Traditional topography vrs. LiDAR. Accuracy, time and costs.

Doing a job with LiDAR could be more accurate than with conventional topography? If it reduces times, in what percentage? How much does it reduce costs?

Times have definitely changed. I remember when Felipe, a surveyor who did the fieldwork I came with a notebook of 25 pages of cross sections to generate with it a map of contour lines. I did not live the time to interpolate on paper but I remember doing it with AutoCAD without using Softdesk yet. So he interpolated with Excel to know how far to place the elevation between the two elevations and, these points placed them in layers of different levels and colors, to finally join them with polylines that turned into curves.

While cabinet work was crazy, it did not compare with the fieldwork that was an art, if you wanted to have enough data to make an acceptable modeling when the altimetry was irregular. Then came SoftDesk, the antecedent of AutoCAD Civil3D that simplified cabinet and Felipe was in one of my courses learning to use a total station, which reduced the time, increased the volume of points and of course the precision.

Stage Drones for civil use Breaks new paradigms, under a similar logic: Resistance to change in topographical techniques always seeks to reduce costs and guarantee accuracy. So we will analyze in this article two hypotheses that we have heard there:

1 Hypothesis: Doing topography with LiDAR reduces times and costs.

Hypothesis 2: To do topography with LiDAR implies loss of precision.

The experimental case

The magazine ALL systematized a work in which a work was done in the data collection of a dike, using conventional method along 40 kilometers. Separately, in a second task a few days later it was developed using topography with LiDAR along 246 kilometers of the same dam. Although the sections were not equal in distance, the equivalent section was compared to make a comparison in similar conditions.

Conventional topography

The topographic survey was collected in cross sections at each 30 meters, matching existing stations. The transverse points were taken at distances less than 4 meters.

Georeferenced the work with points of the geodetic network, which were validated with geodetic GPS along the axes, and from these the transverse points were raised using a combination of virtual reference stations and RTK. It was necessary to take additional points in special places of change of slope and shape to ensure consistency of the digital model.

To deal topography

The residual differences between the known points and the coordinates obtained by the GPS were those shown in the table, confirming That conventional lifting is very accurate.

Maximum Residual Minimum residual square
Horizontal 2.35 centimetres. 1.52 centimetres.
Vertical 3.32 centimetres. 1.80 centimetres.
Three Dimensional 3.48 centimetres. 2.41 centimetres.

The LiDAR survey

This was done with an Autonomous Unit flying at a height of 965 meters, with a density of 17.59 points per square meter. They retrieved 26 known control points and crossed them against 11 additional first-order points that were read with geodetic GPS.

With these 37 points the adjustment of the LiDAR data was made. Although it was not necessary since the coordinates taken by the UAV that comes equipped with GPS receiver and controlled by base stations, obtained at all times a minimum of 6 satellites visible and a PDOP smaller to 3. The distances to the base station were never greater than 20 kilometers.

A set of 65 additional control points served to validate the accuracy of the LiDAR data. With respect to these points, we obtained the following vertical precisions were:

In Urban Area: 2.99 cm. (9 points)

In open field or low grass: 2.99 cm. (38 points)

In forest: 2.50 cm. (3 points)

In shrubs or tall grass: 2.99 cm. (6 points)

To deal topography

The image shows the large difference in density between the points taken with LiDAR against the cross sections marked in green triangles.

Differences in Precision

The finding is more than interesting, contrary to the hypothesis that the LiDAR survey does not reach the precisions of a conventional survey. The following are RMSE (Root mean square error) values, which is the error parameter between the captured data and the reference control points.

Conventional topography LiDAR lifting
1.80 centimetres. 1.74 centimetres.

Differences in Time

If the above has surprised us, see what happened in terms of time reduction in a comparative way between the LiDAR method and the traditional method:

The data collection in the field with LiDAR was only the 8%.

  • Cabinet work was only 27%.
  • Summing the field + flight + LiDAR cabinet hours against the field data + Conventional topography cabinet, LiDAR required only the 19%.

To deal topography

As a consequence, the 123 hours of work per kilometer of conventional topography were reduced to only 4 hours per kilometer.

In addition, if the total of captured points is divided between the time consumed in the capture and cabinet processes, the conventional method obtained 13.75 points per hour, against 7.7 million points per hour of LiDAR.

Differences in Time

The costs of these modern equipments, with those sensors capturing that amount of points, suppose that the work should be more expensive. But in practice, the reduction of time and expenses of mobilization that implies the conventional topography, The final cost to the customer of the 246 kilometers resulted with LiDAR 71% lower than the total cost of the 40 kilometers with conventional topography !.

It seems incredible, but the price per linear kilometer with LiDAR was just 12% compared to conventional topography.

Conclusion

Does the topography with LiDAR totally replace the traditional topography? Not in total, because the work with LiDAR always occupies some topography for control points, but it can be concluded that with all the advantages of cost, product quality and time, the work with LiDAR generates results with almost the same precisions of the topography conventional.

There will always be pros and cons; The high accuracy of conventional topography is nostalgic, but the complications of asking for permission to enter private property, risks of location in irregular sites, need for gaps in high grass and obstacles ... is crazy. Of course, the density of forest cover also has its disadvantages in the case of LiDAR, nor are the same parameters of relation between extremely small projects.

In conclusion, we are pleased to know how technology has advanced to the degree that for large projects like the one raised, it is necessary to have an open mind and willingness to opt for new and more creative ways of doing topography.

8 Replies to “Traditional topography vrs. LiDAR Accuracy, time and costs. ”

  1. Good morning ... friends .... referring to the use of the drones to generate a survey ... .who would be in sensor and / or equipment indicated to raise a large area (1000 Has or more) with dense or very dense vegetation? where access is very difficult.
    Excellent article!!

  2. Very good information and gives me a better point of view of this technology, I also conclude that for designs is a great tool, but for experiences lived in the execution the conventional topography with total stations takes great importance, requiring to make number of adjustments in lines bases in dimensions and coordinates that give the required presicion for a project in execution phase where parameters less than 0.05m of error are required. regards

  3. JOHAM

    I LIKE A LOT OF CLARION ALONG AND PERSISTS THE DOUBT IF YOU CAN OBTAIN THE SAME PRECISION.

  4. It is important to know the reality in highly populated urban environments, since not all types of projects can generalize the precisions and times.

  5. Excellent article…!!! I think it's a doubt that we all have at some point

  6. THANKS FOR THE CLARIFICATION WAS WITH THE QUESTION OF WHICH WOULD BE THE MOST ACCURATE
    GOOD CONTRIBUTION

Leave an answer

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.