Contactless Position Measurement for Highly Reflective Components

The next generation of particle accelerators, and the first generation of productive fusion reactors, will use highly-reflective metallic components with superconducting interior surfaces. These components must be arranged precisely along a line and fastened together in a procedure we call “string assembly”. The slightest amount of dirt or residue upon their polished interior surfaces will compromise their efficiency. To avoid such contamination, we would like to avoid touching the parts by hand while their interior surfaces are exposed. If we place the components on motorized stages, we can, in theory, move them into position for assembly. But we will not know how far to move the components unless we know their initial positions precisely, and the problem is: no existing computer vision system is capable of measuring the position of these highly-reflecting components with sufficient accuracy. We propose a computer vision system that will permit us to measure the location of such components to within a fraction of a millimeter by finding the edges of silhouette images. We will place a uniform, infrared backlight behind each component. We will view each component with two low-aberration, infrared cameras. Each of these stereoscopic cameras will be located within a string assembly coordinate system we set up with reference platforms, light sources, and survey cameras distributed around the perimeter of the string assembly room. We will obtain stereoscopic silhouette images of each component, and we will use these to determine the position of the component within our string assembly coordinate system. Once we know where they are, we can move the components to where they are supposed to be, check they are in the right place, and bolt them together. In Phase I, we will determine the feasibility of a silhouette-based computer vision system for string assembly. We will build prototype infrared backlights and cameras. We will produce an image analysis program that traces the outline of silhouette images and uses them to measure component position. We will set up typical string assembly components on motorized stages to test. By taking silhouette images of components, and applying our outline-tracing image analysis, we will determine whether or not such silhouette images will permit us to measure component position with an accuracy of a fraction of a millimeter. In Phase II, we will construct a full test stand in which half a dozen highly reflecting components must be brought into contact for assembly. We will set up a coordinate system, stereoscopic cameras, infrared backlights, and mount all components on stages that provide precise indication of position and rotation. Through repeated placement and measurement we will test our analysis of silhouette outline location. We will perfect this system until it is ready for application in a real-world string assembly clean room. If we are successful in Phase II, we will have a contactless position measurement system for highly reflective components that may be applied to superconducting string assembly.