PIX Moving

There is a moment in the pitch for PIX Moving where the whole thing either clicks or it doesn't, and it goes like this: the bus pulls up to the curb, and it is not exactly a bus. It is a box on wheels that drove itself here, and depending on the hour and the demand, it is a coffee bar, or a pop-up store, or a small gym, or a library, or a lounge. The vehicle came to you, and then it became a place. This is either the future of cities or a very elaborate way to sell a shuttle, and PIX Moving has spent about a decade betting on the first one.

The company was founded in 2014 by Angelo Yu - who also goes by Chuan Yu - and here is the detail that explains almost everything else: he was trained as an architect. Architects think about space and how people move through it. They do not, generally, start from the constraint that a vehicle must have a hood, a trunk, and four seats facing forward. So when Yu looked at the car, he saw a room that happened to move, and he asked why the room couldn't be reconfigured. That question is the company.

To build reconfigurable rooms you need a flexible foundation, and PIX's foundation is the "skateboard" chassis - a flat, self-contained slab that holds the batteries, motors, steering, and braking, all controlled by wire rather than by mechanical linkage. Put the drivetrain in the floor and the space above it becomes free. You can bolt on a passenger cabin, or a retail counter, or a lounge, and the chassis underneath doesn't care. It is, more or less, a Lego baseplate for vehicles.

Printed, Not Stamped

The genuinely unusual part is how PIX makes the thing. Traditional automotive chassis are stamped out with enormous, expensive steel dies - tooling that costs a fortune and takes months, which is a large part of why building a new vehicle is a large-company sport. PIX instead uses metal 3D printing and generative design, letting an algorithm shape the structure and a printer build it. The company says this cut its R&D costs by roughly 60% compared with the traditional route. It unveiled what it billed as the automotive industry's first 3D-printed chassis at the Michelin Movin'On summit back in 2019.

Skipping the die is not a small trick. It means small runs become economical, and geometry that would be impossible to stamp becomes printable. It is the difference between a factory that makes one thing a million times and a factory that can make a thousand things a thousand times each. PIX calls the surrounding toolchain PAM and RTM - its own names for the algorithmic-modeling and realtime-manufacturing steps - which is the kind of proprietary-acronym move you'd expect, but the underlying claim is real enough: they print cars.

The Contrarian Lane

Most of the autonomy industry spent the last decade sprinting toward one prize: the robotaxi. Replace the driver, keep the car, charge per ride. PIX went sideways. It decided the interesting move was not to automate the taxi but to reimagine the vehicle - to sell the reconfigurable space itself. This is a harder story to tell in a headline, which is probably why PIX is less famous than it is interesting. But it also means PIX isn't standing in the same crowded lane as every well-funded robotaxi outfit on earth.

The flagship is the RoboBus, an electric autonomous shuttle on the PIX chassis whose body detaches and swaps so the same vehicle can be a bunch of different things across a day. There's also RoboShop and RoboVan, aimed at mobile retail and last-mile delivery. And there's the chassis business itself: PIX will sell you the skateboard - PIXKIT - so you can build your own autonomous vehicle on top, which turns PIX into something like a picks-and-shovels supplier for everyone else's autonomy experiments.

From Guizhou to Everywhere

PIX has an unusual geography. It's rooted in China - R&D and manufacturing in Guiyang, mass production now in Huzhou - but it started with Silicon Valley DNA, having gone through the SOSV/HAX hardware accelerator, and it has been global almost from the start. Its RoboBus and RoboShop have run pilots and operations across roughly 30 countries, a list that includes Italy, Japan, South Korea, the UAE, the United States, India, and Spain. The logic is that every city has the same congestion-and-space problem, so you build for the problem, not the postcode.

The money has followed, if modestly by autonomy standards. PIX went through early rounds with SOSV, then raised an $11 million Series A in 2022 led by the Japanese listed IT company TIS Inc., with participation from a Guizhou transportation design institute and SOSV. It later raised a Series B, reported around the end of 2024, to fund global commercialization. This is not the mountain of capital that robotaxi firms have burned; PIX is running a leaner, roughly 40-person operation, which is either a constraint or a discipline depending on how you feel about it.

The Boring, Important Part

The most recent moves are about making deployment unglamorous - which, in hardware, is the whole game. In October 2025 PIX signed a strategic partnership with the LiDAR maker RoboSense to integrate its sensors and push RoboBus and RoboShop toward mass deployment. The latest RoboBus reached mass production in Huzhou and started operating in Chinese cities including Yiwu, Guiyang, and Shenzhen. The company showed up at IAA Mobility 2025 in Munich, and in early 2026 ran public trial rides in Kagoshima, Japan. The stated five-year ambition is almost cartoonishly large - 100 cities, 1,000 application scenarios, more than a billion rides a year - but every one of those numbers still starts with a single chassis coming off a print bed.

What PIX is really selling, underneath the shuttles and the printers, is a reframing: that a vehicle is a space, that autonomy is a physical layer of a city rather than a feature of a car, and that the commute could be worth something instead of being time you throw away. Whether cities and operators actually pay for that is the open question. But it is a genuinely different question than most of the industry is asking, and PIX has spent ten patient, slightly weird years building the machinery to answer it.