The Institute for Transportation Research and Education at North Carolina State University conducted a follow-up study to a previous North Carolina Department of Transportation (NCDOT) project, comparing mobile inventory data collection vehicles to manually-collected data techniques. In the previous studies, sign retroreflectivity readings were either captured with low degrees of accuracy or not captured at all.
The follow-up study focused mainly on automated sign retroreflectivity capture, but also looked at vendor capabilities regarding other sign features. The results show that vendors can accurately locate the majority of signs. However, while vendors were unable to consistently capture sign retroreflectivity readings within 10% accuracy, a comparison of MUTCD pass/fail ratings for signs for these vendors showed that they captured ground-mounted signs with 88% and 97% accuracy, and overhead signs with 100% accuracy. Combining the sign location rates and the accuracy of the pass/fail comparison results in an overall accuracy ranging from 63% to 70% which is comparable to the accuracy achieved by other sign management methods.
Vendors also showed some consistency in capturing the lower retroreflectivity readings, which should be more important to the NCDOT, as MUTCD thresholds for failing signs are set at the lower level readings. Following location of the sign, vendors showed promise collecting many of the other sign features, such as MUTCD code and roadside orientation, which showed significant improvement from the previous study.
This study shows that there is still room for improvement, but also exhibits the improvements that vendors have already made in capturing all sign features.
