After the intelligent driving evasion fails, Tu Xiao'an acts as the third-level fallback for vehicle-road coordination.

The Chinese Academy of Engineering has repeatedly pointed out a reality in research and judgment related to autonomous driving: when L3/L4 technologies move from demonstration to mass production, what hinders their large-scale implementation is not "smooth operation under normal conditions", but the safety fallback mechanism in extreme scenarios. The vehicle-road-cloud integration solution realizes perception collaboration, but fails to fully address one key issue — when the perception of intelligent driving fails, collision avoidance fails and the vehicle can no longer brake, what can be relied on to protect the lives of people on board?

Tuxiao'an is precisely built to provide this "third-level safety fallback": it does not replace active safety, nor does it compete in developing large-model intelligent driving systems. Instead, when a collision is unavoidable, it uses smart flexible guardrails to convert severe impact into graded buffering, and uses vehicle-mounted perception terminals to clarify "who is in the accident, where it happens, how the collision occurs, and who should be dispatched to the site" within seconds.

A Car Crash on a Cold Winter Night Spawned a Systematic Solution

Han Yingquan, the founder, has been engaged in product design for more than 30 years. He holds a Master of Engineering degree from Tsinghua University, and has witnessed countless defects and improvement opportunities of various products. But what really inspired him to launch Tuxiao'an was a car accident he encountered when returning home for the Spring Festival.

"That winter, I drove home. When passing through a culvert, the road surface was icy, my car lost control and crashed into a traditional rigid guardrail. I was lucky enough to survive, but what happened later was really terrifying." He recalled that he called the police in the cold wind, but there was no sign in the culvert so he could not clearly state his location. He walked out for a kilometer with injuries in sub-zero temperatures, only to find the name of the culvert and call back the emergency response center. It had been more than 40 minutes before the traffic police arrived at the scene.

Standing by the roadside in the freezing snow, what came to his mind was not "I'm so unlucky", but the systematic perspective he developed over 30 years of product design experience: the traditional guardrail only blocks impacts rigidly without absorbing energy; there is no automatic perception after an accident; the location cannot be transmitted; rescue dispatch relies on verbal description from people; the whole alarm response process lacks verification. "The whole industry is studying how to avoid collisions, but no one is seriously studying — in case a collision happens, how to keep people alive."

The starting point of Tuxiao'an is exactly this sentence: Being unable to survive a collision and getting no timely rescue after being injured are the real problems faced by hundreds of millions of drivers and passengers.

Three-level Closed Loop: Road, Vehicle and Personnel Each Take Their Own Responsibilities

In the traditional path, active safety is responsible for reducing the occurrence of accidents, while passive safety has long been regarded as a supporting facility for road construction. Tuxiao'an splits the system into three layers from the "human-vehicle-road" collaboration perspective:

Road Side: Flexible composite high-safety guardrail. The self-developed composite material integrates energy absorption and guidance, and the designed collision acceleration is reduced from the common 10G level in the industry to 5G level. It is upgraded for 150km/h smart expressways and accident-prone road sections, solving the problem that "the guardrail may hurt people after a collision".

Vehicle Side: Emergency collision avoidance module for autonomous driving. When active safety fails, it combines the status of roadside guardrails and vehicles to generate an optimal controlled collision angle, so that the vehicle can get multi-point buffering and reduce secondary accidents. For automakers, this type of module can be included in the Safety Case discussion and safety redundancy design.

User Side: Vehicle-mounted intelligent perception terminal (Tuxiao'an / LifeLine). It is triggered at the millisecond level after a collision, automatically reporting the position, vehicle speed, collision direction and vehicle overturn status, and connecting the traffic police, 120 and emergency contacts to the same event, to guard the 30-minute golden rescue window.

Different from companies that simply "sell guardrails", Tuxiao'an positions itself as a provider of autonomous driving safety infrastructure: while active safety works to reduce collisions, Tuxiao'an is responsible for "preventing serious consequences even if a collision occurs".

Collect Collision Timing Sequence at 5 Milliseconds, Every Real Collision Generates Scarce Data

There are countless simulation data, but real collision timing sequence data is extremely scarce.

