|Job Type||Surveying Road and Intersections|
|Project Date||October 2014|
|Project Size||1,800 feet in length, 10 feet of each side|
|Drone||Matrice 100 with Sony ILCE-6000|
|Flight Altitude||75m (246ft) & 130m (426ft)|
|Accuracy||Better than two centimeters (5-6 hundredths of a foot)|
|DatuSurvey savings||5 Days|
The required task was to survey and draft a 3D CAD plan, with a measurement accuracy of better than two centimeters (5-6 hundredths of a foot), of Charles street in Morristown, Tennessee –
1,800 feet in length, 10 feet of each side – and three nearby
Original Effort Estimate
The original effort estimate for the surveying project, using conventional surveying techniques, was 16 days for two crews. This appraisal was based on 8 days in the field, using a Total Station device from several different stations, and 8 days in the office to draft the detailed CAD plans of the street and the intersections based on the Total Station measurements and sketches done in the field.
A key challenge related to the survey was occupational safety hazards due to traffic on the road during most of the day.
Actual Required Time Using DatGram™3D
The actual time to survey and prepare a 3D CAD model of the street and the three intersections was only 5 days.
The time spent in the field was 2 days: one day to photograph the road and the three intersections, and another day to measure several control points along the road and in the intersections. The road and the intersections were photographed
with a regular Sony ILCE-6000 camera (24-mega pixel
resolution) with a 16-mm wide-angle lens. To make the
surveying work more efficient, the street was divided
into six segments. For each segment 40 to 50 images
were made from the road’s shoulders toward the
center. In addition, a total of 100 images of the three
intersections were captured from the perimeter of each intersection
towards its center, with an image taken every 15 meters (50 feet). To minimize foreground obstructions in the images and expedite fieldwork, the camera was mounted on a small quadcopter that flew at an elevation of 35 to 45 meters (120 to 150 feet) above ground.
It is important to point out that, by surveying the road with a camera and quadcopter, the
field crew was able to avoid entering the busy road. This permitted the survey to be done
during daytime while maintaining the strictest occupational safety standards.
A total of 80 points were measured along the road using a prism-less Total Station device. Ten
of these points were used as control points for the geo-referencing of the images for each job
(road segment). The selected control points were high-contrast, well-defined objects such as
corners of road signs, corners of white marks painted on the road, etc.
In addition, in each intersection a total of about 30 points were measured using a prism-less
Total Station device. Ten of these points were used as control points for the geo-referencing of the images for each intersection.
The actual time in the office to geo-reference the images and draft the 3D CAD plan of the whole road and the three intersections directly on the images was three days.
Nine 3D CAD models of the street and adjacent intersections in DXF format were generated
by drafting directly on the oblique images using the DatuGramTM3D application. In addition to
the DXF file, a list of the measured points in PNT format was generated. This included the
point names, codes, descriptions, coordinates, and their measurement accuracies in all axes.
Overall, 646 new measurements were made from the images for the street and another 437
new measurements were made for the three intersections. The measurement accuracy of all
points was better than 2 centimeters (5-6 hundredths of a foot) in both position and elevation.
Based on the DXF files that were exported from DatuGramTM3D, a CAD model of the full length
of the street and adjacent intersections has been created in AutoCAD.
The final survey map of the road made based on DatuGRamTM3D: