Implementation of a 3-D Sensing Technology for Automated Pavement Data Collection in Connecticut
-
2018-06-25
Details:
-
Creators:
-
Corporate Creators:
-
Corporate Contributors:
-
Subject/TRT Terms:
-
Publication/ Report Number:
-
Resource Type:
-
Geographical Coverage:
-
Edition:Final Report
-
Corporate Publisher:
-
Abstract:This report summarizes research on the impact of replacing older ARAN equipment (2-D Strobe cameras, ultrasonic rut bar, and inertial profiler) by the newest 3-D laser technology (Pavemetric LCMS and RoLine sensors) on the accuracy and precision of pavement performance data collected by the Connecticut Department of Transportation (CTDOT). The main goal of the project was to develop transfer functions and/or correction factors relating cracking, rutting, and roughness values reported by the older and newer systems. In addition, precision and accuracy of the new technology was evaluated by comparison of data collected by two identical 3-D systems. This was to investigate comparability of historical and newly collected data. Therefore, one older and two newer ARAN vans were run simultaneously on about 127 km of preselected pavements to collect spatial location data and pavement surface images. The data were processed by the Fugro’s Roadware Vision software to obtain cracking length, rut depth, and roughness (IRI) values. Next, statistical significance of differences between 2-D and 3-D pavement performance datasets was evaluated. Overall, this project revealed the new 3-D laser scanning technology produces a much larger dataset and ultimately a much greater degree of resolution on the profiles due to an increase in the number of data points measured in both longitudinal and transverse directions. This substantially larger dataset created a smoothing effect (averaging of hundreds of values in the old system versus thousands of values in the new system) and under-estimating of the rut depths by the new system, for example. The new height sensors and accelerometers were also found to be extremely sensitive to undesirable vehicle movements that occur in response to obstacles, such as manhole covers, railroad crossings, and bridge joints. Speed gradient, including low speed and high acceleration or deceleration also adversely affected the output. The same undesirable movements apparently affected 3-D profile measurements that were used to generate 3-D images of pavement surfaces, thus, leading to differences in brightness, color, and range (depth) of images generated by the two supposedly identical ARAN systems (Van 8 and Van 9). Nevertheless, the research team believes that the precision and accuracy of the ARAN measurements can be positively improved by the adjustment of the settings and configurations of both hardware and software, the development of repeatable/reproducible system/vehicle operation routines, and the stricter use of quality assurance principles during data collection and data processing.
-
Format:
-
Funding:
-
Collection(s):
-
Main Document Checksum:
-
Download URL:
-
File Type: