UVeye uses vision language models to detect vehicle defects at scale

Overview

Automated vehicle-inspection company UVeye processes over 700 million high-resolution images each month and applies vision language models to convert this visual data into structured condition reports, detecting subtle defects, modifications or foreign objects. UVeye detects 96% of defects compared with 24% using manual inspection methods.

The challenge

Traditional CNN-powered computer vision systems are tuned to spot specific visual anomalies but lack the multimodal ability to translate what they see into text, making it hard to turn high volumes of vehicle-inspection imagery into searchable, structured insight.

The solution

UVeye processes over 700 million high-resolution images each month to build one of the world's largest vehicle and component datasets, applying vision language models (VLMs) to convert this visual data into structured condition reports that detect subtle defects, modifications or foreign objects.

Reported business value

UVeye detects 96% of defects compared with 24% using manual inspection methods, enabling early intervention to reduce downtime and control maintenance costs.