Automatically detecting surface defects from images is an essential capability in manufacturing applications. Traditional image processing techniques were useful in solving a specific class of problems. However, these techniques were unable to handle noise, variations in lighting conditions, and background with complex textures. Increasingly deep learning is being explored to automate defect detection. This survey paper presents three different ways of classifying various efforts. These are based on defect detection context, learning techniques, and defect localization and classification method. The existing literature is classified using this methodology. The paper also identifies future research directions based on the trends in the deep learning area.
Additive manufacturing (AM) technologies have been widely used to fabricate three-dimensional (3D) objects quickly and cost-effectively. However, building parts consisting of complex geometries with curvatures can be a challenging process for the traditional AM system whose capability is restricted to planar layered printing. Using six degrees-of-freedom (DOF) industrial robots for AM overcomes this limitation by allowing the material deposition to take place on nonplanar surfaces. In this paper, we present trajectory planning algorithms for 3D printing using nonplanar material deposition. Trajectory parameters are selected to avoid collision with printing surface and satisfy robot constraints. We have implemented our approach by using a 6DOF robot arm. The complex 3D structures with various curvatures were successfully fabricated with a good surface finish.
Additive manufacturing (AM) technologies have been widely used to fabricate 3D objects quickly and cost-effectively. However, building parts consisting of complex geometries with multiple curvatures can be a challenging process for the traditional AM system whose capability is restricted to planar-layered printing. Using 6-DOF industrial robots for AM overcomes this limitation by allowing materials to deposit on non-planar surfaces with desired tool orientation. In this paper, we present collision-free trajectory planning for printing using non-planar deposition. Trajectory parameters subject to surface curvature are properly controlled to avoid any collision with printing surface. We have implemented our approach by using a 6-DOF robot arm. The complex 3D structures with various curvatures were successfully fabricated, while avoiding any failures in joint movement, holding comparable build time and completing with a satisfactory surface finish.
Using a 6-DOF robotic manipulator for material extrusion additive manufacturing (AM) enables conformal printing onto complex surfaces while orienting material fibers at desired directions. Compared to a traditional 3D printer, conformal printing can produce parts with good surface finish, less number of layers and enhanced mechanical properties. In this paper, we present a three-nozzle extrusion system which can be attached to a robot. This system allows the robot to achieve conformal 3D printing with multi-resolution. One extruder is for printing support materials and the other two extruders are for fabricating structural material with different resolutions. The interior regions of the part are fabricated with a bigger diameter nozzle and the surface of the part is printed with a smaller diameter nozzle. This multi-resolution printing not only speeds up the build times but also produces the good surface finish. In previous work, we demonstrated multi-resolution 3D printing using two robots. The contributions of this paper are design and fabrication of a three-nozzle extrusion system, calibration process for the system, and its validation. The validation of the calibration was done by printing linear and circular patterns using different nozzles, and evaluating relative dimensional accuracy. This work will enable a single robot to perform multi-resolution conformal 3D printing.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.