How to Build a Micromouse Robot - Mechanical, Hardware, Software
How to Build a Micromouse Robot - Mechanical, Hardware, Software
This detailed guide explains how to build a complete Micromouse robot from scratch.
Mechanical Design Considerations of Micromouse
Chassis
The chassis determines weight distribution, sensor placement, and ground clearance. Most competitive teams design a custom PCB that doubles as the chassis - this saves significant weight over acrylic or aluminium plates. Good mechanical design is critical.Key Mechanical Constraints
- Footprint: Must fit inside a 25 × 25 cm square (the maze cell opening is 16.8 cm wide)
- Ground clearance: 2?5 mm to avoid catching on wall posts
- Wheel base: 70?90 mm between drive wheel centers for stable turning
- Weight: Lighter is faster ? aim for under 100 g including battery
- Center of mass: Keep low and centered over the drive axle to minimize wheel slip during acceleration
- Weight Distribution : Aim for balanced weight near the center.
-. Stable turning-. Reduced wheel slip-. Better acceleration
- Ground Clearance
-. Too high: Unstable-. Too low: Scrapes maze floor-. Recommended: 1–3 mm clearanceDifferential Drive vs Omni
Most Micromice use differential drive (two driven wheels, one or two passive front casters). Omni-wheel designs are rare and mechanically complex. For your first build, a two-wheel differential drive with a ball caster is the standard approach.
PCB-as-Chassis Approach
Design your main PCB in KiCad or EasyEDA with mounting holes for motors and sensor mounts. Use 1.6 mm FR4 PCB material ? stiff enough to serve as structural chassis while being light. Route motor traces on copper pour to handle current without extra wiring.
Recommended Dimensions
For half-size Micromouse:
- Width: 7–9 cm
- Length: 8–12 cm
- Height: under 5 cm
Wheels and Tires
Wheel quality significantly affects performance.
Important Factors
- Grip
- Diameter consistency
- Low vibration
- Lightweight
Common Wheel Sizes
- 20–35 mm diameter
Silicone tires are widely used for excellent traction.
Hardware Design Considerations of Micromouse
Microcontroller
The microcontroller is the brain of the robot.
Common Choices
MCU Advantages STM32 Fast and powerful ESP32 Wireless support Teensy High-speed processing Arduino Beginner-friendly
Recommended
STM32 is widely used in competitive Micromouse robots due to:
- Fast ADC
- Hardware timers
- Interrupt support
- High processing speed
Sensors
Sensors are the mouse's eyes. You need to reliably detect walls in three directions: front, left, and right. Some advanced designs also use diagonal sensors for wall-following accuracy during turns.
Infrared (IR) Emitter / Detector Pairs
The classic approach uses an IR LED paired with an IR phototransistor. The maze walls reflect IR light, and the detector reading indicates distance. You need at least three pairs: front-left diagonal, front-right diagonal, and optionally a straight-ahead pair. Advantages include high speed (no I²C overhead), low latency, and low cost.
The main challenge with IR is ambient light interference. Solve this by rapidly toggling the emitter (e.g., at 10 kHz) and reading the difference between the "LED on" and "LED off" ADC values ? this cancels ambient noise.
Time-of-Flight (ToF) Sensors
VL53L0X or VL53L1X sensors use laser ranging and return absolute distance in mm over I²C. They are more accurate and immune to ambient light, but add I²C bus latency. Multiple sensors on a single I²C bus require addressing via XSHUT pins on startup.
Sensor Placement
- Diagonal IR sensors: Mount at 45° on each front corner to detect side walls as the robot enters a cell
- Front sensor: Aimed straight ahead, detects the wall at the end of the current cell before arrival
- Mounting height: Aim sensors at the midpoint of the wall height (approximately 3 cm off the ground)
Motors
A fast, precise motor system is critical. You need not just speed but exact speed control- both wheels must turn at precisely calculated rates to navigate straight lines and accurate 90° turns.
