Automated Transport System in TT Scale
Commercial systems for road vehicles have fundamental limitations. The Faller Car System operates at a single speed with stopping controlled by passive magnets. Belt-driven systems run at a fixed speed with fixed spacing between vehicles. Any form of active control is either overly complex or completely absent.
The goal was to design a system from scratch with full control over speed, direction, and behavior of each individual vehicle.
MechanicsVehicles move on the road surface, while underneath there is a magnetic track with a motor-driven carrier. Permanent magnets both hold and guide the vehicle. The entire system is designed to be manufacturable at home using only 3D printing – no off-the-shelf axles or specialized components.
ElectronicsEach carrier is controlled by an ESP32-C3. Communication is handled via ESP-NOW, a lightweight protocol optimized for the ESP ecosystem without the overhead of a WiFi server. A higher-level ESP32-S3 processes inputs and distributes commands to individual vehicles.
Since a control system was already required, it was extended to rail vehicles as well. Custom locomotives replace traditional DCC decoders, using the same communication layer and unified architecture across the entire system.
Each vehicle accelerates and decelerates smoothly, without abrupt speed changes. Direction changes are handled via a controlled ramp with a neutral pause, reflecting real-world vehicle behavior.
Current StateA functional prototype is running on a test layout (quarry scenario). Vehicles handle slopes, tight curves, reversing, and Y intersections under manual control. Position detection is implemented using Hall sensors.
This is the first fully working version of the system. A key design constraint was simplicity – everything must be manufacturable at home with a standard 3D printer. No proprietary parts, no specialized components.
Dispatch LayerThe top-level system will run on a standard PC (Ubuntu). Instead of pre-programmed sequences, it will track the real-time position of each vehicle (via Hall sensors), dynamically assign routes, and resolve conflicts based on current traffic conditions.
This includes block allocation, schedule generation, signal control, and automated announcements. The system reacts to what is happening, not to a predefined script.
The goal is centralized control of three transport modes simultaneously – rail, bus, and freight – with mutual interaction. For example, a train can wait for a connecting bus, or a freight vehicle can yield based on the current schedule.
The communication chain is intentionally simple: the PC sends commands via USB to the ESP32-S3, which distributes them via ESP-NOW to individual vehicles. No cloud dependencies, no external services – everything runs locally.
Manual override is always possible. A user can take control of any vehicle at any time, while the rest of the system continues operating and adapts without interruption.






Comments