China Speed: Caresoft’s View of the Execution Model and Cost Competitiveness

This article is a reconstruction of a keynote delivered by Matthew Vachaparampil, CEO of Caresoft Global, at Automotive World 2026, focusing on “China’s execution model.” The talk itself was compelling - but what stood out even more were the depth and specificity of the data and examples contained in the slides. Because the presentation time did not allow every slide to be fully explained, AEM preserved the flow of the talk while editorially reinforcing several key slide-driven points to support reader understanding. What follows is that reconstructed version.
 


Introduction: Not documents - tear-downs and real proof
Caresoft, which collaborates with companies like Microsoft, is active in Japan together with Sanyo Trading. Broadly speaking, Caresoft’s work falls into two categories: benchmarking and cost reduction. In that context, Caresoft offers two primary products - Iceberg and Eureka. The company maintains a fleet of more than 25 of the world’s latest vehicles so customers can see them firsthand. The goal is not “benchmarking from documents,” but benchmarking you can “touch, disassemble, and verify.”
At Automotive World 2026, Caresoft’s central theme was China: why understanding China matters, what China is doing, what its execution model looks like - and further, how software and AI are being embedded into that structure. From Caresoft’s perspective, understanding China’s approach can help major Asian manufacturers, including Japan.
In short, China is changing the rhythm of competition through scale and speed.


Part 1. China’s scale: EV production and sales already dominate the world
Let’s begin with the big picture. China now sits at the center of global automotive production volume. Even China’s domestic market alone behaves like a world of its own - massive enough to set its own rhythm. EVs are where that dominance is most visible: more than 65% of the world’s EVs are produced in China. That is the starting point for understanding today’s landscape.
Even on a monthly basis, China’s EV market runs at the scale of “millions.” For example, EV sales in October 2025 were around 1.9 million units - up 8% year-over-year, down 10% month-over-month. The month-to-month fluctuations matter less than the fact that this level of volume is China’s baseline weight class.
Another point is equally striking: aside from a few models such as Tesla’s Model Y and Model 3, the top ranks of EV sales are largely filled by Chinese brands. Here is a comparison worth emphasizing: in 2025, Europe’s best-selling EV, the Volkswagen ID.4, sold fewer units than the 14th best-selling EV model in China. This is not a simple growth curve - it signals that the order of the market has already shifted.


Part 2. Share shift: electrification and total volume both move toward China
“China is rising” should be understood through numbers, not impressions. One lens is global volume share in electrified vehicles; the other is global volume share in overall automotive production.
In electrified vehicles, China grew from 21% in 2019 to 43% in 2024, becoming a pillar approaching half the global market. In overall automotive volume share, China also rose from 14% in 2019 to 27% in 2024.
What happened elsewhere? In electrified share, Europe and Japan moved downward relatively. Japan fell sharply - from 52% to 21% - and Korea also slipped from 7% to 5%. The U.S. rose modestly from 10% to 12%.
The message is straightforward: China is no longer merely a regional leader - it has become a true global player, and it is executing global expansion.


Part 3. Global expansion: echoes of Japan’s overseas plant strategy in the 1980s - 1990s
Chinese OEMs are already expanding overseas. In Brazil, BYD acquired Ford’s plant near Camacari (near Salvador). GWM also secured facilities in the Sao Paulo region - assets associated with past Mercedes operations - to expand its footprint. Thailand, with its large pickup-truck market, has seen GWM establish a factory, and BYD has also opened operations there. In Europe, BYD has opened a plant in Hungary.
This resembles how Japanese OEMs expanded into the U.S. in the 1980s and 1990s. The difference is that Japan’s expansion was driven primarily by “quality,” while China combines cost + technology + speed as a single package.







