This project evaluated the use of drone photogrammetry for determining the verticality and structural tilt of a tower by comparing its accuracy with conventional theodolite measurements.
Project objective
The goal was to determine the displacement of the structure's faces relative to the design axis rather than relying solely on angular measurements traditionally used for verticality assessment.
The project also compared the accuracy of drone-based surveying with conventional theodolite measurements.
Equipment and software
- DJI Phantom 4 RTK with a 20 MP camera.
- PIX4Dmapper.
Processing workflow
The workflow consisted of:
- Generating a dense 3D point cloud.
- Placing measurement markers on the reconstructed 3D model.
To improve measurement accuracy, each marker was identified on multiple photographs before its coordinates were calculated.
The calculated marker coordinates were exported to DXF format for further engineering analysis.
Deviation analysis
Following the methodology used by Vodafone, deviations were evaluated in plan view by analyzing the displacement of markers on each level relative to the base of the structure along the X and Y axes.
The calculated average values were compiled into a table used for structural tilt analysis.
Comparison of survey results
The following executive diagram illustrates the structural deviations obtained from the DJI Phantom 4 RTK survey.
The diagram below shows the results obtained using the traditional theodolite surveying method.
Project timeline
- Fieldwork: Approximately 30 minutes.
- Automatic data processing: Approximately 3 hours.
- Manual measurements: Approximately 30 minutes.
Results
The drone survey showed strong agreement with the traditional theodolite measurements.
The observed differences between the two methods were:
- Up to 2 cm along the X-axis.
- Up to 1 cm along the Y-axis.
Advantages of drone surveying
- Clear visibility of all faces of the structure without distortion caused by observation distance.
- No requirement to position the instrument at a specific location, unlike conventional theodolite surveys.
- Rapid collection of a large number of measurements during a short field visit.
- Ability to measure virtually any point on the structure after processing.
- Complete photographic documentation of the inspected object.
- Reduced risk of human error, particularly during instrument alignment.
Current limitations
- Image quality depends on lighting conditions.
- The drone-based workflow requires further methodological refinement.
- Signal interference from nearby communication towers may affect maneuverability when flying closer than approximately 3 meters.
Recommendations
- Validate the results using an additional independent survey, such as a total station.
- Continue improving the methodology together with an independent engineering organization or scientific institution.
Testing drones without RTK
The project also evaluated whether drones without RTK positioning could be used for verticality assessment.
Testing included:
- DJI Phantom 4 Pro.
- DJI Mavic 2 Pro.
When verticality deviation diagrams were generated from these datasets, the calculated deviations exceeded acceptable engineering tolerances by a significant margin, incorrectly indicating that the structure was in critical condition.
Software verification
To verify that the errors were not caused by processing software, the same datasets were processed using several photogrammetry platforms:
- PIX4Dmapper.
- Agisoft Metashape.
- DJI Terra.
All software packages produced similarly inaccurate results when RTK or PPK positioning was unavailable.
Conclusion: drones operating without RTK or PPK positioning should not be used for structural verticality assessment.
However, they remain highly effective and cost-efficient tools for inspection, documentation, and asset inventory applications.
