Fast inline inspection: vision system for geometric deviation measurement and early defect detection in smart manufacturing Academic Article in Scopus uri icon

abstract

  • Part inspection is a critical activity in manufacturing because it seeks to prevent defective parts from being delivered to the end user. Inspection processes are generally carried out by a combination of gauging, offline measurements, and sampling runs aided by statistical analysis techniques. Gauging is a fast process, but it is intended to validate compliance with specs and render limited measurement information. Other full measurement processes are time-consuming. There is a trend in the industry to verify all critical dimensions of all parts without affecting production rates. Current inspection practices are not suitable for this new demand. In this work, an inline inspection process for geometric deviations of critical features was built by implementing Industry 4.0 tools (Fog and Cloud Computing, Internet of Things, and Artificial Intelligence), and computer vision. This system can collect measurement data from image analysis, compare it to GD&T specifications as parallelism and flatness, and report the results within the manufacturing process time, making information available in a timely manner, without hindering production rates and reducing cycle time by more than 90% compared to other measurement methods. Measuring cycle times are within the range of typical manufacturing processing times used in applications such as automobile production. © 2025 Informa UK Limited, trading as Taylor & Francis Group.

publication date

  • January 1, 2025