Part 4. Case 1: BYD Seagull  - not a “cheap car,” but “cheap-by-design engineering”
Last year, Caresoft brought in BYD’s Seagull, priced around $8,000. It’s small - almost like a K-car class vehicle. But the point is not simply that it’s cheap. When you look at the specifications, it becomes more disruptive.
The Seagull, with its 38.8 kWh battery (Flying Edition), claims 405 km CLTC range. Curb weight is around 1,240 kg (2,734 lb). It is front-wheel drive, with a 55 kW (75 hp) motor producing 135 Nm, and it uses BYD’s Blade LFP battery.
Even the safety and base features are difficult to dismiss: six airbags, rear disc brakes, EPB, and ESP are included. The body structure is presented as 61% high-strength steel, with 1,500 MPa+ hot-forming in key areas. In other words, this is less a “low-cost EV” than a competitiveness designed to be low-cost - and given its size and range, it could be compelling even in Japan.


Part 5. Case 2: Xiaomi YU7 - consumer-electronics speed builds a car
Xiaomi YU7 is described as a “Porsche killer.” It packs an aggressive mix of latest technologies. The price is presented around $35,000, yet the performance is striking: the AWD version is 690 hp, 0 - 100 km/h in 3.2 seconds, and a 253 km/h top speed. The RWD version is 315 hp with a 240 km/h top speed. The battery is presented as CATL NMC, with a claimed 670/750/760 km CLTC range depending on version.
The key word here is 800V high-voltage architecture - enabling claims such as adding up to 620 km in 15 minutes.
ADAS messaging points toward Level 3 intent, while the cockpit emphasizes conversational AI, large displays, and deep smartphone ecosystem integration - essentially transplanting the “consumer electronics company tempo” into the vehicle. Xiaomi is not historically an automaker; it began as a smartphone company. Yet it is building cars at that speed. That is the execution tempo China is demonstrating.


Part 6. Case 3: Yangwang - “doing the impossible” as a real product
BYD’s luxury brand Yangwang is also worth highlighting. The slogan is “Dare to Leap.” The U8 is introduced as a luxury EV emphasizing features such as emergency flotation and a claimed 960 km+ driving range. The U9 is framed with a symbolic line - “a car that jumps over potholes.”
Behind that symbolism is technology such as BYD’s DiSus suspension, an active body-control concept that uses sensors, controllers, and actuators to manage vertical motion (ride), lateral roll, and longitudinal pitch (dive/squat). The key point is that this is not merely talk - it is already embodied in real products.


Part 7. Why does the U.S. lose money on a $70,000 truck
- while China profits on an $8,000 EV?
Here is the central question: why can a $70,000 truck lose money in the U.S., while an $8,000 EV can generate profit in China?
If you answer this with “China just makes things cheaper,” you miss the core. The key is not unit price - it’s structure. China integrates system blocks - thermal management, wiring, electronics - so the cost structure falls, then locks that advantage in through vertical integration, commonization, and high-volume manufacturing. And critically, it pushes all of it with speed. The shorter the development and decision-making loop becomes, the faster cost drops - and the faster features rise.
So how was this structure built? An analogy helps.


Part 8. The metaphor: China’s Olympic strategy and its automotive strategy
To understand how China arrived here in autos, think about how it became an Olympic powerhouse. In the 1988 Seoul Olympics, China won only five gold medals. Then it made a national decision: host the 2008 Beijing Olympics and become No. 1 in gold medals. It selected events - especially individual sports like diving and gymnastics - where it could build comparative advantage, avoiding direct collision with U.S.-dominant categories.
The result: 48 gold medals in Beijing, and after that, a sustained level near 40 gold medals while competing at the top with the U.S. A similar “choose, focus, execute” logic has operated in China’s automotive rise.


Part 9. China’s “30-year plan” for autos
In the 1990s, China established joint ventures with Western and Asian OEMs: GAC in Guangzhou, GM in Shanghai, Brilliance-BMW in Shenyang, Volkswagen with FAW, and more. The market was vast, and Western, Japanese, and Korean OEMs all entered. China learned how to build cars and absorbed best practices - while the process itself served as a plan to cultivate Chinese OEMs.
In the 2000s, China faced environmental and pollution pressures, and pushed a national decision to grow EVs. At the same time, it began vertical integration, strategically tying together mining, batteries, and the lithium value chain to secure the foundation. Now, China has entered global expansion - and recently, it has been shifting the competitive axis toward software, AI, and advanced technologies.


