On September 28, China Changan Automobile Group signed 6 major projects in Chongqing in one go, and the group headquarters office space was also commissioned on the same day. The signing targets range from Liangjiang New Area, Shapingba, Hechuan all the way to CCB, CMB, Chongqing Yufu. It sounds like routine business, but when you lay out the list, you will discover that it is not that simple.

Chenzhi Group signed for headlamps, electronics/electricals, and battery packs with the three districts respectively, aiming for a trillion-yuan level parts platform. The underlying logic of this is not "building more factories", but transaction costs — Chongqing already has 19 whole vehicle enterprises and over 1200 registered above-scale parts enterprises, local matching rate stable above 45%, the shorter the transport distance, the lower the friction costs of inventory and model changes. The true dividend of industrial clusters has never been subsidies, but compressing "who to find to do it, how long to arrive". Changan was able to pull intelligent driving all the way down to the 130,000-level Q06, relying precisely on this flattened cost curve.
The group jointly initiated an industrial fund with Yufu, explicitly investing in automotive-grade chips, new automotive materials, and digital intelligent technology — the fund does not build cars, it creates "supply", burying future tight links early into the local area. On the other hand, Changan Minsheng Logistics signed a memorandum with the Port Logistics Office to equip the "Chongqing Vehicles Going Global" with a supply chain: parts are consolidated and packaged at Guoyuan Port before export, saving 40% of logistics costs compared to whole vehicle transport; KD parts go via the New Land-Sea Corridor in the West to the Thailand factory, reducing time and costs by 30% and 20% respectively. Changing from whole vehicle export to capacity export, rewriting the accounts of tariffs, freight, and rhythm simultaneously.

Wanyou Auto signed with Yuzhong District to build and operate an automotive new media center. A central state-owned enterprise writing "content creation" into an industrial agreement is rare in the automotive industry. Usually, marketing is an expense, once spent it ends; becoming a standing institution, it moves from expense item to asset item — content can be reused, can be accumulated, and can be allocated to each vehicle. When parameters from various manufacturers cannot be distinguished, whoever can continuously produce credible content at low cost holds the pricing power of customer acquisition costs.
Tianshu Navigation adopts a one-stage end-to-end approach, compressing perception, prediction, and planning into one network, equipped with 560 TOPS chips and 256-line LiDAR, relying on 10Hz high-frequency inference for redundancy, identifying earlier than human eyes in low light, backlight, and rain/fog.
However, the extended-range version uses 400V plus 3C, not that 800V set; the entry-level 602 Yuexiang Edition is also not equipped with LiDAR and Tianshu Navigation. In-house development is never cheap, intelligence technology invested over 60 billion over ten years, a team of 7,500 people, these fixed costs ultimately need to be amortized bit by bit through scale.

There has always been another voice in the industry: intelligent driving solutions can be bought externally, why carry this cost yourself. There is no standard answer to this question, but Q06 turned it into an arithmetic problem — once the installation volume goes up, the marginal cost of in-house development will approach zero, while procurement will always have to pay that premium.
This new central state-owned enterprise, established in July last year with registered capital of 20 billion yuan, has achieved full brand revenue of 359.8 billion yuan, there is capital to amortize. So the bet of this signing is clear: rely on localization to press costs, rely on in-house development to grasp technology, rely on scale to recoup investment, if one of the three links breaks, the game cannot be played alive.