• assembly inspection
• automatic inspection
• expert system for group technology
• parametric vision system
• variant vision system learning
• vision feature clustering
• vision system

The ability to automatically perform inspection and quality control in manufacturing lines, is a valuable resource, which reduces manufacturing costs and improves product quality.

This work presents a successful industrial application of machine vision technology for the automatic inspection of halfshaft in an assembly cells. The implemented system is the result of a collaboration between the Department of Civil and Industrial Engineering (DICI) of Pisa and GKN Driveline Firenze. Its aim is to allow a 360° inspection of the parts in a reliable way, reducing time and cost, and decreasing the possibility of human errors.

This dissertation, after an introduction to the problem addressed and some considerations on the state of the art in the field of industrial vision systems (respectively, chapters 1 and 2) will go into the merits of the solutions adopted, both at the hardware level (chapter 5), including an overview of the configurations taken into consideration (chapter 4), and at software level (chapter 6 and 7), to meet the requirements specified by GKN (presented in third chapter).

The peculiar elements claimed in this thesis that distinguish this system from the multitude of industrial solutions already implemented consist in its hardware modularity, the software parametricity and the adoption of support tools, such as the Expert System for halfshaft Classification, to simplify and speed up machine configuration.