Robot Arm

Mechanical Design

During the ongoing development of this arm I have gained experience with designing and building electomechanical systems. While the arm is far from complete and will likely go through notable changes to the design, I already have gained valuable that has been applied to other projects of mine.

Control System

Even though I have completed the Systems and Control theory course at the University of Guelph, which focused more on theoretical systems, this arm presented a new challenge that required me learn how to apply what I learned in class to a complex real system. While the PID loop I implemented to control J0 is basic and far from tuned, I learned a lot about PID controllers and how they apply to real systems.

Posts

Robot Arm: Fundamental Issues

Progress on the arm ultimately slowed when the project coincided with one of the busiest academic terms of my program. While J0 was successfully assembled and tested, the remaining joints—particularly J1 and J2—never progressed beyond early prototyping and printed components before the project had to be paused for exams and other priorities. Even within the completed base joint, tuning the controller proved difficult. The low mass of the partially assembled arm combined with

Robot Arm: Controlling J0

Development began with the base joint, J0, which was the first axis to be fully assembled and wired. The joint used a DC motor with an encoder to provide feedback, allowing the system to implement a closed-loop PID controller for position control. Running on the Arduino Mega at roughly a 100 Hz update rate, the controller read encoder feedback, calculated position error, and adjusted motor output to drive the joint toward its target position. Implementing this control loop pr

Robot Arm: Design So Far

This robotic arm was an early personal project aimed at building a fully 3D-printed, six degree-of-freedom manipulator as a learning platform. The arm was designed to reach roughly 60 cm from the base (J0) to the end effector and was powered by DC motors with integrated encoders that I already had access to. Control was handled by an Arduino Mega (ATmega2560), which coordinated motor control and sensor feedback while serving as the main development platform for experimenting

Justin Michaud