Xiaomi car accident, smart driving trembles!

On March 29, 2025, a serious traffic accident occurred on the Chiqi section of the Anhui Deshang Expressway, where a Xiaomi SU7 standard version collided with a guardrail at 97 km/h and caught fire, resulting in the deaths of three occupants.

The incident has attracted widespread attention online. This article outlines the background of the "Xiaomi SU7 highway accident," how the incident and public opinion have evolved, what responses have been made by various parties, how the market reacted today, and what this means for the intelligent driving industry.

What Happened?

• On the night of March 29, a Xiaomi SU7 standard version was involved in a serious traffic accident on a highway in Tongling, Anhui.

According to "Zhengzai News," at 10 PM on March 29, three passengers were driving from Hubei to Anhui to take an exam in a Xiaomi SU7 when the accident occurred on the Deshang Expressway from Zongyang to Qimen. According to the Tongling City Zongyang County Government Office, the accident resulted in three fatalities.

According to the official Xiaomi Automobile website, the Xiaomi SU7 standard version is equipped with Xiaomi Pilot Pro for intelligent driving, featuring one Orin chip with a computing power of 84 TOPS, no lidar, and one millimeter-wave radar, which are significantly different from the SU7 Pro and SU7 Max.

• Until the 31st, the incident continued to escalate online.

Some netizens claiming to be relatives of the passengers stated: "The vehicle collided with At the same time, Xiaomi stated: "After the collision occurred, we immediately contacted the car owner to understand that it was not the owner driving. We also called for emergency rescue for the passengers in the car, reported to the police, and called 120 emergency services."

• Family Statements

On March 29, it was reported that several family members spoke out online, stating that their daughters and nieces were driving a Xiaomi SU7 from Wuhan to Chizhou, Anhui, to participate in a civil service exam. During the journey, the vehicle collided with a guardrail, leading to a fire, resulting in the deaths of three girls in the blaze.

According to Southern Metropolis Daily, on April 1, Mr. Luo, the father of one of the victims, told a reporter, "My daughter and her classmates were burned alive; the third girl was sent to the hospital but died despite rescue efforts. I saw my daughter's body at the funeral home, and it was already carbonized."

• Tongling City Transportation Bureau

A reporter from Red Star News learned from the Tongling City Transportation Bureau that the cause of the accident is under investigation, and a special working group has been established in Tongling City for this incident. Staff from Congyang Huarui Hospital, who arrived at the scene after receiving the police report, confirmed that the emergency department of the hospital was responsible for the incident.

Four Major Public Opinion Focuses of This Incident and What It Means for the Smart Car Industry

According to comprehensive information from China Entrepreneur Magazine, the public is generally concerned about the following four issues regarding the Xiaomi car accident:

• First, whether the collision that led to the vehicle catching fire and casualties was due to improper operation by the driver or a quality issue with the Xiaomi car;
• Second, there is widespread concern about whether the doors were locked and could not be opened at the time of the accident;
• Third, how much relevance does the intelligent driving level of the standard version of the Xiaomi SU7 have in this accident;
• Fourth, there was a 2-second reaction time before the vehicle collided with the concrete guardrail; is this enough time to take action to avoid danger?

A widely noted detail in this incident is that "the vehicle was in NOA intelligent assisted driving mode, traveling at a speed of 116 km/h. The road section where the incident occurred was closed due to construction, diverting traffic to the opposite lane."

Yang Xinmiao, deputy director of the Traffic Research Institute at Tsinghua University, believes that "after the construction of highways, regular maintenance is necessary, and traffic safety issues during repair processes are a global challenge." Previously, Yang Xinmiao had worked with Beijing traffic police to establish traffic safety technical standards for road maintenance operations within the Beijing area.

"Our experience is that in these areas, the situation is very complex, and autonomous driving may not be able to handle the job. There is a term in the autonomous driving field called Corner case, which refers to rare cases, because the road is not in a normal operating state. At this time, it often requires detours or changes in direction, and relying on human judgment is very challenging. Whether autonomous driving assistance technology can handle this environment is actually questionable." " Yang Xinmiao told China Entrepreneur.

In Yang Xinmiao's memory, after the introduction of autonomous driving assistance technology in China, there were many accidents related to road construction operations. In 2016, the first fatal accident involving autonomous driving occurred in Handan, Hebei, also due to autonomous driving colliding with a road cleaning vehicle. This itself is a difficult problem that has not been solved well worldwide.

In addition, Yang Xinmiao stated that although the country has unified standards for the safety of areas with road occupation, the execution varies greatly. This poses a disastrous test for autonomous driving companies, as the actual situations are not uniform, making it difficult for autonomous driving systems to recognize them properly.

However, in Yang Xinmiao's view, there is another possibility: "When a vehicle collides with an obstacle at a speed of 97 km/h, it is possible that the person may not survive, even though there are protections like airbags, the impact force is too great. However, this needs to be confirmed by accident and collision experts, but such situations cannot be ruled out."

Regarding the fourth focal issue, analysis indicates that there was only a 2-second gap from the time the intelligent driving mode reminded the driver to take over until the accident occurred, and only 1 second from the driver's manual takeover to the accident. At high speeds, if the driver is distracted, a few seconds are simply not enough to make an appropriate response. Nowadays, domestic new energy vehicle companies are competing on how intelligent their intelligent driving is, and the advanced intelligence showcased in various press conferences and demonstration videos has led some consumers to develop a cognitive misconception that "autonomous driving has matured."

With the Xiaomi car accident igniting public opinion, the safety issues of intelligent driving have also sparked new reflections within the industry: Are we moving too fast with intelligent driving?

On March 29, the annual China Electric Vehicle 100 People Forum was held in Beijing. Almost all the automotive leaders present mentioned the same thing in their speeches: intelligent driving.

According to the "Prism" public account, when intelligent driving becomes standard, under the slogan of universal intelligent driving, what users actually apply is mostly just NOA assisted driving on highways, far from true high-level intelligent driving.

Intelligent driving is divided into five levels from low to high: L1-L5, where L1 and L2 can only be called assisted driving because they require human takeover at any time. L3 is partial autonomous driving, where vehicles can drive autonomously on closed highways and some urban roads, but human intervention is still needed in emergencies. L4 and L5 are fully autonomous driving.

From the user's perspective, the most commonly used intelligent driving functions are still L2 level features such as lane keeping, automatic lane changing, and adaptive cruise control, and they are generally used on highways, with little willingness to try on urban roads. To allow users to confidently hand over their vehicles to the intelligent driving system on urban roads, automotive companies still have a long way to go and need to invest significant efforts.