19. Final Project Requeirments¶
What does it do?¶
This project is a quadruped robot that walks using eight servos, two per leg, controlled through a web server interface. The interface allows remote control of the robot’s movements over Wi-Fi, letting users adjust direction and motion in real time. The system combines electronics, mechanical design, and programming to demonstrate coordinated robotic locomotion.
Who’s done what beforehand?¶
This Fab Final Project by Karam Khrais is similar in spirit, though his design took a more spider-like approach.
Additionally, I drew inspiration from Petoi Nybble, a commercial open-source quadruped robot, studying its servo layout and gait logic to inform my own system’s design.
What did you design?¶
I designed a custom PCB to connect the Xiao ESP32C6 to a servo shield, allowing the control of eight servos powering the robot’s four legs.
I also designed the mechanical structure, including the leg attachments and body that house all electronics and wiring.
The PCB was created in KiCad, and the robot body and legs were modeled in Fusion 360 and fabricated using 3D printing.
What sources did you use?¶
- Fab Academy archives for reference on PCB design, embedded networking, and milling workflows.
- Petoi Nybble documentation for gait pattern research and servo placement.
- Adafruit and Espressif documentation for ESP32C6 hardware and web server examples.
- My own prototype testing data from single- and dual-leg setups to refine movement sequences.
What materials and components were used?¶
- Bambu PLA filament for 3D-printed body and leg structures
- Single sided copper board for the milled PCB
- Xiao ESP32C6 microcontroller
- PCA9685 Servo Shield for multi-servo control
- Eight micro servos (SG90/MG90S)
- Jumper wires, pin headers, and connectors
Where did they come from?¶
- PLA sourced from the Bambu Lab website
- Servos and servo shield purchased from Amazon
- ESP32C6 and electronic components provided through the Fab Lab
How much did they cost?¶
The total project cost is approximately $46.44 USD, covering the ESP32C6, servos, servo shield, PCB materials, and power supply components.
The PLA filament used for printing was lab-provided and therefore excluded from the total cost calculation.
What parts and systems were made?¶
I fabricated and assembled:
- A custom-milled PCB for signal routing and testing
- 3D-printed body and leg assemblies
- A servo control system using the PCA9685 shield
- A web server interface for wireless robot control
- The software handling gait logic, network communication, and motion synchronization
What processes were used?¶
- PCB milling for custom circuit fabrication
- 3D modeling and additive manufacturing for the mechanical design
- Embedded programming on the ESP32C6 using Arduino
- Networking for Wi-Fi-based web control
- System integration to bring together mechanical, electrical, and software systems
What questions were answered?¶
The main design question was how to coordinate eight servos to mimic a natural, stable walking gait.
The answer was to sequence movements diagonally and incrementally, ensuring that balance was maintained while transitioning between steps.
What worked? What didn’t?¶
The servo shield worked extremely well for controlling all eight servos simultaneously and simplified power distribution.
Initially, attempting to run multiple servos directly from the ESP32C6 caused instability and inconsistent power delivery, which was solved after integrating the shield and optimizing the control code.
The web server interface also worked smoothly after refining the non-blocking control logic.
How was it evaluated?¶
Evaluation was done through real-time motion testing. The robot’s gait cycle, servo timing, and response to web-based input were observed and adjusted until smooth, balanced walking was achieved.
Performance was measured by how consistently the robot could walk forward and turn without losing stability or servo alignment.
What are the implications?¶
This project demonstrates how digital fabrication, embedded electronics, and networked control can merge to create fully interactive robotic systems.
It serves as a foundation for future development in legged robotics, web-based control systems, and educational robotics, showing how accessible components can be used to design complex, functional machines.