The vehicle-mounted terminal of Tuxiao'an can collect the full cycle of collision at a frequency of 5 milliseconds: including the driver's operation and vehicle status 15 seconds before the collision, the trajectory, overturn, posture and occupant risk signals 10 seconds after the collision, covering multiple vehicle models, road conditions and weather conditions. These data are currently used for load calculation, dangerous driving analysis, and accident authenticity auxiliary judgment. After financing, 1,000 seed users will be arranged to run under real working conditions to verify data integrity, accident identification and rescue collaboration effects.

"On the surface, we are developing guardrails and automatic alarm rescue systems, but at the bottom, we are producing what the autonomous driving industry lacks — real accident evidence." Han Yingquan put it plainly: intelligent driving algorithm companies have abundant simulation long-tail data, but lack boundary samples generated from real collisions; when automakers carry out vehicle safety design, they also lack real vehicle data under the boundary conditions of local roads, ice and snow, culverts and guardrails. Tuxiao'an obtains data entry through hardware deployment, and then converts these samples into desensitized, structured and licensable data sets.

Hardware is the Admission Ticket, Private Domain Operation and Data Bring Compound Returns

The business model of Tuxiao'an can be analyzed from two layers.

In the short term, cash flow comes from two types of gross profit: the price difference of vehicle-mounted perception terminal products, and the price difference of high-safety guardrail products. Once roads are built and terminals are sold, revenue will be generated.

But what really turns this project from a "one-time transaction" into a "repurchasable, superposable, non-linearly appreciable business with expanding scale" is the flywheel design behind it:

The vehicle-mounted terminal is priced at a low level, and gross profit does not come from it. When users buy the hardware and bind the APP, they unlock a set of life-saving capabilities in critical moments — automatic alarm, precise positioning, simultaneous connection to traffic police and 120, family notification, hospitalization subsidy, ambulance fee reimbursement, and customized insurance provided by the platform. Since the core rescue functions cannot work without supporting hardware, the private domain retention rate of Tuxiao'an is significantly higher than that of ordinary vehicle-related APPs.

Once car owners join the platform, the platform can connect to daily vehicle services such as refueling, maintenance, insurance and road rescue in daily scenarios; B-end service providers can obtain high-intent traffic of "drivers whose maintenance is about to expire, who have just had an accident, or whose insurance is due for renewal". The platform earns diversion commission, service provider admission fee and membership fee, and then feeds back the profit to hardware customer acquisition and rescue cloud operation and maintenance.

The more important layer lies in the industrial side. As real collision samples accumulate, the desensitized data can be used for automakers' safety design, extreme scenario iteration of intelligent driving algorithms, and insurance risk pricing. The more users there are, the more valuable the data is; the more valuable the data is, the more willing automakers and insurance companies are to pay — which in turn reduces the cost of C-end hardware and services.

"Acquire customers with low-cost hardware, retain users with rigid demand for rescue, achieve cross-subsidy through the after-market, and realize long-term monetization through real data." Han Yingquan summarized, "Hardware is the admission ticket, private domain operation is the foundation, and real collision data is the source of subsequent compound returns."

Current Progress and Financing Plan

The core team of Tuxiao'an has run in for 5 years, and has invested a total of 6 million yuan (1.2 million yuan from government special funds, 4.8 million yuan self-raised), completed guardrail structure verification, terminal prototype and collision avoidance algorithm framework, and it is not in the PPT stage. At present, it has started communication on deployment of access kits with traffic police and 120 in many regions, and has laid out invention patents around flexible guardrail materials, collision avoidance algorithms and perception terminals.

This round is the angel round, with planned financing of 12 million yuan, releasing 10% equity, and the post-investment valuation is 120 million yuan. The funds will be used for three things: pilot test of flexible guardrails and smart expressway demonstration, small-batch mass production of vehicle-mounted terminals and deployment of 1000 seed users, structuring of real collision data sets and construction of compliance framework.

In the overlapping window of vehicle-road-cloud integration, L3/L4 safety access and digitalization of emergency rescue, Tuxiao'an has not chosen a crowded hot track, but an infrastructure layer that has been skipped by the whole industry.

Active safety is responsible for "avoiding collisions", while Tuxiao'an is responsible for "ensuring that people can be rescued even after a collision".

"There are 450 million drivers and passengers in China, and it is impossible for them to never encounter ice and snow, never pass through culverts, or never be forced to the guardrail by other vehicles. At that critical second, whether the system can provide you with a safety fallback is the difference between life and death."