RBE PhD Dissertation Defense
Sean McGovern
Continuous Coverage Motion Planning on 3D Freeform Surfaces
Monday, July 31st, 2023
10:00 AM - 11:00 AM
Zoom Link: Email smmcgovern@wpi.edu
Abstract: Many industrial robotic applications require a manipulator’s end-effector to fully cover a 3D surface region in a continuously constrained motion. The manipulator must continuously satisfy surface task constraints imposed on the end-effector while maintaining joint constraints. Surface coverage in this context is focused on employing commonly used coverage patterns onto the surface for the manipulator to follow. While substantial research has been conducted on achieving autonomous coverage on planar surfaces, the available methods for constrained coverage of 3D surfaces are limited to parametric or spline surfaces. Additionally, these methods do not adequately address coverage feasibility given both manipulator and task constraints. This indicates there is a lack of fundamental research to address the general problem: given a manipulator, a 3D surface, and task constraints, whether there exists a feasible continuous motion plan to cover the surface, and if so, how to produce a coverage path for uniform surface coverage which best satisfies task constraints. Furthermore, there is a pressing need for analysis tools that can empower a human worker, who is not a robotics expert, to compare material surface coverage results and identify critical factors that contribute to optimal production quality.
This work presents an approach to systematically determine continuous coverage feasibility and generate singularity-free manipulator paths to follow uniform coverage patterns on 3D freeform surfaces. We propose a novel method for generating a uv grid on 3D freeform surfaces which facilitates coverage feasibility checking, automatic surface coverage pattern placement, and analysis of material surface coverage. Moreover, we provide an interactive virtual environment to allow a domain expert worker, who is not a robotics expert, to simulate material surface coverage using our methods. Experimental results demonstrate the efficiency of our coverage feasibility checking algorithms and the versatility of our uv grid generation on 3D freeform surfaces for achieving uniform coverage patterns. Finally, we showcase the effectiveness of our approach in achieving uniform material coverage on 3D freeform surfaces compared to other methods through results obtained from a virtual spray painting case study.
Advisor:
Professor Jing Xiao, Worcester Polytechnic Institute (WPI)
Committee:
Professor Berk Calli, Worcester Polytechnic Institute (WPI)
Professor Emmanuel Agu, Worcester Polytechnic Institute (WPI)
Professor William Michalson, Worcester Polytechnic Institute (WPI)