Wheeled autonomous mobile robot

3D printer

Qualcomm Dragonwing™ QCS6490

Atmega328P+A4950

AidLux

ROS 2 Humble

3D Printed parts 

JY61P Accelerometer electronic gyroscope module

LDS-E110-R

Screwdriver set

Speed control for DC motors from 3 to 35V 5A

<html><body><h3></h3><figure><iframe src="https://youtu.be/KrpXISVBSSs?si=ZaZnq4vO3aeGZr-r"></iframe></figure><h3>Getting to Know mBot</h3><h3>What is mBot?</h3>mBot is a medium-sized mobile robot platform designed for intelligent robotics research and education, based on a classic two-wheel differential chassis design with a 30cm diameter. The name "mBot" embodies the concept of a "mobile robot basic platform," aiming to provide an outof-the-box experimental platform for ROS2 learning and autonomous navigation algorithm validation. <figure><img src="https://projects.arduinocontent.cc/f71ccfca-131f-41ce-b168-f9956d9e2cdc.png"/><figcaption></figcaption></figure> <figure><img src="https://projects.arduinocontent.cc/0f453f9e-bf9a-4f9a-9073-d7774571cf32.png"/><figcaption></figcaption></figure>Analogy : If we compare mBot to an autonomous vehicle, then: <ol><li>The QCS6490 development board is the brain, responsible for thinking and processing </li><li>information </li><li>The two motor driver boards are the engine and transmission, enabling movement </li><li>The JY61P gyroscope is the inertial navigation system, knowing its own attitude </li><li>The LDS-E110 LiDAR is the eyes, perceiving the surrounding environment </li><li>The CH344 serial chip is the nervous system, connecting various sensors </li><li>The battery is the fuel tank, providing power </li></ol>Core Specifications: <ol><li>Dimensions: 30cm diameter × 25cm height </li><li>Weight: Approximately 2.5kg </li><li>Drive System: Two-wheel differential drive + caster wheel </li><li>Control Core: Qualcomm QCS6490 (8-core processor, 8GB RAM) </li><li>Operating System: AidLux (with built-in Ubuntu Desktop) </li><li>Software Framework: ROS2 Humble </li><li>Sensors: LiDAR, IMU, encoder motors </li><li>Power: 12V lithium battery (approximately 2 hours runtime) </li></ol><h3>Hardware Components</h3><h3></h3><figure><img src="https://projects.arduinocontent.cc/4e81609a-4ff3-4e1f-bf38-27a2d96d54d9.png"/><figcaption></figcaption></figure>Hardware Configuration and Interface Allocation The mBot robot adopts a modular design, with all peripherals connected to the main control via serial ports expanded through the CH344 chip. To ensure fixed device names upon plugging/unplugging, we use udev rules to map serial ports to fixed names: <figure><img src="https://projects.arduinocontent.cc/cccad1b2-27f5-43dd-a517-98b7a8df93f5.png"/><figcaption></figcaption></figure> <figure><img src="https://projects.arduinocontent.cc/7ab77b40-c5fd-4168-845b-0046cc1d084c.png"/><figcaption></figcaption></figure> <figure><img src="https://projects.arduinocontent.cc/338e1cac-ec88-4f6c-adda-04934a1ff27d.png"/><figcaption></figcaption></figure><h3>Software Environment</h3>Software Requirements <ol><li>AidLux Operating System: Runs on the QCS6490 development board, based on Android underlying layer but provides a complete Ubuntu Desktop environment </li><li>ROS2 Humble: Robot operating system, provides distributed communication framework </li><li>Modern Web Browser: For accessing AidLux's built-in web development interface </li></ol><h3>Project Architecture</h3><pre>~/evobase_ws/src/ </pre><pre>├── evobase_bringup/ # Robot launch files </pre><pre>│ ├── config/ # Configuration directory </pre><pre>│ ├── launch/ # Launch file directory </pre><pre>│ ├── scripts/ # Python scripts </pre><pre>│ └── udev/ # udev rules </pre><pre>├── evobase_navigation2/ # Navigation configuration </pre><pre>│ ├── config/ # EKF configuration </pre><pre>│ ├── launch/ # Navigation launch files </pre><pre>│ ├── maps/ # Saved maps </pre><pre>│ └── param/ # Navigation parameters </pre><pre>├── lidar_odom/ # PLICP LiDAR odometry </pre><pre>│ ├── config/ # PLICP configuration </pre><pre>│ ├── launch/ # Launch files </pre><pre>│ └── src/ # Source code </pre><pre>├── bluesea-ros2/ # Blue Sea LiDAR driver </pre><pre>├── slam_toolbox/ # slam_toolbox mapping </pre><pre>└── navigation2/ # Navigation2 source code </pre><h3>Core Algorithms Explained in Depth</h3><h3>Two-Wheel Differential Kinematics — How the Robot Moves</h3>Imagine pushing a shopping cart: if you push both handles with equal force, the cart goes straight; if you push only the left handle, the cart turns right. The mBot robot's motion is similar, except it's driven by two independently controlled motors. mBot uses two-wheel differential drive , the most classic configuration in mobile robotics. Two wheels can independently control speed and direction, achieving various motions by adjusting the speed difference between the two wheels. Basic Motion Formulas: Let left wheel speed be $v_l$, right wheel speed $v_r$, and the distance between the two wheels (wheelbase) be $L = 0.3m$. The robot's overall motion can be decomposed as: <figure><img src="https://projects.arduinocontent.cc/d21a3e8a-a365-4877-804c-4b5afefae32c.png"/><figcaption></figcaption></figure>Intuitive Understanding: <ol><li>When $v_l = v_r$, both wheels have the same speed, robot moves straight forward or backward </li><li>When $v_l = -v_r$, wheels have opposite speeds, robot rotates in place (like a tank turn) </li><li>When $v_l &gt; v_r$, robot turns right (left wheel faster than right) </li><li>When $v_l &lt; v_r$, robot turns left (right wheel faster than left) </li></ol> <figure><img src="https://projects.arduinocontent.cc/1ad1a43e-407e-48dc-980b-9d41f38cd915.png"/><figcaption></figcaption></figure>This formula tells us: the greater the speed difference between the two wheels, the smaller the turning radius; the closer the two wheel speeds, the larger the turning radius. When the two wheel speeds are exactly equal, the turning radius is infinite, i.e., straight-line motion. <h3>Dead Reckoning — How the Robot Knows Where It Is</h3>If you walk with your eyes closed, you can only estimate how far you've walked and how much you've turned by sensing your own steps. This is the basic idea of dead reckoning — estimating position by integrating motion. In mBot, the dead reckoning process is as follows: <ol><li>1. Obtain actual RPM of left and right wheels from encoders </li><li>2. Calculate current linear velocity v and angular velocity ω </li><li>3. Multiply these velocities by time interval Δt to obtain small displacements </li><li>4. Accumulate these small displacements to get current position </li></ol>Mathematical Expression: Over a small time interval $\Delta t$, the robot's pose updates as: <figure><img src="https://projects.arduinocontent.cc/b25a3672-9e02-42a9-8e82-70687c5bcc92.png"/><figcaption></figcaption></figure> <figure><iframe src="https://www.youtube.com/watch?v=1W2SDhMKY38"></iframe></figure> </body></html>

mBot Robot - A Hands-on Tutorial for Building a Smart Mobile Robot with ROS2 from Scratch