Hangzhou’s 15 SUPCON T2 robots have issued more than 170,000 warnings since May 2026, according to the company, but they remain traffic assistants rather than police officers: they cannot arrest people or issue fines,... The wheeled, 1.88 metre, 98 kilogram T2 uses cameras, radar and onboard computing to detect helm...
Published byEdited with GPT-5.6 LunaImages generated with GPT Image 1.5
Research answer

Create a landscape editorial hero image for this Studio Global article: What are China’s new humanoid traffic-police robots, including the SUPCON T2 deployed in Hangzhou, capable of doing; how do they use cameras. Article summary: China’s new “traffic-police” robots are primarily automated public-service assistants, not law-enforcement officers. The Hangzhou SUPCON T2 is a wheeled, 1.88 m, 98 kg humanoid-style robot that detects selected road viol. Topic tags: general, news, general web, user generated. Style: premium digital editorial illustration, source-backed research mood, clean composition, high detail, modern web publication hero. Use reference image context only for broad subject, composition, and topical grounding; do not copy the exact image. Avoid: logos, brand marks, copyrighted characters, real person likenesses, fake screenshots, UI text, readable text, watermarks, charts w
China’s new traffic robots look like a step toward automated policing, but the SUPCON T2 currently occupies a narrower role. In Hangzhou, the wheeled robot monitors selected road behaviour, reinforces traffic signals, provides public information and passes emergencies to human officers. It is an automated safety-and-service assistant—not an autonomous police officer. 1
The T2 is designed for repetitive work at busy intersections and tourist areas. Its reported functions include:
The distinction between detection and enforcement matters. The robot can observe, warn and assist, but reporting on the Hangzhou deployment says it has no power to arrest people or issue fines. Consequential enforcement decisions remain with human officers. 1
The T2’s cameras provide visual information about road users, vehicles, signals and lane markings. Radar adds information about nearby movement and objects. Onboard computing then processes those inputs to identify the limited set of traffic situations the system has been designed to recognize. 1
That combination is useful because traffic scenes are dynamic: a system must distinguish, for example, an e-bike rider from a pedestrian, interpret whether a vehicle has crossed a marked line, and relate a person’s movement to the current signal. The available reporting describes the equipment and detection categories, but it does not provide a complete public technical specification or an independently validated test protocol.
Some online reports repeat a recognition-accuracy figure above 95%, but that figure is presented as a company claim and has not been independently verified in the supplied reporting. 5
6 It should not be treated as a general guarantee of performance in every intersection, weather condition or crowd.
The T2 is intended to handle routine duties autonomously, reducing the need for an officer to perform every repetitive warning or traffic-direction task. Its arms make the system visibly legible to road users: rather than relying only on a speaker, it can reproduce traffic-control gestures alongside the signals. 1
2
This is an assistive form of autonomy. The robot can monitor and respond within its programmed role, but it does not replace human judgment where an incident requires legal authority, discretion or physical intervention. A warning also has no direct legal force unless a human authority separately takes action.
SUPCON Information deployed 15 T2 robots around Hangzhou’s West Lake area from May 2026. The company says the fleet has issued more than 170,000 warnings and helped reduce some helmetless-riding and stop-line violations by more than 40%. 1
Those figures indicate operational activity, not proof that the robots caused the reduction. They are company-reported results, and the supplied sources do not establish how detections were independently checked, how many warnings were correct, or whether other changes in traffic management contributed to the outcome.
The available reporting also does not establish a reliable public figure for daily working hours, battery endurance or independently measured accuracy. Claims about those details should therefore be treated cautiously.
A traffic robot’s performance depends on more than its camera and radar hardware. Real deployments must account for:
The reported warning total does not show that every alert was accurate, nor does the reported decline in violations establish a causal effect. Those questions would require transparent evaluation data and testing outside the company’s own operating environment.
The opportunity is broader than traffic enforcement. A bounded robot could potentially guide vehicles and pedestrians at large junctions, tourist zones, airports and major events; provide multilingual information; issue non-coercive safety reminders; and connect incidents to human dispatchers.
Chery-backed AiMOGA has said it sees demand in the Middle East and Southeast Asia. Reuters reported that the company had deployed 110 humanoid police robots across multiple Chinese cities and had delivered more than 3,000 robots globally, including 2,000 overseas, across more than 60 countries and regions. 17
That does not mean a Chinese traffic robot can be transferred directly into another country’s police system. The more realistic export model is a public-service or safety robot with tightly limited authority, local supervision and clear escalation procedures.
A buyer would need to define what the robot is legally allowed to observe, say and record. It would also need to assign responsibility for collisions, incorrect warnings, missed hazards, bad directions and failures to escalate emergencies. In most practical deployments, the robot would need to remain advisory while human officers retain enforcement authority.