Part 10. The formula: Imitate, Improve, Increase
China’s strategy is often described as imitate, improve, and increase. The important word is not “imitate,” but the fact that “improve” and “increase” follow at high speed. There is also a culture that enables it: an execution style of fail fast and learn - try quickly, learn quickly, and iterate without hesitation.


Part 11. Japan also “learned and perfected”
Let’s shift gears briefly to Japan. The Toyoda family started in textiles and transitioned into automobiles, visiting Detroit to study Ford’s assembly lines. But Japan’s approach was not mere copying. It deeply adapted what it learned into its own methods - and ultimately perfected manufacturing systems. The Toyota Production System (TPS) became a global benchmark.
In the U.S., the late 1980s to early 1990s saw a major theme of “learning from Japan,” with TPS, Kaizen, and TQM becoming key study targets. Japan rose through learning - and then perfecting.


Part 12. China’s structure: volume + intense competition creates learning speed
Back to China: the EV market is massive and crowded. Geely, Zeekr, BYD, AITO, Li Auto, NIO - countless players compete, with some describing the ecosystem as approaching “nearly 100” intertwined participants. Japan, by comparison, has far fewer major players.
China’s competition and survival pressure are extreme - and that accelerates learning speed. It resembles the U.S. in the 1910s and 1920s, when many brands existed until consolidation and reorganization created giants like GM. China may also move toward consolidation, but the key point is that during the process, technology adoption and execution speed can surge dramatically.







Part 13. Tech/cost example 1: cooling systems - the West/Japan keep “hoses,” China moves to “integration”
Cooling systems offer a clear example. Tesla introduced an integrated cooling architecture represented by manifolds and the octo valve. The 23MY Tesla Model Y is simplified around a single manifold structure, and China pushes further - toward integrating even the reservoir.
In one comparison, the 23MY XPeng G6 is presented with a single manifold plus reservoir integration, while a 24MY Western OEM (legacy) example still shows complexity such as 3.6 meters of hose length, 14 segments, 18 quick connects, 26 band clamps, and 48 fasteners (type 13). This structural gap is presented as leading to roughly $60 - 65 cost difference per vehicle.
Ultimately, this is not about “a few parts” - it is about structure that shapes assembly efficiency, process variation, and quality dispersion.


Part 14. Tech/cost example 2: “18-month learning” - how fast XPeng catches Tesla’s Model Y
Similar patterns appear in body structures and casting. The essence is not “it looks similar,” but that XPeng absorbed Tesla’s structural ideas (integration and simplification) into real product design within about 18 months.
In BIW figures, Tesla is presented as 333.7 kg BIW / 1,980 kg vehicle (16.8%), while XPeng is presented as 299 kg BIW / 2,095 kg vehicle (14.3%). Casting efficiency comparisons also show XPeng’s front and rear castings as lighter. Again, the key is not who did it first, but how quickly learning becomes implementation.


Part 15. AI/ADAS: transformer diffusion and the speed of DeepSeek integration
Now to vehicle software - especially AI in ADAS. The key is not “who has the best algorithm,” but how quickly a new AI approach is integrated into vehicle functions, deployed as software, and iterated through improvement loops.
Lane keeping, adaptive cruise, highway assistance, and navigation-on-autopilot (NOA/Navigation Pilot) may look similar on the surface. Internally, however, performance and scalability depend on how sensor data - camera video, radar, ultrasonic - are fused into perception, decision, and control.
A transformer-based approach can be understood as a shift toward processing complex data more holistically, and bringing situation understanding and planning into a more scalable form. Tesla began such a transition around 2021, and Chinese players quickly followed.
In 2021, GWM (Haomo) is described as entering mass production flow with transformer-based ADAS/highway pilot. In 2022, XPeng and NIO followed; in 2023, Li Auto continued the trend; and from 2024 onward, many Chinese OEMs moved into deploying NOA/navigation pilot systems based on newer AI networks.
The point is not “who was first,” but that the rhythm of learn → apply → deploy has become shorter. Software is not finished in one shot; it matures through real driving data and user feedback. The shorter the deployment cycle, the faster the feature maturity can rise.
A similar integration tempo appears in cockpits/IVI. After DeepSeek launched, within about three months Geely, Zeekr, IM Motor, Changan, and Chery integrated it into IVI - and progressed to test software releases in a relatively short timeframe. The key is not the model’s raw capability, but the execution loop itself: attach quickly, validate quickly, release the next version quickly.







