A large-scale malfunction of Baidu’s Apollo Go robotaxis in Wuhan on March 31 left passengers stranded in live traffic on elevated ring roads and city highways, intensifying questions over how rapidly expanding driverless taxi fleets handle failures in complex urban conditions.

Get the latest news straight to your inbox!

Robotaxi Malfunction in Wuhan Strands Passengers in Live Traffic

System Glitch Freezes More Than 100 Robotaxis

Publicly available information from Chinese media and international outlets indicates that more than 100 autonomous taxis abruptly stopped operating across multiple districts of Wuhan on the evening of March 31. The vehicles, part of Baidu’s Apollo Go ride-hailing service, reportedly came to a halt within minutes of each other around 9 p.m. local time.

Reports describe cars freezing in place on elevated ring roads, expressways and major arterial streets, in some cases occupying center lanes while human-driven vehicles continued to flow around them. In several accounts shared with Chinese outlets and on social media, riders said their in-car display suddenly showed a system error message and urged them to remain seated while remote staff were dispatched.

Coverage by domestic technology and business publications characterizes the incident as a system-wide malfunction affecting a significant portion of the Apollo Go fleet in Wuhan, where Baidu is believed to operate hundreds of robotaxis. Police statements cited in multiple reports describe it as a rare breakdown of this scale for China’s autonomous driving programs.

Preliminary descriptions suggest that the vehicles executed an automatic safety stop when the system detected an internal fault, but the failure pattern, the geographic spread of disabled cars and the duration of the outage are still under review by relevant departments, according to news reports.

Passengers Trapped in Moving Traffic Lanes

Passengers experienced the outage not as a lab test but as a real-time breakdown in heavy evening traffic. Accounts gathered by Chinese media describe riders sitting for up to 90 minutes in stalled vehicles, some perched in the overtaking lane of elevated highways with trucks and cars streaming past on either side.

Several riders recounted hesitating to exit because of the danger of stepping directly into fast-moving traffic from the middle lane, as well as uncertainty over whether leaving the vehicle might violate safety instructions displayed on in-car screens. In some cases, passengers eventually opened doors and made their way to roadside barriers or waited for assistance from traffic personnel and accompanying Apollo Go staff.

Video clips circulating on Chinese social platforms and referenced in international coverage show long lines of immobilized robotaxis with hazard lights flashing, creating bottlenecks on ring roads and ramps. At least a few minor collisions were reported, including rear-end impacts where human-driven vehicles struck stationary robotaxis whose sudden stops were not anticipated by following drivers.

No serious injuries have been reported in connection with the incident, according to published reports citing local police and city authorities. Nonetheless, the images of unmanned vehicles stalled in live traffic with nervous riders inside have quickly become a focal point in the debate over how autonomous fleets should manage failures.

Wuhan at the Center of China’s Robotaxi Rollout

The malfunction hit in a city that has become one of China’s most prominent testbeds for fully driverless taxis. Wuhan, with a population approaching 14 million, has hosted early commercial deployments for multiple autonomous driving companies and is a flagship market for Apollo Go. Public information from Baidu and municipal announcements over recent years highlights large-scale pilots and gradually expanding service zones, including busy ring roads and business districts.

Analysts note that the incident underscores the challenge of moving from limited, highly supervised trials to citywide operations where robotaxis share crowded roads with buses, trucks, motorbikes and pedestrians. As fleets scale into the hundreds and thousands of vehicles, even rare software or network faults can rapidly manifest as visible disruptions across an entire urban corridor.

China’s regulators have encouraged controlled deployment of driverless services in select cities while maintaining safety oversight. Industry watchers suggest that the Wuhan breakdown is likely to prompt a detailed technical probe, as well as fresh scrutiny of how operators design redundancy, remote supervision and incident response plans for large fleets.

The episode also lands at a time when Chinese robotaxi firms have been seeking to showcase their technology globally, with new routes and pilot services announced in parts of the Middle East and Europe. Observers say the Wuhan outage may influence how foreign regulators and city planners evaluate similar services in their own jurisdictions.

Global Autonomous Fleets Face Parallel Safety Questions

The Wuhan malfunction joins a growing list of high-profile disruptions involving autonomous taxis in other markets. In the United States, widely reported incidents involving Cruise and Waymo vehicles in San Francisco, including blocked intersections and stoppages linked to power or connectivity problems, have already raised questions about how driverless fleets behave in edge cases.

International technology and automotive coverage has tracked these events as early stress tests of a technology still in its commercial infancy. While reported crash and injury numbers for robotaxis remain low compared with human-driven traffic overall, each visible breakdown or traffic snarl has outsized influence on public perception, particularly when images of empty cars blocking lanes circulate widely online.

Experts cited in previous analyses of autonomous driving incidents argue that regulators and operators must look beyond average safety metrics to examine how systems fail, how quickly they recover and how clearly they communicate with passengers and other road users when something goes wrong. The Wuhan event, with dozens of cars stopping almost simultaneously at night in live traffic, is now being viewed through that lens.

For travelers and commuters, the episode highlights a tension at the heart of driverless mobility: the promise of safer, more efficient transport on most days versus the risk of rare but disruptive outages that can leave riders stranded at the very moment they most rely on the system.

Implications for Travelers and Urban Mobility

For residents and visitors in cities rolling out robotaxis, the Wuhan breakdown serves as a cautionary case study in how to plan trips and assess risk. Travel advisers and mobility researchers often recommend that newcomers to a city treat autonomous taxis as a supplement, rather than a sole option, particularly when catching flights, long-distance trains or time-critical appointments.

From an urban mobility perspective, the incident raises questions about how traffic management systems, emergency responders and private operators coordinate when large numbers of autonomous vehicles fail simultaneously. Observers point to the need for clear playbooks on lane closures, passenger evacuation procedures and communication channels so that stranded travelers are not left to interpret cryptic error messages alone.

Industry analysts expect Baidu and peer companies to review their software update processes, remote intervention tools and user messaging after the Wuhan outage. Publicly available commentary from Chinese transportation experts suggests that future regulations may require more robust fail-safe behaviors, such as guiding vehicles to the nearest shoulder or low-speed lane whenever possible, rather than stopping abruptly in high-speed traffic.

As cities worldwide pursue ambitious timelines for driverless services, the images from Wuhan are likely to feature in policy debates, investor briefings and consumer discussions. For now, the sudden standstill of more than 100 robotaxis in one of China’s largest cities offers a vivid reminder that the road to fully autonomous travel is still under construction.