arduinoCarProject

Dijkstra Path-Following Arduino Car

What: An Elegoo/Arduino robot car that follows a shortest path computed with Dijkstra. How: Press “9” on the IR remote to enter Path Mode, then stream commands over Serial: S, moves like R3 / U2, then E. The car executes the route using time-based motion.

Repository layout arduinocarProject/ ├─ arduino/ │ ├─ DijkstraPathCar/ │ │ └─ DijkstraPathCar.ino │ └─ extras/ │ └─ PathMode/ │ └─ PathMode.ino # optional helper for factory sketches └─ python/ ├─ grid_generator.py ├─ dijkstra.py └─ bridge_send_path.py

Hardware

Arduino UNO (Elegoo UNO R3 compatible)

Motor driver: TB67H450 / TB6612 family (IN1/IN2 + PWM)

IR receiver (default pin 12)

Two DC motors with battery pack

Default motor pins in DijkstraPathCar.ino: Left PWM 5, IN1 2, IN2 4; Right PWM 6, IN1 7, IN2 8.

Software requirements

Arduino IDE (2.x recommended)

Library: IRremote by Armin Joachimsmeyer

Sketch uses #include <IRremote.hpp> (v3+). If using v2, change to #include <IRremote.h> and adjust API if needed.

Python 3.9+

Packages: matplotlib, pyserial

Install Python packages:

python3 -m pip install matplotlib pyserial

Arduino: upload the sketch

Open arduino/DijkstraPathCar/DijkstraPathCar.ino in Arduino IDE.

Tools → Board: Arduino Uno. Tools → Port: your /dev/cu.* (macOS) or COMx (Windows).

Install IRremote if prompted.

Upload. Open Serial Monitor at 115200; on reset you should see RDY.

IR key setup: Press 9 on the remote. If the sketch prints PATH:ENTER, you’re done. If not, read the RAW hex value printed for your “9” key and set it in the sketch:

#define IR_CODE_9 0xYOUR_HEX_CODE

Re-upload.

Python: generate a grid and plan a path

From the python/ directory:

Unweighted grid (all steps cost 1)

python3 grid_generator.py --max 10 --k 10 --weighted none --json grid.json

Or weighted grid (random 1..4)

python3 grid_generator.py --max 10 --k 10 --seed 1 --weighted random --wmin 1 --wmax 4 --json grid.json

Compute path to the closest 'R' point and export results

python3 dijkstra.py --grid grid.json --color R
--save-path path.json
--save-moves moves.json
--save-moves-txt moves.txt

Outputs:

path.json — waypoints ({"path":[[x0,y0],...]})

moves.json — "moves": "URRDL..."

moves.txt — the exact lines the car expects:

S U1 R3 ... E

Send the path to the car

Find your serial port (macOS example):

ls /dev/cu.* PORT="/dev/cu.usbmodemXXXX" # replace with yours

Put the car’s battery on. Press 9 on the remote to enter Path Mode. Then:

python3 bridge_send_path.py path.json --port "$PORT" --baud 115200 --handshake

Configuration (in DijkstraPathCar.ino) // IR #define IR_RECV_PIN 12 #define IR_CODE_9 0xFF52AD // replace with your remote’s RAW code for "9" #define IR_CODE_EXIT 0xFF9867 // optional exit key

// Motors #define L_PWM 5 #define L_IN1 2 #define L_IN2 4 #define R_PWM 6 #define R_IN1 7 #define R_IN2 8

// Timing (tune for your surface/grid) int DRIVE_SPEED = 180; // 0..255 int TURN_SPEED = 170; // 0..255 unsigned long CELL_MS = 600; // ms per grid cell unsigned long TURN90_MS = 350; // ms per ~90° turn

Calibration tips printf "S\nU1\nR1\nE\n" > tiny_moves.txt python3 bridge_send_path.py tiny_moves.txt --port "$PORT" --baud 115200

If distance per cell is off → adjust CELL_MS.

If turns aren’t square → adjust TURN90_MS.

If a side runs backward → swap that motor’s IN1/IN2 or reverse wiring.

Troubleshooting

No serial port → try another USB cable/port; CH340-based boards may need the CH340 driver (macOS).

Pressing 9 does nothing → print IR values in Serial Monitor, set IR_CODE_9, re-upload.

“Port busy” → close Serial Monitor before running the Python sender.

No movement on USB → motors require the battery pack.

License

Add your preferred license (MIT/Apache-2.0/etc.).