Session: 03-15-02: Smart Manufacturing and Robotics for the Future II

Paper Number: 165948

Defect Mitigation for Robot Arm-Based Additive Manufacturing Utilizing Intelligent Control and IoT 

This research presents an innovative approach to integrating a robotic 3D printing system with intelligent control and defect mitigation capabilities using the ViperX 300S 6-DOF robotic arm. It presents an IoT-driven intelligent system for low-cost data acquisition and processing with unsupervised machine learning (ML) techniques to develop an anomaly detection interface and enhance the reliability and precision of large-scale 3D printing. The end effector of the robot has been modified by replacing its gripper with a fused deposition hotend, allowing controlled extrusion of filament. The hotend temperature is controlled by a microcontroller, which activates a heating element through a relay system, ensuring precise thermal control. Various control strategies including conventional threshold-based switching, proportional-integral-derivative (PID) control, and reinforcement learning (RL)-based models are explored to regulate the temperature efficiently, aiming to improve stability and response time.

Beyond temperature regulation, this research investigates different methods for controlling filament extrusion, ensuring consistent material deposition. The extruder, driven by an A4988 stepper motor driver, is controlled using traditional logic, PID control, and potentially RL-based approaches to optimize filament flow. These techniques are evaluated to determine their impact on print quality, extrusion precision, and adaptability to different printing conditions. By comparing conventional and intelligent control strategies, the study aims to identify the most effective method for dynamic and adaptive extrusion regulations.

To enhance print quality further, a defect detection and mitigation system is integrated into the setup. A camera, positioned to monitor the print bed from the top view, captures real-time images of each printed layer. Using computer vision techniques, the system identifies defects or irregularities in the printed structure in real-time. Various defect detection approaches were compared, allowing a comparative analysis of their effectiveness. Upon detecting a defect, the system maps its coordinates and transmits them to the robot arm, which then moves to the affected area using forward kinematics via ROS2 and deposits filament to repair the flaw. The effectiveness of defect mitigation is analyzed across different control paradigms, providing insights into the advantages of AI-driven correction mechanisms over traditional rule-based approaches.

This research contributes to the advancement of robotic additive manufacturing by integrating intelligent control strategies for both large-scale printing and defect correction. The study explores the potential benefits of optimizing the printing process, defect detection, and adaptive repair, while also evaluating conventional methods for comparison. The findings are expected to provide a comprehensive understanding of how different control strategies impact the efficiency, accuracy, and adaptability of robotic 3D printing. The proposed system represents a step toward fully autonomous, self-correcting additive manufacturing, with potential applications in space manufacturing, repair processes, and adaptive fabrication systems.

Presenting Author: Matsive Ali University of Louisiana at Lafayette

Presenting Author Biography: Sen Liu, is an Assistant Professor at the Department of Mechanical Engineering, University of Louisiana at Lafayette. He was a postdoc scholar at Stanford University and received PhD degree from Colorado School of Mines. His research interests include metal additive manufacturing, machine learning, robotics for manufacturing, in-situ monitoring and control for manufacturing processes.

Authors:

Matsive Ali University of Louisiana at Lafayette
Blake Gassen University of Louisiana at Lafayette
Sen Liu University of Louisiana at Lafayette