The thesis explores the development and implementation of an automated height measurement system for trucks using computer vision and AI at Volvo's Tuve plant. The manual measurement process, currently a bottleneck, is prone to errors and inefficiencies. The proposed system leverages advanced technologies like YOLOv8 and Mask R-CNN for real-time, accurate height measurement.

The study includes detailed investigation methods, data collection, model training, interface development, and extensive testing to ensure system reliability. The system demonstrated high accuracy, reducing the error margin significantly compared to manual methods, and improved operational efficiency by reducing measurement time from 60 seconds to 120 milliseconds. The system was capable of measuring the height of the truck within the range of -+5cm of variation which is 50% lower compared to current manual measurement.

This advancement not only enhances productivity but also aligns with Volvo's commitment to technological innovation and is submitted as a proof of concept, mostly replacing the current measurement method.