Fulton Hogan Abstract
Rapid advances in machine learning and cloud based computing are already having an outsize impact on the Public Works community. While these types of advances are commonly seen in the tech sector, their adoption is becoming more widespread in other industries.
Every Road Authority that is responsible for their road management knows the complexity in developing an accurate and comprehensive road management program. However, more and more communities are adopting the use of Artificial Intelligence based methods to create reliable and objective road condition assessments.
Here we present unique technology originally developed at Carnegie Mellon University, Pittsburg, successfully deployed in 200+ communities around the world including Australia and New Zealand. Using a windshield mounted smartphone camera, to identifying all the major surface distress categories - potholes, longitudinal/transverse cracking, surface deformation etc.
The machine learning artificial intelligence platform continues to develop, meeting pavement managers needs and has been trained to identify individual distresses automatically using the video data captured from the smartphone. With these advances, road condition and individual distress data can be collected easily and more frequently since it only requires a simple drive down the road at highway speeds.
The dataset is also delivered inside a simplified online mapping software, where condition ratings and individual distresses (potholes, cracking and surface defects) are presented and reported every 3 meters. With this dataset, communities then develop their road programs using their up-to-date and accurate road condition information. This can more easily be translated into robust action pavement management plans/programs nominating corresponding surface treatments.
Public Works Engineers are better able to estimate their forward maintenance costs, but also to better project the likely deteriorations of the road over time. All of this enables a more effective use of rate-payer money to managing and maintaining road infrastructure.