Browser-Based City Simulator Runs an Arduino Uno Sketch on a Six-Wheeled Delivery Robot

Six wheels, one Arduino Uno microcontroller and an entire 3D city in a single browser tab. The authors of the project, which was presented on Habr, built a simulation with cars, pedestrians and traffic lights. The delivery robot there is controlled by a real Arduino sketch, not a toy template. The target audience is children and adults in robotics clubs.

The announcement gives few details: the technical part is in the full article on Habr. Below is what the authors claim, and what it means for educational robotics.

What is claimed

The simulated city is alive. Traffic moves along the roads, traffic lights work at intersections, and pedestrians walk on the sidewalks. The headline promises that pedestrians yield the right of way, which means the simulated people have their own behavior logic rather than a scripted trajectory. The robot has six wheels, and physics is claimed for it. That implies wheels, traction and inertia affect the outcome, and sloppy code leads to a crash, not a pretty animation.

The core of the project is the "real sketch." The student writes code in the familiar Arduino C++, and it controls the virtual chassis. The announcement does not say how this is implemented. Projects like this usually emulate the AVR core in JavaScript or WebAssembly, but that is our assumption, not a fact from the publication.

Key parameters

• Platform: browser, no installation required.
• Environment: a 3D city with traffic lights, cars and pedestrians.
• Robot: a six-wheeled chassis, delivery scenario.
• Control: an Arduino Uno sketch.
• Audience: robotics clubs, children and adults.
• Open questions: license, source code, sensor set and emulation accuracy.

Why this is useful for clubs

A physical robotics classroom runs into the same constraints again and again. Boards cost money, batteries run down, motors burn out, and the marked-out field has to be stored somewhere. One track for everyone gets boring fast. A simulator removes almost all of these problems: a child can break the robot as many times as they like, and a teacher can change the scenario in a minute. The same sketch remains for the move to a real board.

The approach has a weak spot, too. A simulation always diverges from hardware: sensors are noisy, power sags, wheels have backlash. If the authors did not build in these effects, students will come away with an inflated idea of how easy it is to write a working algorithm. That is why it matters that sensor noise and motor error can be switched on.

Given all the talk about autonomous couriers and self-driving vehicles, the choice of scenario makes sense. The task "deliver the package across the city without hitting a pedestrian" clearly illustrates the basics of autonomous system behavior: perception, rules, stopping and priorities. And it needs neither an expensive test rig nor neural networks.

How to use it

Teachers should read the full article and check three things: whether the code is open, whether you can add your own maps and scenarios, and which sensors are available to the robot. If all of that is in place, the project will work for first Arduino lessons and for a hackathon inside a club. For developers of educational platforms, the architecture is worth a look: browser-based microcontroller emulation with 3D physics remains a niche but workable way to provide access to "hardware" without the hardware itself.