Login
Sign Up
Woofun AI reports that the 萝卜快跑 robotaxi service in Wuhan has emerged from a 60-day operational freeze, marking a critical juncture for Baidu's autonomous driving ambitions following a period of intense public scrutiny and regulatory adjustment. The suspension, which began in late March, left rows of driverless vehicles parked in remote lots under sunshades, creating a visible absence of the service that had become integral to the city's transportation landscape. While the official cause was cited as a system failure, the incident exposed deeper tensions between rapid technological deployment and public trust, necessitating a comprehensive overhaul of operational protocols before any return to service could be authorized.
Prior to the shutdown, the service had achieved significant milestones that underscored its growing maturity and market penetration. Launched in August 2022, the fleet reached a break-even point for individual vehicles in Wuhan by the second quarter of 2025, a rare feat for the industry. Baidu's Q1 earnings report released in May 2026 highlighted a surge in demand, with 萝卜快跑 completing 3.2 million orders in the first quarter, representing a year-on-year increase of over 120%. The cumulative operational data is equally impressive, with the fleet having autonomously driven over 330 million kilometers to date, including more than 220 million kilometers executed entirely without human intervention. These figures illustrate a trajectory of rapid scaling that positioned the service as a dominant force in the Wuhan market before the disruption.
The primary driver for the public's demand to resume service remains the stark economic advantage offered by the robotaxis compared to traditional transportation options. A one-way trip of 10 kilometers via 萝卜快跑 costs between 4 and 16 yuan, whereas a comparable ride in a traditional taxi ranges from 18 to 30 yuan. This price differential, combined with the consistent cleanliness, quietness, and comfort of the autonomous vehicles, created a strong user base that actively lobbied for the service's return on social media platforms.
However, the path to resumption was complicated by the aftermath of the incident, which included malicious complaints orchestrated by MCN agencies and the spread of false information regarding the technology's safety, forcing Baidu to file formal reports against these 'black armies' to protect its reputation.
As operations began to tentatively resume in other regions, the situation in Wuhan remained distinct due to the severity of the local disruption. In Guangdong, specifically in Dongguan and Shenzhen, test vehicles were spotted as early as late May, with photos of active operations appearing on social media throughout June. These resumption efforts were conducted relatively low-key, often discovered by users rather than announced through official channels. The vehicles returning to the road featured noticeable exterior updates, including large areas of blue and green paint, while retaining the "Raboo Fast Running – Safer" slogan on the front doors. Interior modifications were equally significant, with new safety instructions clearly marking the locations and usage methods of emergency pull rings and escape hammers, signaling a shift toward tangible safety enhancements rather than mere cosmetic changes.
Despite these visible improvements in other cities, the operational status in Wuhan remained ambiguous for the duration of the suspension. Although the official 萝卜快跑 APP continued to list Wuhan as an 'open city' where users could select pick-up and drop-off locations, the interface consistently displayed a message stating "No available vehicles currently." Customer service responses were uniform, confirming that no vehicles were available in the current area. This discrepancy between the app's functionality and actual service availability led to confusion, with some users reporting that even when the system indicated an order could be fulfilled within 1 minute, no vehicle arrived. In some instances, users only learned after contacting support that services were suspended city-wide due to weather conditions, prompting criticism regarding the lack of prior notice and the frustration of waiting in vain.
The resumption of service has elicited mixed reactions from different stakeholders, highlighting the complex social dynamics surrounding autonomous driving. Many passengers who missed the service emphasized its advantages, noting the absence of driver-related issues such as rude attitudes, refusal to pick up passengers, or requests for extra fees. They praised the smooth navigation through Wuhan's challenging traffic conditions and the superior comfort of the clean, spacious vehicles. Conversely, some traditional taxi drivers argued that the robotaxis' limited capabilities in complex road conditions negatively impacted traffic efficiency during peak hours, with some claiming that traffic had actually improved since the service was suspended. Other passengers who experienced the previous service interruptions expressed skepticism, recalling frightening incidents where vehicles became stuck on elevated roads for extended periods and received only a 50% discount coupon as compensation, leaving them wary of a full return.
