Physix FrontierXiaomi Auto Pengcheng N90 MAX Revealed with Over 3-Meter Wheelbase
Let's first sort out the hard parameters from the news. A wheelbase exceeding 3 meters means this is a large SUV, likely targeting the Li Auto L9 or AITO M9 segment. Xiaomi SU7 has been on the market for just over half a year, and the second car goes straight to a full-size SUV, with "Pengcheng" in the name—likely indicating an extended-range powertrain solution with an internal combustion engine. For production lines, extended-range is more complex than pure electric: it adds an engine, fuel tank, exhaust system, and thermal management modules. If Xiaomi plans to produce both pure electric and extended-range versions on the same line, the welding, painting, and assembly sections all need flexible modifications, and the investment isn't linear growth—it's a jump.
Short Term: Technical Difficulty and Cost Pressure of Production Line Modification Are Unavoidable Hurdles
Xiaomi Auto's Phase 1 factory in Yizhuang, Beijing, has a planned annual capacity of 150,000 units and currently produces only the SU7. To introduce a new car with a wheelbase over 3 meters, the side body fixtures, floor positioning, and door/lid lines on the welding line all need to be replaced. Changes in wheelbase mean changes in body dimensions—is there enough space margin in existing hemming islands and laser welding stations? I've seen too many projects where R&D arbitrarily set parameters, only to find during integration that the welding gun stroke was insufficient, robot bases needed raising, and pipeline packages had to be rerouted—all burning cash.
More critical is the integration of the extended-range powertrain line. The SU7 is a pure electric platform, with technologies like integrated die-cast rear floors and CTP battery packs already locked into the process. The N90 MAX adds an engine and transmission, requiring at least 1.5x the investment of a pure electric line for powertrain assembly. Where does the engine come from? If Xiaomi develops it in-house, machining lines for cylinder blocks/heads and assembly/test lines could cost billions; if sourced externally, say using a mature supplier's 1.5T range extender, then docking processes, fluid filling, refrigerant cycle tests, etc., on the final assembly line all need new stations. In my experience, adding 15-20 stations to the final assembly line is very normal. At an integration cost of 800k per station, that's 12-16 million, not counting conveyor system modifications.
One more thing: To compress inventory cycles, Xiaomi SU7 uses a "real-time scheduling" logic similar to Tesla, i.e., order-driven with few ready cars. If the N90 MAX launches both extended-range and pure electric versions simultaneously, line-side material management will become extremely complex. Mixed flow of the same body requires AGV delivery systems to identify VIN codes and automatically switch glass sunroofs, exhaust systems, and rear axle differential parts. I've seen a new EV startup factory where poor material switching logic caused mixed-flow takt time to drop from 60 JPH to 35 JPH, taking three months to recover. Xiaomi's team has limited experience in automation integration—can they handle a 3-meter wheelbase, co-production of extended-range and pure electric, while maintaining high takt times all at once? I have doubts.
Long Term: Integrated Die-Casting and Flexible Production Lines Are Key to Breakthroughs, But Depend on Scale
From a body process perspective, for cars with wheelbases over 3 meters, the ideal process is large-tonnage integrated die-cast rear floors + front engine bays. Xiaomi has already deployed 9,100-ton die-casting machines, and the SU7's rear floor is a die-cast part. The N90 MAX is larger, so the rear floor might need to be cast in two pieces and welded, or use a larger die-casting machine. I know a domestic ultra-large tonnage die-casting machine manufacturer can already build 12,000-ton equipment, but mold costs, mold life, and yield rates are unknowns. In the long run, if die-casting matures, it can save hundreds of welding robots, significantly improving takt time, and actually lowering integration costs. But how long is "long term"? At least two years of process tuning.
Another long-term variable: Xiaomi's self-built electric drive and battery pack lines. If the N90 MAX uses Xiaomi's self-developed Super Motor V8s and 800V batteries, the integration complexity of the e-drive assembly line will be lower than external sourcing because internal component interface standards are more consistent, making robot grasping poses, tightening torques, and air-tightness test programming smoother. But this assumes Xiaomi is willing to bring motor and battery capacity next to its own factory for vertical integration. Otherwise, they'll fall into the "purchased part tolerance hell"—every batch's dimensional fluctuations need compensation on the line, which integrators fear most.
Finally, a broader perspective: Xiaomi Auto's current sales volume. SU7 barely broke 10,000 monthly sales since launch, still far from the break-even point. R&D investment, mold development, and production line modifications for the N90 MAX conservatively require 8-10 billion RMB. If sales don't pick up, these production line investments will go down the drain sooner or later. Conversely, if the N90 MAX captures over 20,000 monthly sales, the fixed asset amortization pressure on the production line will decrease significantly, and integrators won't be pressured daily to cut costs.
After doing production line integration for so many years, my deepest feeling is: automakers' product definition and process planning must proceed synchronously. Many new EV startups ask us for integration solutions before even finalizing length, wheelbase, or door gap lines, demanding a plan in 3 months. Result: drawings revised three times, two sets of fixtures scrapped, ending in blame games between integrators and OEMs. Since the Xiaomi N90 MAX has already appeared in MIIT filings, tooling development should have started, but I still want to ask—
3-meter wheelbase, extended-range power, semi-hidden door handles, frameless mirrors—each of these details brings corresponding takt time losses and cost increments on the production line. Has Xiaomi's process team calculated the total bill, or are they just building it first and figuring it out later?
Original Link: https://www.ithome.com/0/975/520.htm