A conventional car carries the memory of a driver even when no driver is present. The seats face forward. The cabin points toward a windshield. Pedals, mirrors, and a steering wheel organize the room around a job that software is supposed to inherit. Chuan Yu looks at that arrangement and sees an architectural mistake: a new kind of machine still inhabiting an old floor plan.
Yu, also known as Angelo, is the founder and CEO of PIX Moving. He is an architect by training, a robotics entrepreneur by practice, and a stubborn reframer of the automobile. His company makes drive-by-wire platforms and compact robotic vehicles whose bodies can become shuttles, shops, delivery spaces, or other small rooms in motion. The premise sounds futuristic. Its design logic is pleasantly ordinary. If the driver disappears, why should the space continue to behave as if the driver were there?
This is the distinction that gives Yu's work its shape. Much of autonomous driving begins with the car and asks how to automate it. He begins with the city and asks which spaces ought to move. One path produces a car that drives itself. The other can produce a six-seat shuttle, a retail pod, or a small electric vehicle assembled over the same underlying logic.
“When automobiles replaced horse-drawn carriages, the greatest transformation was not the vehicle itself - it was the city.”Chuan Yu
The line is less a slogan than a professional tell. Architects are trained to see circulation, use, and the relationship between people and space. Yu applies that habit to machines. For him, mobility is not only movement between two points. It is an organizing layer that decides where services appear, how much street a vehicle occupies, and whether a neighborhood must travel to a shop or a shop can travel to the neighborhood.
Before the wheels, there were wings
Yu's route into automotive hardware began far from the assembly line. He studied at Chongqing University from 1996 to 2000 and worked as an architect. An early profile says he had been interested in programming since primary school. In 2011, he won a national cloud-computing competition. A year later he entered the drone business.
The result was Hex, a small flying robot pitched on Kickstarter. Its campaign raised about $570,000, and the product reached buyers in more than 100 countries. Those figures matter, but the more revealing part is the sequence: code, architecture, a crowdfunded robot, then vehicles. Yu did not arrive in mobility through a traditional automaker. He arrived through the question of how digital ideas become physical products.
PIX Moving emerged from that earlier venture in 2014. Instead of building a familiar car around a novel sensor stack, the young company concentrated on the skateboard: the flat electric platform holding motors, batteries, steering, braking, and control systems. The form separates the expensive, safety-critical foundation from the body above it. A common base can support multiple uses without pretending every use needs the same cabin.
shuttle
shop
pod
EV
That separation creates a useful business argument. Traditional auto manufacturing earns its efficiency through repetition. A dedicated line makes enormous numbers of nearly identical things. Cities, however, contain awkward, small-batch needs: a campus route, an exhibition shuttle, a pop-up store, a short logistics loop. Yu's answer is not to beat mass production at its own game. It is to use modularity and digital manufacturing where variation is the point.
A factory made of instructions
PIX couples its modular chassis with generative design, industrial robots, and metal 3D printing. Software can propose forms under a set of constraints. Those forms become manufacturing instructions. The goal is a shorter route from a specific use case to a physical vehicle, especially when the order is too small to justify a conventional production line.
Move variation into software. Keep the costly physical base stable. Let the customer's use case determine the final layer.
Yu has described this as decentralized and participatory manufacturing. The strongest version of the idea lets a design file travel farther than a finished chassis: build closer to the place of use, reduce the tooling needed for a small run, and invite outside designers into the system. It does not make atoms behave like bits. Certification, materials, assembly, and road testing remain stubbornly physical. It does relocate some of the creative leverage upstream.
In 2018, PIX tested the social side of that theory with KuaiKai, a hackathon in Guiyang that challenged participants to build a full-sized self-driving car in seven days. The event gathered engineers, designers, and autonomous-driving enthusiasts around a deadline and a shared machine. Yu later co-authored a book chapter about it. The title sounds almost mischievous, but the exercise exposed a serious belief: a platform becomes more valuable when outsiders can understand it, alter it, and make it do something the original team did not specify.
