Behind the "100 Million Milestone": SAIC is Mining a Data Goldmine with Scale!

On May 26, 2026, the National Intellectual Property Administration announced an invention patent—"A Method, Device, and Controller for Engine Diagnosis", applicant SAIC Group, publication number CN122082896A.

 

Two days later, on May 28, at the Shanghai North Bund World Living Room, SAIC delivered the first 100 millionth vehicle (IM LS9 Hyper) to Momenta CEO Cao Xudong. SAIC thereby became the first automotive group in China to exceed 100 million cumulative production and sales.

 

Notably, prior to reaching the 100 million milestone, the 99,999,999th vehicle of SAIC Group was the SAIC Volkswagen ID. ERA 9X. This model exceeded 7,000 deliveries just one month after launch, and in April, it ranked among the top 3 in the over 300,000 RMB extended-range large premium SUV segment. This is a brilliant achievement of the fusion of SAIC Volkswagen's "In China, For China" strategy and "Global Wisdom + China Speed".

 

The patent announcement and the 100 million delivery occurred in the same week; it is not entirely a coincidence.

 

What does 100 million vehicles mean? Over 70 years, nearly 4,000 vehicles produced per day on average. But what is more noteworthy is the structural change: The tens of millions added in recent years are mostly equipped with independent electronic control systems, capable of data collection and upload. They are scattered across over 170 countries and regions, driving in various real-world conditions such as plateaus, extreme cold, congestion, and high speeds. The engine operation data of every vehicle is a training sample in SAIC's R&D system; this is something no laboratory can replicate.

 

That engine diagnosis patent happens to provide a window for observation, pointing to a shifting logic: When a car manufacturer's cumulative ownership reaches the 100 million level, scale is no longer just a tool for diluting manufacturing costs, but will begin to feed back into technical R&D.

 

01 The Confidence of 100 Million Vehicles

 

To understand the technical meaning of 100 million vehicles, one must first look at how these 100 million were accumulated.

 

In 1958, Shanghai workers used hammers to produce the first "Phoenix" sedan, achieving a breakthrough from zero in Shanghai's car manufacturing. In 1983, the first Santana was assembled manually and rolled off the line, opening the era of joint venture cooperation and driving the establishment of a modern parts system. Over the following decades, SAIC deepened ties with Volkswagen, General Motors, etc., completing the original accumulation of manufacturing systems, supply chain management, and talent resources in the wave of "Market for Technology".

 

The real turning point occurred after the rise of independent brands. The Roewe brand was launched in 2006, the world's first internet car, Roewe RX5, debuted in 2016, and the premium intelligent electric brand IM Motors was established in 2020. By January to April 2026, SAIC had cumulatively sold 1.302 million vehicles, ranking first among Chinese car manufacturers for four consecutive months. Independent brand sales accounted for nearly 70%, overturning the past structure that long relied on joint ventures.

 

 

Meanwhile, SAIC's products and services have entered over 170 countries and regions, with cumulative overseas deliveries exceeding 7 million units. From the UK to Indonesia, from Thailand to Pakistan, SAIC has established 4 overseas manufacturing centers and 3 R&D innovation centers. SAIC Anji Logistics owns 42 ro-ro ships, with 8 international routes covering Southeast Asia, Europe, and the Americas.

 

Actually, more persuasive than the number itself of 100 million is the structural change. In 2015, SAIC's independent brand share was only about 38%, new energy share was almost negligible, and overseas sales share was about 10%. By the first four months of 2026, these three numbers had jumped to nearly 70%, 31.7%, and 35.3% respectively. This means that the tens of millions added to the 100 million in recent years are vehicles truly equipped with SAIC's independent electronic control systems, independent data collection architectures, and independent cloud platforms.

 

 

Over the past decade, SAIC has cumulatively invested more than 150 billion yuan in R&D, holding nearly 26,000 valid patents. Since the beginning of 2026 alone, nearly 100 new patents have been authorized. Behind these numbers is a "data-driven R&D" model that is taking shape.

 

Said plainly, SAIC's 100 million vehicles are the largest "mobile laboratory" in the world and the starting point of "data-defined technology". In the future, it will not be engineers guessing what problems users might encounter, but the operating data of users' vehicles directly telling engineers where the problem lies, what the priority is, and what the optimal solution is.

 

02 "Data Goldmine" in the Engine Diagnosis Patent

 

Returning to the patent itself, its technical solution doesn't sound flashy: Within a predetermined time period, when the conditions for both the first diagnostic strategy and the second diagnostic strategy are met simultaneously, only one fuel cut request is sent, and both diagnostic strategies are executed simultaneously within the same continuous time period where the engine is fuel-cut.

