Micromouse

Design a micromouse PCB from scratch and implement maze floodfill algorithms on it!

Table of Contents

• Micromouse
  • Table of Contents
  • About The Project
  • Project Workflow
  • Hardware Used
  • Perfboard Testing
  • PCB Model
  • CAD Models
    • Features
  • Floodfill Algorithm Simulation
  • Software Used
  • Future Work
  • Contributors
  • Resources
  • Acknowledgements
  • License

About The Project

The aim of this project is to design and assemble a full size micromouse which will use floodfill and other path finding algorithms to solve a maze.

The dimensions of our full-size micromouse is 11.5 x 10 cm

Project Workflow

Research

• Learning and understanding Embedded C by referring to the firmware used in SRA WallE and MazeBlaze
• Understanding what Micromouse is and how maze solving algorithms like floodfill actually work
• Deciding the constraints for our micrmouse and selecting hardware and electronic componenets accordingly. We did a detailed review of existing micromouse designs and features and arrived at our present design.

CAD and PCB Design

• Designing a CAD Model on Onshape.com. We went through various iterations and design options. After weighing the pro's and cons of a 4 motor design, we settled for a 2 motor design with a suction fan in the centre.
• PCB Design keeping mechanical design in mind during routing. Once the pcb design and routing is finalised, the order for the PCB is placed.

Perfboard Testing and PID

• Testing our pcb schematic on a perfboard and recreating the final micromouse design on a perfboard.
• Implementing PID on the perfboard Micromouse to refine the code and understand the implementation of control systems.

PCB Soldering and Assembly

• Once the PCB is delivered, we will solder the components onto the PCB and assemble the bot with the 3D-printed mounts
• After this, the testing of the final Micromouse in a 6ft x 6ft maze will begin.

Hardware Used

Components	Description
ESP32 WROOM-E	Microcontroller
VL 5310X	ToF sensor4
DRV8833	Motor Driver
N12 600 RPM With Encoders	Motors
IR Sensors	Infrared Emitter-Receiver Pair

Perfboard Testing

PCB Model

• View of the PCB Model front and back:

• View of the PCB Routing - front and back:

CAD Models

• First design iteration was a four motor design with a suction fan. However this design was not suitable for fitting an enitre Esp-Devkit.

• Second and final design is a two-motor design with a suction fan. Certain drawbacks of the four-motor design was addressed in this design.

Features

• Suction Fan to improve traction and potentially turn at faster speeds without flipping over.
• Custom mount to hold the motors, shaft and suction fan in place and reduce the distance between the plate and the ground.
• Gears to run 4 wheels on two motors

Floodfill Algorithm Simulation

The path solving algorithm floodfill has been successfully implemented in the software simulation.
Floodfill simulation

Software Used

• Mechanical Design - Onshape
• PCB Design - KiCad 8.0.4
• Embedded C and Firmware - ESP-IDF

Future Work

• Implementing the flood-fill algorithm on it
• Optimizing the bot to achieve high speeds and taking turns as diagonals

Contributors

• Ojas Patil
• Varun Nigudkar
• Vishal Mutha

Resources

• UCLA Micromouse Playlist
• Mushak Github Repository
• Mackorone MMS Github Repository to run and test the floodfill simulation.

Acknowledgements

• SRA VJTI Eklavya 2024
• Special thanks to our mentors Atharva Atre, Suraj Sonawane, and all the seniors at SRA, VJTI for their constant support and guidance throughout the project.

License

MIT License