An account from the DeAuto community places Yu at a computer on a summer night, frustrated that the vehicle on his screen looked like every other car. He noticed a copy of Eric S. Raymond's The Cathedral and the Bazaar, the classic argument for open-source development, and began asking whether a car could be created more like a shared software project. The anecdote is neat because it connects his habits: study the inherited form, find the hidden constraint, then change who is allowed to participate.
“A lot of DAO projects today are virtual, so I wonder if we can do something in the tangible, real economy.”Chuan Yu
That question led PIX to initiate DeAuto in 2022, an experiment in community-designed vehicles and robotics. The language of DAOs and NFTs belonged unmistakably to its moment. The more durable part was older than Web3: Yu wanted a bazaar of contributors around a modular physical platform. PIX's Moving Community grew to more than 300 designers and engineers across over 20 countries.
The useful tension
Hardware eventually edits every manifesto. Batteries have weight. Roads have rules. A vehicle that performs beautifully in a demo still has to be produced, maintained, insured, and placed into a real route. Yu knows the discipline this imposes. Recalling PIX's early days, when little more than a dozen people held the company together, he described a deliberate split: conservative about survival, aggressive about creation.
It is a good sentence because neither half cancels the other. Survival without creative risk yields another generic machine. Creative risk without survival yields a prototype and an auction of the office furniture. PIX's platform strategy is an attempt to resolve the tension in the product itself. Reuse the chassis. Reuse the control logic. Change the room.
By 2022, PIX said it had served roughly 200 customers in 17 countries. That year brought substantial financing and a partnership with the Japanese IT company TIS. The plan included robotic vehicles for local services and communities in Japan, with PIX supplying vehicles and operating support. Switzerland followed as another base for partnerships in Europe. The geography reflects the product: small autonomous vehicles are rarely dropped into a city alone. They require operators, regulators, route owners, technical partners, and a place willing to make the first loop work.
The RoboBus is the clearest expression of the idea. It has no conventional cockpit, pedals, or mirrors. Its symmetrical form can move in either direction. The published specification is modest by car-ad standards: six passengers, a speed range of 15 to 30 kilometers per hour, and a range of 70 to 100 kilometers. Those numbers describe a machine for bounded daily work, not a fantasy highway sprint. In 2025, the design received an iF Design Award.
At IAA Mobility that year, PIX put the RoboBus beside RoboShop and Beastie, a two-seat micro EV with a generative, biomimetic frame. The family resemblance was philosophical rather than cosmetic. Each asked what happens when a vehicle is treated as a small urban tool instead of a private living room with an engine.
After the robotaxi
Yu's recent writing pushes the frame farther. In a 2026 essay, he argued that the future of autonomy is the city, not merely the robotaxi. In another, he used the term “Silicocene” for an age in which AI stops behaving only like a tool and becomes a companion embedded in everyday systems. The word is grand; the company beneath it remains occupied with practical objects - sensor stacks, redundant controls, low floors, turning radii, charging times.
That combination is characteristic. Yu thinks in civilizations and chassis dimensions at once. He talks about mobility as living infrastructure, then builds a platform with a CAN interface so another developer can tell it what to do. He imagines “mobility cells” that sense and connect, while pursuing a concrete target of 100 cities, 1,000 communities, and more than one billion rides.
The ambition contains a real design question for cities. A robotic shop that appears on demand can add convenience, but it can also add another object competing for finite curb space. A fleet of small shuttles can fill gaps in transit, but only if routing, access, and operations make the service dependable. The value of a moving space will not be decided by its novelty. It will be decided by what the space replaces, whom it serves, and how gracefully it joins the street.
Yu's architectural instinct is useful here because architecture is never only the object. It is the object in context. PIX's machines are proposals about context: that a vehicle can be a room, that manufacturing can begin with a digital file, that one stable platform can support many city jobs, and that outside creators can help decide what gets built.
The familiar autonomous-car question is when the machine will drive as well as a person. Yu keeps a different question on the table: once it does, what deserves to move? The answer will not come from a sensor alone. It will come from watching neighborhoods, noticing small frictions, and designing the right room for the trip. That is the architect's work. Chuan Yu has simply put it on wheels.