Robot path planning with sensor feedback for industrial applications
Date
2021-05Author
Kulkarni, Rekha G.
Advisor
Boldsaikhan, EnkhsaikhanMetadata
Show full item recordAbstract
The research goal of this study is to investigate the use of sensor feedback for robot path planning
and optimization. Robots are ideal automation tools for handling highly repetitive manufacturing
tasks with unparalleled accuracy and agility. Manufacturing tasks that can be handled by robots
include polishing, material handling, assembly, inspection, etc. The study uses ABB Robotstudio
simulation software and an existing collaborative robot equipped with a conventionally available
3D machine vision sensor. Regular digital cameras capture two-dimensional images. 3D vision
sensors can capture three-dimensional images with the depth perception. The research findings are
expected to enhance the domain knowledge and make industrial robots more collaborative and
intelligent with advanced sensor feedback.
The approach involves two scholarly tasks in achieving the research goal.The first task is
to investigate robot path planning for a painting application with the use of the ABB Robotstudio
simulation software. The aim of the path planning is to cover the entire surface of a flat panel
evenly. For the simulation purposes, a flat panel may contain arbitrary pockets and/or holes which
shouldn’t be painted. The simulation uses line sensors that assimilate proximity sensors for
detecting the geometry of the flat panel so that the robot path can be optimized in accordance with
the workpiece geometry. The second task is to investigate path planning methods along with the
machine vision feedback for the same robotic painting application. A UR-10e robot and a SICK
ranger camera are used for the second task. A Canny edge detection algorithm is studied in
conjunction with identifying the workpiece geometry.
Description
Thesis (M.S.)-- Wichita State University, College of Engineering, Dept. of Industrial, Systems, and Manufacturing Engineering