Introduction to data validity

The validity of the data produced with computer vision is dependent on the quality of the video used as input for the analysis. Video quality affects all data types that RoadAI produces, with some being more susceptible to specific aspects of video quality than others.

Video quality depends on three key factors:

• The technical quality of the video produced by the phone
• The recording setup in the vehicle used for surveying
• The state of the road surface, as well as weather and lighting conditions

To ensure optimal technical quality, use the recommended phone models for recording. Follow the setup instructions for mounting and positioning the phone in the vehicle.

Since RoadAI automates a visual survey, the assessment is subject to weather and lighting conditions, just as in conventional windscreen surveys. Lighting and the state of the road surface significantly affect how features on the surface are visually perceived.

For example, a wet road surface tends to show less detail compared to a dry surface. Direct sunlight makes road surface features more visible than on an overcast day. The direction of sunlight also affects perception. Direct sunlight can cast dark shadows from nearby vegetation onto the road, which may obscure details and affect assessment results.

To ensure valid and repeatable results, follow the recommendations in the Recommendations for best surveying results.