Monitoring changes in road conditions caused by single heavy load through mobile laser scanning systems

No Thumbnail Available
Journal Title
Journal ISSN
Volume Title
Insinööritieteiden korkeakoulu | Master's thesis
Date
2022-12-12
Department
Major/Subject
Highway Engineering
Mcode
Degree programme
Master's Programme in Geoengineering (GEO)
Language
en
Pages
49+18
Series
Abstract
This study presents the feasibility of MLS system to accurately identify changes in the condition of low-volume roads that are not designed for special heavy loads. The observation of changes in condition was based on monitoring settlement of the road surface. Two MLS systems, Riegl VMX-2HA and Trimble MX9, were used to detect settlement caused by heavy loading. For each MLS system, two-point clouds were produced, one before and one after loading. The data was analyzed visually as well as analytically. The visual analysis made it possible to visualize long enough sections of the road in longitudinal direction. The settlement was tire wide and occurred mostly on the right wheel path due to its proximity to the right edge of the embankment. This occurred as a result of the loaded truck using the right side of the road, leaving the left side of the road unaffected. Moreover, the roads are usually more compacted in the center whereas closer to the edges the embankment tends to be less stable. Five locations were picked where the settlement in the road was clearer to be identified. Analytical analysis of the changes that happened as a result of heavy loading was performed. The goal of this investigation was to determine the magnitude of the vertical displacement and assess the MLS systems' accuracy. In all five sections, the right wheel had the highest settlement. In most cases, both MLS systems detected the right wheel path at the same place with approximately the same magnitudes of settlement. For each segment, the settlement on the right wheel path was measured. It was discovered that the Trimble MX9 provides slightly higher settlement values than the Riegl VMX-2HA. However, both MLS systems gave the maximum settlement on the right wheel path in the same section which was 23.7 mm by Riegl VMX-2HA and 26.2 by Trimble MX9. According to calculations, there is an average difference of 2.7 mm between the two MLS systems' settlement measurements on the right wheel path for all the sections. Accuracy is the most reliable and conclusive measure of data quality. Therefore, a 1.5 m accuracy evaluation window that was made on the left side of the road and put 20 cm farther away from the left wheel of the vehicle to evaluate the accuracy of both MLS systems. This side of the road was not affected by the loaded vehicle; hence no vertical displacement was expected on this side of the road. However, the results from both MLS systems on this location showed some amount of vertical displacement. Therefore, three primary error sources that contributed to the vertical displacement in this location were identified. Such as alignment error that resulted from the incorrect alignment of point clouds with respect to one another. The average vertical displacement values received from the accuracy assessment window were used to evaluate the magnitude of the point clouds alignment error. The average values in this region of the road varies for each section, and it ranges from -0.01 mm to -4.168 mm for Riegl VMX-2HA and -0.354 mm to -5.518 mm for Trimble MX9. The internal error of the system indicated in the system specifications is the second error source. The accuracy assessment window's maximum and minimum values were taken to represent the internal inaccuracy of the MLS systems. The average internal error resulting from the Riegl VMX-2HA was 4.2 mm, whereas the average internal error resulting from the Trimble MX9 was 3.1 mm. The last error source was identified as originating from unbound materials that were moving around the tires of the scanning vehicle as it was driving. The laser beam detects those particles and assign the position to each individual point. The rest of the roughness in the result was due to this error. This error could be fixed by accurately removing the noise by using noise filters and by averaging procedure which filter away the local displacement.
Description
Supervisor
Cannone Falchetto, Augusto
Thesis advisor
Bussman, Sebastian
Gustavsson, Henry
Keywords
settlement, visual, analytical, accuracy, alignment, internal error
Other note
Citation