The Kansas Department of Transportation (KDOT) faced data challenges when looking to meet the 2026 Model Inventory of Roadway Elements (MIRE) Fundamental Data Elements (FDEs) deadline, an essential component for federal funding eligibility under the Highway Safety Improvement Program (HSIP). As part of the FHWA’s Data and Analysis Technical Assistance (DATA) Teams program, which provides technical assistance to states for safety data issues and supports the advancement of safety data programs, VHB employed Artificial Intelligence (AI) to help KDOT collect reliable and cost-effective roadway data.
In Kansas, a significant hurdle was the unknown mileage of local roads that lacked surface type classification data. This was typically gathered manually, was labor-intensive, and often produced unreliable results. Using digital maps and Geographic Information Systems (GIS), the VHB team leveraged high-resolution aerial imagery and machine learning models to identify road surface types with a 90-95% accuracy rate. This method provided a comprehensive and efficient data collection solution, vastly improving decision-making for future roadway improvements. The integration of AI not only enhanced data quality and saved costs but also allowed Kansas to meet federal requirements while paving the way for a safer and more reliable road system.