Cameras and radar can capture information about people, vehicles and movement even when facial identification is not the product’s stated purpose. An overseas deployment would therefore need clear rules for data minimization, public notice, retention, deletion, access, auditing and human review. Cross-border data transfers and remote system access could receive additional scrutiny.
A connected robot operating near police networks, traffic signals or other critical systems would need strong controls around authentication, software updates, remote access, supply-chain security and incident response. Buyers may also require security testing and detailed visibility into how data and updates are handled.
Export versions may need to meet local requirements for machinery safety, radio equipment, batteries, electromagnetic compatibility, public-space mobility and cybersecurity. The robot would also need to understand local road markings, signal conventions, languages, accessibility expectations and traffic behaviour.
The robot’s value depends on whether it can safely connect to traffic-signal controllers, dispatch systems, mapping, communications and approved video-management platforms. Those interfaces may be fragmented, proprietary or restricted for security reasons.
A humanoid form can make a machine easier to notice, but a visible camera-equipped robot may also raise concerns about surveillance, bias and overcollection. Public agencies would need transparent notices, auditable alerts, clear limits on identification and a human decision-maker for consequential actions.
The SUPCON T2 shows how humanoid-style machines can move from demonstrations into narrow municipal roles. Its strongest near-term use is not replacing traffic police, but handling repetitive observation, guidance, warnings and public assistance while human officers retain authority.
For overseas buyers, the central question is not whether a robot can look and gesture like an officer. It is whether the system can operate safely, lawfully and transparently inside local traffic, privacy, cybersecurity and emergency-response frameworks.
Studio Global AI
This page includes a source-backed answer you can continue inside Studio Global.
Hangzhou’s 15 SUPCON T2 robots have issued more than 170,000 warnings since May 2026, according to the company, but they remain traffic assistants rather than police officers: they cannot arrest people or issue fines,...
Hangzhou’s 15 SUPCON T2 robots have issued more than 170,000 warnings since May 2026, according to the company, but they remain traffic assistants rather than police officers: they cannot arrest people or issue fines,... The wheeled, 1.88 metre, 98 kilogram T2 uses cameras, radar and onboard computing to detect helmetless e bike riding, extra passengers, stop line violations and pedestrians crossing against signals.
The same model can synchronize arm gestures with traffic lights, answer questions about directions and parking, issue spoken warnings and connect emergencies to human police.
Hangzhou’s 15 SUPCON T2 robots have issued more than 170,000 warnings since May 2026, according to the company, but they remain traffic assistants rather than police officers: they cannot arrest people or issue fines,... The wheeled, 1.88 metre, 98 kilogram T2 uses cameras, radar and onboard computing to detect helm...
Published byEdited with GPT-5.6 LunaImages generated with GPT Image 1.5
Research answer

Create a landscape editorial hero image for this Studio Global article: What are China’s new humanoid traffic-police robots, including the SUPCON T2 deployed in Hangzhou, capable of doing; how do they use cameras. Article summary: China’s new “traffic-police” robots are primarily automated public-service assistants, not law-enforcement officers. The Hangzhou SUPCON T2 is a wheeled, 1.88 m, 98 kg humanoid-style robot that detects selected road viol. Topic tags: general, news, general web, user generated. Style: premium digital editorial illustration, source-backed research mood, clean composition, high detail, modern web publication hero. Use reference image context only for broad subject, composition, and topical grounding; do not copy the exact image. Avoid: logos, brand marks, copyrighted characters, real person likenesses, fake screenshots, UI text, readable text, watermarks, charts w
China’s new traffic robots look like a step toward automated policing, but the SUPCON T2 currently occupies a narrower role. In Hangzhou, the wheeled robot monitors selected road behaviour, reinforces traffic signals, provides public information and passes emergencies to human officers. It is an automated safety-and-service assistant—not an autonomous police officer. 1
The T2 is designed for repetitive work at busy intersections and tourist areas. Its reported functions include:
The distinction between detection and enforcement matters. The robot can observe, warn and assist, but reporting on the Hangzhou deployment says it has no power to arrest people or issue fines. Consequential enforcement decisions remain with human officers. 1
The T2’s cameras provide visual information about road users, vehicles, signals and lane markings. Radar adds information about nearby movement and objects. Onboard computing then processes those inputs to identify the limited set of traffic situations the system has been designed to recognize. 1
That combination is useful because traffic scenes are dynamic: a system must distinguish, for example, an e-bike rider from a pedestrian, interpret whether a vehicle has crossed a marked line, and relate a person’s movement to the current signal. The available reporting describes the equipment and detection categories, but it does not provide a complete public technical specification or an independently validated test protocol.