The operational landscape across China's major cities remains fragmented as of mid-June. While cities like Suzhou, Chongqing, Chengdu, and Shenzhen have resumed normal operations, others listed as "open" in the official APP, including Nanjing, Hefei, Ezhou, and Jiaxing, still do not offer the service. This uneven rollout suggests a cautious approach to scaling, prioritizing stability in specific markets over rapid expansion. The situation in Wuhan, as the most affected city, appears to face a more stringent approval process, with reports indicating that while vehicles remain parked in remote lots, preparations for resumption may be underway in nearby areas like Gegdian Town, just dozens of kilometers away.
Parallel to these operational shifts, Baidu's Intelligent Driving Group (IDG) underwent significant internal personnel adjustments in mid-June that signal a strategic pivot. Li Yuxin, formerly the leader of Baidu's intelligent entity business and the Xinxiang App, was transferred to IDG to take charge of the 萝卜快跑 product line, reporting directly to Wang Yunpeng, the Vice President of Baidu Group and President of IDG. Li Yuxin brings a background in product implementation and business integration, having previously led a team to develop and launch the general AI intelligent entity product, the Xinxiang App, within about 30 days. His appointment suggests a clearer role definition for the product head, focusing on transforming requirements into actual products and managing business lines effectively.
Simultaneously, Chen Jingkai, a technical expert who had worked at Baidu for 24 years, announced his resignation. Chen had been instrumental in the development of core technologies including search, advertising, and mapping, and played a key role in the early stages of Baidu's autonomous driving efforts, particularly within the Apollo project. His departure coincides with a period where the company is placing greater emphasis on operational stability and compliance with updated safety regulations for L4 technology. This leadership transition marks a shift from a focus on technical performance metrics like "cumulative test mileage" and "takeover rate" to a strategy driven by commercial viability and positive gross margins under Wang Yunpeng's leadership.
The regulatory environment has also tightened significantly, influencing the strategic direction of 萝卜快跑. On June 16, the compilation of the mandatory national standard "Safety Requirements for Autonomous Driving Systems of Intelligent Connected Vehicles" was completed. The new regulations permit L4 systems to use "remote assistance" mechanisms but explicitly state that autonomous driving systems should not rely on remote assistance to perform dynamic driving tasks. Crucially, the regulations mandate that vehicles must handle risk scenarios such as "shutdowns" independently, relying on local systems to perform safe control actions like stopping safely under the least risky conditions when communication is unstable or remote assistance fails. This requirement ensures that vehicles do not lose control completely due to a lack of backend connection.
Furthermore, the draft regulations stipulate that autonomous driving systems should not arbitrarily exit autonomous driving mode when the vehicle is not stationary, unless they can safely switch to predefined states such as automatic parking or safe stopping. This provision aims to prevent complex situations from being left to passengers, traffic police, or remote customer service to handle, thereby reducing safety risks in emergencies. These comprehensive material review requirements compel car manufacturers to submit systematic explanations, including risk identification and fault handling logic, proving that their systems can handle various abnormal situations safely rather than performing well only under ideal test conditions.
Technologically, Baidu is aligning its hardware strategy with these new regulatory and commercial demands. Robin Li previously indicated that Apollo Go would evolve toward a "pure vision" approach, relying more heavily on cameras for environmental perception. From a cost perspective, the price of the sixth-generation driverless vehicle, the RT6, has dropped to around 200,000 yuan, a significant reduction from the previous generation. This cost decrease is partly attributed to the elimination of certain sensors, such as lidars and millimeter-wave radars, to lower the overall hardware cost. Pure vision systems utilize algorithms to determine distances, speeds, and spatial relationships, making them more dependent on lighting conditions and the quality of training data compared to radar systems, which offer more direct and stable performance in complex weather and lighting conditions.
The combination of new safety mandates and a shift toward pure vision technology presents both challenges and opportunities for the Robotaxi industry. Companies like 萝卜快跑 must now demonstrate that their vehicles can operate safely without backend support, a requirement that will be critical in regaining public trust. The new regulations introduce a rigorous review process that goes beyond ideal test conditions, demanding proof of robustness in real-world scenarios. For Wuhan residents, the desire for the return of the service is not merely about convenience or cost savings but about accessing a reliable system that will not stop operating in the event of problems. As Baidu navigates these regulatory and operational complexities, the focus has clearly shifted from rapid expansion to ensuring stable, compliant, and commercially viable operations in a limited number of cities. This marks a definitive transition in the company's approach to autonomous driving, prioritizing long-term sustainability over short-term growth metrics.