Abstract:
This project details the design, development, and evaluation of the Dingo
quadruped robot, with a focus on educational applications and small-scale
automation tasks. The robot integrates mechanical, electronic, and control
systems to achieve dynamic and stable locomotion across diverse terrains. A
custom gait control algorithm based on inverse kinematics ensures smooth
leg coordination. The control system takes input from a 9-axis IMU sensor,
continuously monitoring orientation and movement to correct joint angles
and maintain stability and controllability during operation. The robot is
primarily programmed in Python, leveraging the ROS Noetic framework for
communication and control. High-level control is managed by a Raspberry
Pi 4B, while an Arduino Nano processes sensor data, including voltage and
temperature monitors, to optimize motor control and energy management.
The quadruped uses DS3240 digital servos for precise leg movements,
allowing traversal of uneven terrains with balance and stability. Sensors
enhance the robot‟s adaptability, making it suitable for tasks requiring
flexibility. Applications range from educational robotics, providing a
platform for learning control systems, to small-scale automation in fields like
agriculture, inspection, and exploration. The Dingo quadruped is designed as
a research platform, offering potential for future improvements. Enhanced
autonomy through AI-based algorithms and machine learning could improve
decision-making. Energy efficiency can be optimized with energy-efficient
motors, low-power sensors, and improved power management systems.
Advanced sensor integration, including LIDAR or stereo cameras, and force
sensors will enhance environmental awareness and stability. Further control
system enhancements will improve gait and navigation. This project builds
on the foundational work provided by the original developers of the Dingo
Quadruped platform.
