Best Paper, ISCIE/ASME International Symposium on Flexible Automation

• “Agile Surface Inspection Framework for Aerospace Components Using Unsupervised Machine Learning,” Arun Nandagopal, Abhishek Kulkarni, Colin Acton, Krithika Manohar, and Xu Chen, Japan Institute of Systems, Control, and Information Engineers (ISCIE) / ASME International Symposium on Flexible Automation

It’s been almost six years since we started working on robotic high-precision inspection. There have been so many exciting results (and unexpected happy surprises) along the path. This segmentation result takes in complex geometries, performs optimization to assure full-surface coverage, and provides occlusion-free image capture. The processes are strategically designed to utilize the optimal number of imaging locations, make in-focus image acquisitions, and maintain their applicability across different robots.

Method-wise, the developed camera-parameter-based mesh segmentation divides the inspection area into manageable sectors based on camera parameters; a ray-tracing viewpoint placement helps determine the best positions for imaging; and our robot-agnostic viewpoint planning allows the system to be versatile and applicable across different robotic platforms.

Way to go! More to come!