 

This effect has two layers: First, it can reduce the total number and duration of engine fuel cuts, thereby saving fuel consumption. Second, it can limit the diagnostic execution process to the shortest time period, reducing the duration of single fuel cuts, avoiding potential safety risks caused by fuel cuts lasting too long.

 

Traditional engine diagnosis adopts a "time-sharing diagnosis" logic. Simply put, it involves troubleshooting item by item. For each item checked, there is a brief fuel cut; there are many times and the duration is long. Previously, this might not have been obvious for fuel vehicles, but now hybrid vehicles frequently start and stop. Every fuel cut feels like a lurch in the car body, and power delivery lags half a beat. SAIC's patent solution is essentially merging multiple checks into a single fuel cut.

 

This solution works only if two capabilities are possessed: First, knowing which diagnostic items can be safely executed in parallel, meaning knowing which checks can be done together without issues; Second, setting the shortest and most appropriate time for each fuel cut for every engine model and usage scenario. These two capabilities both rely on sufficiently extensive real-world operating data. Laboratories may not be able to simulate all road conditions, but the real data generated by 100 million vehicles can.

 

This is what is called "scale feeding back into technology". 100 million vehicles generate data every moment. This data allows algorithm engineers to continuously optimize diagnostic strategies, and the optimized strategies are pushed to vehicles via OTA, bringing lower fuel consumption, a smoother driving experience, and more accurate fault diagnosis. Users barely notice the existence of the diagnosis, yet they genuinely enjoy the convenience brought by diagnostic optimization.

 

03 Why the Hybrid Era Needs More "Engine Understanding"

 

Against the backdrop of continuously rising penetration rates of pure electric vehicle models, a natural question arises: Why does a car manufacturer actively transforming towards electrification still invest resources to delve deep into engine diagnosis technology?

 

In 2025, SAIC launched the overseas "Glocal Strategy", and models equipped with the new HEV hybrid powertrain system will cover major global market segments. In March this year, MG held a technology day in Frankfurt, Germany, globally launching two core technologies: SolidCore semi-solid-state battery and Hybrid+ hybrid. The MG Hybrid+ hybrid technology, which balances low fuel consumption and high performance, has exceeded 20,000 units in overseas monthly sales. Meanwhile, SAIC's DMH Super Hybrid system has also been applied to multiple models including Roewe and MG, with its hybrid dedicated engine thermal efficiency exceeding 46.3%.

 

Similarly, including extended-range vehicle models, the EA211 Golden Range Extender equipped on SAIC Volkswagen ID. ERA 9X is essentially a high-efficiency engine power generation system. Its diagnostic strategy also requires precise control of fuel cut timing and duration. The ID. ERA 9X exceeded 7,000 deliveries one month after launch, indicating that user acceptance of this range-extending solution is not low, which also means the underlying engine diagnosis logic is being verified on a large scale.

 

One of the core technical challenges of hybrid vehicle models is the coordination of start-stop between the engine and the motor. Every start-stop involves the triggering and execution of diagnostic strategies. If the traditional "time-sharing diagnosis" logic is adopted, frequent fuel cut diagnostics will severely affect the smoothness of hybrid driving; users will feel inexplicable lurching or lag. At the same time, every fuel cut also means that the engine stops outputting power at that moment, and fuel economy will be discounted.

 

The characteristics of "reducing fuel cut times" and "shortening single fuel cut duration" in SAIC's patents are precisely to solve this major problem. This logic holds true for range-extender models as well. When the range extender starts, diagnostic efficiency directly relates to the smoothness of power generation.

 

From a broader perspective, internal combustion engines will not completely exit the historical stage due to electrification. Whether it is the demand for hybrid models in the European market or the preference for plug-in hybrid models by Chinese consumers, both point to the same conclusion: Engines will not disappear, but they will become more "intelligent". And what makes engines intelligent is precisely this data and diagnosis.

 

This also explains why SAIC's DMH hybrid dedicated engine could break through a 46.3% thermal efficiency. Optimizing solely at the mechanical level is actually approaching the physical limit. To break through further, one must rely on precise diagnosis and adjustment; every item depends on support from massive amounts of real-world operating data.

 

Standing at the new starting point of 100 million vehicles, the problem SAIC faces is direct: Can these massive data assets be transformed into perceptible user experience and technical barriers? The engine diagnosis patent is just a cross-section; it is small, but also deep enough. When technical competition shifts from hardware strength to algorithms and data, what determines victory or defeat may no longer be which company produced a better prototype in the laboratory, but which company's algorithm is more reliable in actual use.

 

100 million vehicles are not even the endpoint. They are 100 million data sources, 100 million units of user trust, and also the new starting point for SAIC to enter a new stage. From 1955 to 2026, the reason SAIC has lasted more than 70 years is not due to a hit vehicle model or a star technology, but the ability to understand technology and, more importantly, understand users.