happens, the humans are still the ones in charge.
Impact Report: A rocket stage recently made an unplanned mark elsewhere in the solar system. To know more, read our post on SpaceX Falcon 9 Moon Crater: NASA’s LRO Images.
Disclaimer
This article is based on thorough research of the information and source material provided, supplemented by the cited reporting available as of August 20, 2026. The article has not added capabilities, deployments or performance claims beyond the supplied material and supporting sources. Company-reported figures, including SUPCON’s claimed reduction in certain violations, recognition accuracy and number of warnings, are identified as claims where they have not been independently verified. The article is intended for informational purposes and reflects the available reporting rather than an independent technical assessment of the robots.
Source links
- Reuters
- People’s Daily Online
- Global Times (first squad)
- Global Times (Labor Day patrols)
- The Straits Times
- The Next Web
- Economic Times
Image Credit:
- Featured Image: Fumikas Sagisavas / Wikimedia Commons (CC0 Public Domain)
Leave a Reply