Part 16. BYD “God’s Eye”: putting ADAS into low-priced cars
BYD does not treat ADAS as a “premium-only option.” The essence is pushing functions downward, rather than locking them into top trims. BYD is less a company that showcases top-tier tech in one or two halo models, and more a company that spreads segmented packages broadly across price tiers.
BYD’s ADAS lineup under the “God’s Eye” branding is described in three tiers. The entry tier uses a combination of cameras, radar, and ultrasonics - aimed at delivering highly tangible features like highway assistance and parking even in entry-level cars. It breaks the old assumption that base models must have weak ADAS.
The next tier adds higher-end sensors (e.g., LiDAR) and more compute to expand the operational domain - from highways into city driving. City environments raise complexity sharply due to pedestrians, cyclists, cut-ins, and dense intersections, so sensor and compute are upgraded together.
The top tier adds even more LiDAR and compute, with messaging that “targets Level 3.” The key is not claiming immediate full autonomy, but opening the ceiling on top packages while already spreading “usable functions” widely below.
BYD is trying to make ADAS a mass-distributed default experience. The point is not the option list, but the speed at which ADAS moves from option to baseline - and as that speed increases, the market’s standard shifts with it.


Part 17. Development speed: China 18 - 24 months vs legacy 36 - 48 months
A typical Chinese development cadence is presented as: Concept & Design (2 - 4 months), Engineering & Development (4 - 6), Prototyping (2 - 4), Validation & Testing (4 - 6), and Manufacturing ramp-up (2 - 4) - a total of 18 - 24 months.
Legacy OEMs are presented around 36 - 48 months. Li Auto’s rapid model expansion between 2021 and 2024 is cited, along with the idea that new platform development commonly turns in 1.5 - 2 years for Chinese OEMs. Competitive pressure, the “9-9-6” work culture stereotype, and - above all - cross-functional parallel execution combine to drive that pace.


Part 18. The structure of cost advantage: dollars fall from “blocks” like zonal, wiring, battery, thermal
China’s cost advantage is less about “buying cheaper parts” than about re-designing the vehicle’s major blocks. When structure is simplified at the system level - E/E architecture, wiring, battery, body, thermal - savings cascade across part count, assembly processes, inspection points, and variation-related quality costs.
Indicative deltas are presented like: Zonal architecture/electronics integration ($300+), LV wiring architecture ($200+), HV wiring and powertrain integration ($250+), battery structure/modular evolution ($200+), thermal management integration ($100+), and so on.
For example, zonal architecture consolidates electronics by vehicle zones instead of multiplying ECUs by function, which naturally changes wiring harness structure. When harness length, branching, and connector count drop, cost and assembly efficiency drop immediately. “Electronics integration” therefore affects not only ECU cost, but wiring, connectors, and assembly time.
HV wiring and powertrain integration is another lever: in EVs, complexity in HV cables, connectors, sealing, and fixation increases cost and risk. Modularizing and integrating reduces part count and labor steps, stabilizing quality.
Battery structure is similar: it’s not only cell cost, but pack housing, fastening, cooling, and module architecture - simplifying “non-cell costs” can bring major reductions. Thermal integration often shows fast effects because hose/connector/clamp-heavy layouts increase assembly time and variability; integrated designs can reduce cost while also narrowing build variation.







Part 19. Execution model: legacy 19 - 24 weeks, China 4 - 6 weeks
Caresoft works with 22 OEMs worldwide on technical optimization and cost reduction, and sees execution differences very concretely.
For legacy OEMs, a cost-reduction idea can take 19 - 24 weeks to become a decided specification, because reviews proceed sequentially: R&D, manufacturing, purchasing, and other functions take turns. Even when ideas enter early, adoption rates can remain at 10 - 20% even in favorable cases.
China aims to compress that 19 - 24 weeks down to 4 - 6 weeks through parallel execution. R&D, manufacturing, purchasing, and production engineering review simultaneously, cycling internal reviews every 2 - 3 days, categorizing feedback, running joint design reviews, and making final decisions quickly. In three Chinese OEM cases Caresoft worked with, implementation rates were cited around 35 - 50%.
This difference is not “genius.” Competitive pressure and survival pressure force speed into organizations and supply chains - and that speed becomes structurally embedded.