Some online reports repeat a recognition-accuracy figure above 95%, but that figure is presented as a company claim and has not been independently verified in the supplied reporting. 5
6 It should not be treated as a general guarantee of performance in every intersection, weather condition or crowd.
The T2 is intended to handle routine duties autonomously, reducing the need for an officer to perform every repetitive warning or traffic-direction task. Its arms make the system visibly legible to road users: rather than relying only on a speaker, it can reproduce traffic-control gestures alongside the signals. 1
2
This is an assistive form of autonomy. The robot can monitor and respond within its programmed role, but it does not replace human judgment where an incident requires legal authority, discretion or physical intervention. A warning also has no direct legal force unless a human authority separately takes action.
SUPCON Information deployed 15 T2 robots around Hangzhou’s West Lake area from May 2026. The company says the fleet has issued more than 170,000 warnings and helped reduce some helmetless-riding and stop-line violations by more than 40%. 1
Those figures indicate operational activity, not proof that the robots caused the reduction. They are company-reported results, and the supplied sources do not establish how detections were independently checked, how many warnings were correct, or whether other changes in traffic management contributed to the outcome.
The available reporting also does not establish a reliable public figure for daily working hours, battery endurance or independently measured accuracy. Claims about those details should therefore be treated cautiously.
A traffic robot’s performance depends on more than its camera and radar hardware. Real deployments must account for:
The reported warning total does not show that every alert was accurate, nor does the reported decline in violations establish a causal effect. Those questions would require transparent evaluation data and testing outside the company’s own operating environment.
The opportunity is broader than traffic enforcement. A bounded robot could potentially guide vehicles and pedestrians at large junctions, tourist zones, airports and major events; provide multilingual information; issue non-coercive safety reminders; and connect incidents to human dispatchers.
Chery-backed AiMOGA has said it sees demand in the Middle East and Southeast Asia. Reuters reported that the company had deployed 110 humanoid police robots across multiple Chinese cities and had delivered more than 3,000 robots globally, including 2,000 overseas, across more than 60 countries and regions. 17
That does not mean a Chinese traffic robot can be transferred directly into another country’s police system. The more realistic export model is a public-service or safety robot with tightly limited authority, local supervision and clear escalation procedures.
A buyer would need to define what the robot is legally allowed to observe, say and record. It would also need to assign responsibility for collisions, incorrect warnings, missed hazards, bad directions and failures to escalate emergencies. In most practical deployments, the robot would need to remain advisory while human officers retain enforcement authority.
Cameras and radar can capture information about people, vehicles and movement even when facial identification is not the product’s stated purpose. An overseas deployment would therefore need clear rules for data minimization, public notice, retention, deletion, access, auditing and human review. Cross-border data transfers and remote system access could receive additional scrutiny.
A connected robot operating near police networks, traffic signals or other critical systems would need strong controls around authentication, software updates, remote access, supply-chain security and incident response. Buyers may also require security testing and detailed visibility into how data and updates are handled.
Export versions may need to meet local requirements for machinery safety, radio equipment, batteries, electromagnetic compatibility, public-space mobility and cybersecurity. The robot would also need to understand local road markings, signal conventions, languages, accessibility expectations and traffic behaviour.
The robot’s value depends on whether it can safely connect to traffic-signal controllers, dispatch systems, mapping, communications and approved video-management platforms. Those interfaces may be fragmented, proprietary or restricted for security reasons.
A humanoid form can make a machine easier to notice, but a visible camera-equipped robot may also raise concerns about surveillance, bias and overcollection. Public agencies would need transparent notices, auditable alerts, clear limits on identification and a human decision-maker for consequential actions.
The SUPCON T2 shows how humanoid-style machines can move from demonstrations into narrow municipal roles. Its strongest near-term use is not replacing traffic police, but handling repetitive observation, guidance, warnings and public assistance while human officers retain authority.
For overseas buyers, the central question is not whether a robot can look and gesture like an officer. It is whether the system can operate safely, lawfully and transparently inside local traffic, privacy, cybersecurity and emergency-response frameworks.
Studio Global AI
This page includes a source-backed answer you can continue inside Studio Global.
Hangzhou’s 15 SUPCON T2 robots have issued more than 170,000 warnings since May 2026, according to the company, but they remain traffic assistants rather than police officers: they cannot arrest people or issue fines,...
Hangzhou’s 15 SUPCON T2 robots have issued more than 170,000 warnings since May 2026, according to the company, but they remain traffic assistants rather than police officers: they cannot arrest people or issue fines,... The wheeled, 1.88 metre, 98 kilogram T2 uses cameras, radar and onboard computing to detect helmetless e bike riding, extra passengers, stop line violations and pedestrians crossing against signals.
The same model can synchronize arm gestures with traffic lights, answer questions about directions and parking, issue spoken warnings and connect emergencies to human police.