Development of an automated 3D laser inspection system for VIN and engine number verification in automotive manufacturing
Pracharoj, Wanfasai (2026)
2026
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-202604267922
https://urn.fi/URN:NBN:fi:amk-202604267922
Tiivistelmä
This thesis presents the development of an automated 3D laser inspection system for VIN and engine number verification in a confidential automotive manufacturing environment. Before the project, inspection relied on manual rubbing, depth-gauge checking, sampling-based verification, and paper-based records. These practices resulted in long inspection times, risk of human error, weak traceability, and limited support for modern digital quality control. The aim of the project was to replace the previous method with a fully automated inspection process supported by digital data recording and process monitoring.
The developed solution combined a Keyence LJ-S080 3D laser sensor, an LJ-S8000 controller, a Keyence KV-8000 PLC, a UR10 collaborative robot for engine-line scanning, a PostgreSQL database, and a browser-based user interface. The frame-line solution used a fixed scanner with pneumatic actuation, while the engine-line solution used type-dependent robot motion to position the scanner at different stamping locations. The web interface supported data search, visualization of scanned images and measurement results, and SPC-based trend monitoring. The LJ-S080 was selected because it provides high-resolution 3D surface measurement for character recognition and dimensional evaluation, while PostgreSQL enabled reliable storage and retrieval of inspection data.
The system was evaluated through industrial validation under stabilized operating conditions using 800 units in total, comprising 400 frame-line units and 400 engine-line units. Validation was based on com-parison between automated outputs, the previous manual inspection method, and internal engineering acceptance requirements. In the final validated dataset, the developed system achieved 100% OCR ac-curacy under the tested conditions. It also reduced inspection time from approximately 7 minutes for engine-number inspection and 8 minutes for frame VIN inspection to 20 seconds per unit, improved digital traceability through database-connected records, and supported process monitoring through SPC charts in the browser interface.
The developed solution combined a Keyence LJ-S080 3D laser sensor, an LJ-S8000 controller, a Keyence KV-8000 PLC, a UR10 collaborative robot for engine-line scanning, a PostgreSQL database, and a browser-based user interface. The frame-line solution used a fixed scanner with pneumatic actuation, while the engine-line solution used type-dependent robot motion to position the scanner at different stamping locations. The web interface supported data search, visualization of scanned images and measurement results, and SPC-based trend monitoring. The LJ-S080 was selected because it provides high-resolution 3D surface measurement for character recognition and dimensional evaluation, while PostgreSQL enabled reliable storage and retrieval of inspection data.
The system was evaluated through industrial validation under stabilized operating conditions using 800 units in total, comprising 400 frame-line units and 400 engine-line units. Validation was based on com-parison between automated outputs, the previous manual inspection method, and internal engineering acceptance requirements. In the final validated dataset, the developed system achieved 100% OCR ac-curacy under the tested conditions. It also reduced inspection time from approximately 7 minutes for engine-number inspection and 8 minutes for frame VIN inspection to 20 seconds per unit, improved digital traceability through database-connected records, and supported process monitoring through SPC charts in the browser interface.