Autonomous Mobile Robot

Introduction

The aim of this project is to build an autonomous mobile robot capable of localization, mapping and autonomous planning in real-time. This project is still under development.

Main Components

• Raspberry Pi
• Arduino nano
• Arduino RP2040 Connect
• RP Lidar
• Full components list...

Software Details

Program Structure

Robot Controller (RaspberryPi)

Main controller of the robot, handles main logic and runs the control loop.

Control Loop [4/5 Hz]

The control loop is responsible for the continuous operation of the robot, handling perception, planning, and control tasks.

1. Perception

   1. State Prediction:
      • Retreaves odometry data from Arduino RP2040
      • Uses an Extended Kalman Filter (EKF) to predict the new state (x, y, th, v, w).
   2. Visual Odometry:
      • Processes data from the Lidar to create a local map.
      • Uses ICP to compute map transformation.
   3. State Update:
      • Refines the predicted state using visual odometry data.
   4. Mapping:
      • If mapping is enabled, expands the global map with lidar data.

2. Planning

   1. Control Commands: Depending on the control type (position or velocity), sends appropriate commands to the RP2040 to move the robot.

The loop runs continuously at a specified interval, ensuring the robot can navigate and map its environment in real-time.
[robot.py]

Motor Controller (Arduino RP2040) [50 Hz]

Handles velocity control of motors and wheel odomtery. Receives commands from the Robot Controller and translates them into motor actions using a PID control. Sends encoder data to the Robot Controller each loop. The Arduino RP2040 communicates with the Raspberry Pi via UART2. [arduino code, raspberrypi driver]

Developer Console (Client)

Graphical interface to allow the end-user to remotely control the robot while receiving camera streaming and sensors feedback. Connects to RaspberryPI with multiple socket connections.