Part 20. Supply chain: Western hierarchy vs Chinese integrated collaboration - and BYD’s vertical integration
Traditional Western/Japanese supply chains are hierarchical - OEM → Tier 1 → Tier 2 → Tier 3 → Tier 4 - and change slowly. Ten to fifteen years ago, “scale” and alliance strategies mattered most. Now, the key is how fast technology change is absorbed and deployed.
China has shaped structures that enable fast communication and rapid supply chain rollout. BYD, especially, is strongly vertically integrated: batteries, motors/power systems, power electronics, some semiconductors, molds/tooling, interior/exterior components - all heavily internalized. As a result, BYD is presented as controlling 60 - 70% of BOM through in-house production. More importantly, that structure enables very fast change and iteration.
Commonization is also critical. If a part like a brake booster is shared across multiple OEM groups, brands, and segments, volume can jump from 100,000 units to 1 - 2 million units. Moving from 100k to 1M can reduce cost by 15 - 25%.
Together, these factors are used to present China’s total cost advantage as 30 - 40%, decomposed into elements such as: supply chain/commonization/scale (7 - 15%), fast-follower efficiency (about 7%), large-market low-margin model (about 7%), and policy support (7 - 9%).







Part 21A. ADAS trend: reducing cost through vision-centric approaches
ADAS is increasingly moving toward vision-centric strategies to reduce cost - reducing expensive sensors like LiDAR/radar and relying more on cameras and AI perception.
XPeng is described as lowering LiDAR and HD-map dependence and moving toward “pure vision” for some models after 2026. BYD aims to spread vision-centric ADAS downward without LiDAR. Huawei ADS SE similarly approaches the entry/mid segment by reducing sensor configuration to lower cost.


Part 21B. Cockpit trend: smart cockpits evolve at smartphone speed
Smart cockpits follow a similar logic. Displays, IVI, voice/AI, connectivity, and sensors combine into a personalized in-vehicle experience. Huawei/Yinwang provides an end-to-end “smart car stack” - from OS and cockpit software (Harmony-based) to IVI, connectivity, and ADAS integration - for partner OEMs. Players like Desay SV integrate cockpit domain controllers with displays/clusters/IVI as combined hardware and software offerings.
China pushes this domain with the smartphone-industry mindset: more functions at the same price, on a faster cycle - resetting user expectations and shifting the market baseline.


“It’s not the big that eat the small - it’s the fast that eat the slow”

SPEED is the only competitive advantage.
But is speed everything? No - quality, safety, and reliability matter too. China appears to be closing gaps in at least some areas.
When Japan entered the U.S., “quality” was the main vector - Toyota and Honda led with quality in the 1970s and 1980s. In the 1990s, Korea entered with “cost.” Hyundai’s early quality was poor, but strong moves such as 10-year/100,000-mile warranties changed perception, and over time quality and cost became baseline.
China arrives with software, AI, and technology - and it arrives fast. Speed is not everything, but even with everything else, you still have to be fast.
Japan does not need to copy China. Japan has strengths - quality, safety, reliability, monozukuri - and those strengths remain respected and valid. The key is combining them with software, AI, and speed:
Craftsmanship × Software × AI × Speed


 

---

Appendix: bZ3X  - Japanese engineering × Chinese execution
Toyota launched the bZ3X in China. The case is presented under the slide title “Japanese engineering × Chinese execution.” It describes a shift toward China-led R&D for model development and facelifts, new EV platforms entering production, and faster regional decision loops - echoing the Audi case in some ways.
The cockpit is combined with Huawei’s Harmony OS ecosystem, Xiaomi smart-device ecosystem integration, and localized ADAS such as Momenta. Toyota is also described as co-developing a next-generation voice assistant with Huawei and Tencent.
 

---

 

AEM(오토모티브일렉트로닉스매거진)

---