Motor Selection
Pololu N20 micro metal gearmotors are the industry standard for Micromouse. Choose a gear ratio based on your target speed. A 10:1 ratio gives a high top speed; a 30:1 gives more torque for climbing slight imperfections. For 32 mm wheels at 6 V, a 10:1 or 15:1 ratio typically reaches 1-2 m/s. Stepper Motors are less common today due to lower efficiency at high speed.
Motor Driver
The TB6612FNG or DRV8833 dual H-bridge handles both motors. Connect PWM, direction, and standby pins to your MCU. Use PWM frequencies of 10-20 kHz to avoid audible motor whine.
Quadrature Encoders
Magnetic encoders (e.g., Pololu magnetic encoder kit) attached to the motor shaft provide velocity and position feedback. Wire both A and B channels to interrupt-capable MCU pins for quadrature decoding - this gives you direction as well as count. At 512 counts per revolution with a 32 mm wheel, you get approximately 0.2 mm per count resolution.
Power System
A clean, stable power supply is crucial. Motor switching noise can corrupt sensor ADC readings and crash your MCU if not properly managed.
Power Architecture
- Battery: 2S LiPo (7.4 V nominal, 8.4 V full charge) powers the motors directly via the motor driver
- MCU rail: 3.3 V from an LDO regulator (AP2112K or similar, 600 mA) - separate from motor power
- Sensor rail: Some IR emitters may run at 3.3 V; ensure your regulator can supply adequate current
- Decoupling: Place 100 nF ceramic capacitors on every IC power pin, with a 10 μF bulk capacitor near the motor driver input
- Power switch: A physical slide switch for safe power-on before placing in the maze
Battery Management
Add a battery voltage divider to an MCU ADC pin. Implement a low-battery warning (e.g., an LED or buzzer when voltage drops below 7.0 V) to prevent LiPo damage from over-discharge. Never run a LiPo below 3.0 V per cell.
Software Design Considerations of Micromouse
Main Software Modules
| Module | Purpose |
|---|---|
| Sensor processing | Read wall data |
| Motor control | Drive motors |
| PID controller | Stabilize movement |
| Maze mapping | Store maze structure |
| Path planning | Find shortest route |
| Motion profiling | Smooth acceleration |
PID Control System
PID control stabilizes the robot.
PID Formula
Where:
- = proportional gain
- = integral gain
- = derivative gain
Wall Following
The robot continuously adjusts position using side-wall sensors.
Objectives
- Stay centered
- Reduce oscillation
- Maintain high speed
Maze Solving Algorithms
The maze-solving algorithm is the heart of Micromouse intelligence.
1. Flood Fill Algorithm
Most popular method.
How It Works
- Assign values to maze cells
- Goal has value 0
- Neighboring cells increase in value
- Robot follows lowest-value path
Advantages:
- Reliable
- Efficient
- Easy to implement
2. DFS (Depth First Search)
Explores one path fully before backtracking.
Advantages: Simple implementation
Disadvantages: Slower optimization
3. Dijkstra Algorithm
Finds shortest path mathematically.
Advantages: Accurate
Disadvantages: More computation
Motion Profiling
Competitive Micromouse robots use motion profiles for smooth acceleration.
Trapezoidal Velocity Profile
Benefits:
- Smooth movement
- Reduced wheel slip
- Better cornering
Recommended Development Tools
| Tool | Purpose |
|---|---|
| KiCad | PCB design |
| STM32CubeIDE | Firmware development |
| PlatformIO | Embedded development |
| MATLAB | Simulation |
Summary
Building a Micromouse robot is a complete engineering education in miniature. You will learn PCB design, embedded firmware, real-time control theory, and algorithm design - all in a competition format that gives you immediate feedback on every decision.
Whether you are a student, hobbyist, or professional engineer, Micromouse development provides an exciting challenge that improves both technical and problem-solving abilities.
With proper design, tuning, and testing, your Micromouse robot can successfully navigate complex mazes and compete in international robotics competitions.
Start simple: get a robot moving reliably in a straight line. Then add sensors, then mapping, then pathfinding. Each layer builds on the last. The journey from blinking LED to center-cell finish is one of the most satisfying in all of robotics.
"The best Micromouse is not the fastest motor - it is the most accurate map."
Comments
Post a Comment