Dynamic Hexapod Control System

The hexapod in its natural habitat.

Project Overview

This project is a dynamic, 3D-printed, six-legged robot (hexapod) controlled by a XIAO ESP32-S3 Sense microcontroller. It showcases a range of advanced robotics capabilities, including fully independent 6-degrees-of-freedom (DOF) body pose control during locomotion, adaptive speed and gait parameters, and real-time FPV (First-Person View) video streaming from its onboard camera.

Control is achieved through a custom Python PySide6 GUI, facilitating TCP/UDP communication for commands and telemetry. Additionally, the robot can be controlled via the DroidPad mobile application (available on F-Droid) for Android devices, offering an alternative input method.

This system was developed as a graduation project for Göteborgsregionens Tekniska Gymnasium (GTG).

Key Features

• Dynamic Locomotion: Tripod gait with smooth, continuous motion.
• 6-DOF Body Pose Control: Full control over body position (X, Y, Z) and orientation (pitch, roll, yaw) which can be adjusted live, even while walking.
• Adaptive Parameters: Real-time adjustment of maximum speed, acceleration/deceleration, step height, and step time via the GUI.
• FPV Camera Streaming: Live video feed from the onboard camera, viewable through the Python GUI, enabling FPV operation.
• Dual Control Interfaces:
  • Comprehensive Python PySide6 GUI for detailed configuration, telemetry display, and keyboard control.
  • DroidPad mobile app (Android, via F-Droid) for joystick and button-based remote control.
• Modular Firmware & Software: Designed with distinct modules for kinematics, walk cycle, network communication, and hardware control.
• Network Discovery: The robot broadcasts its presence on the network, allowing the GUI to easily discover and connect to it.

Table of Contents

• Quick Start Guide - Get the robot up and running.
• Developer's Guide - Set up, modify, and understand the code.
• Technical Deep Dive: Kinematics & Walk Cycle - Understanding the motion system.
• Technical Deep Dive: Communication Protocol - How the GUI and robot talk.

Core Components

Hardware

• Microcontroller: Seeed Studio XIAO ESP32-S3 Sense (with Grove Shield).
• Camera: Onboard camera of the XIAO ESP32-S3 Sense.
• Servo Drivers: 2x PCA9685 I2C PWM Servo Driver boards.
• Servos:
  • 16x FiTech FT5330M (35kg-cm torque) digital servos (for most joints).
  • 2x FiTech FT5320M (20kg-cm torque) digital servos (for middle leg coxa joints).
• Power System:
  • 1x 2S (7.4V) 5200mAh 100C Hardcase RC LiPo Battery.
  • !!! BATTERY WARNING !!!
    • CRITICAL: This robot uses a high-discharge LiPo battery. LiPo batteries can be DANGEROUS if misused, over-discharged, or damaged.
    • The current voltage reading system on the robot is UNRELIABLE/NON-FUNCTIONAL.
    • You MUST periodically check the battery voltage MANUALLY using a LiPo battery checker or multimeter to prevent over-discharging (damaging) the battery. Do not let a 2S LiPo fall below ~6.4V under load, or ~7.0V resting.
    • Always use a proper LiPo balance charger.
    • Store LiPo batteries safely.
    • Handle with care. Disconnect the battery when not in use or when working on the robot.

Firmware (ESP32 - C++/Arduino)

• Manages all hardware interfaces (servos, camera).
• Implements the walk cycle, inverse kinematics, and body transformations.
• Handles TCP/UDP network communication for commands and telemetry.
• Serves the MJPEG video stream via an HTTP server.
• Manages robot state and configuration.

Control Software & Apps

• Python GUI (hexapod_gui.py):
  • Built with PySide6.
  • Provides comprehensive control over robot movement, pose, and gait parameters.
  • Displays telemetry data received from the robot.
  • Includes an MJPEG video viewer for the FPV stream.
  • Manages TCP/UDP communication via hexapod_comms_client.py.
• DroidPad Mobile App (Android):
  • Available from the F-Droid store.
  • Allows for joystick-based locomotion and button inputs.
  • (QR Code to import button layout in docs/images/droidpad_qr.png)

Quick Links

• Run the Robot (Quick Start)
• Modify the Code (Developer's Guide)
• Technical Deep Dive: Kinematics & Walk Cycle
• Technical Deep Dive: Communication Protocol

License

[License to be determined - MIT or Apache 2.0 recommended for open-source hardware/software projects.]

Acknowledgements

This project was developed as a graduation work for Göteborgsregionens Tekniska Gymnasium (GTG).