Accurate mapping of underground infrastructure is essential for infrastructure maintenance, safety inspections, and urban planning. However, environments with poor lighting, complex geometry, and no GNSS signal pose serious challenges for traditional methods. Total stations, LiDAR, and SLAM-based systems offer high precision but come with significant financial, logistical, and technical constraints.
Our Swiss partner — Pix4D — conducted an experimental study demonstrating that underground structures can be effectively mapped using:
· a smartphone with an RTK GNSS module;
· Pix4D AutoTags;
· the PIX4Dcatch application.
The results showed accurate mapping in challenging tunnel conditions—with minimal cost and without the need for specialized teams.
Experiment conditions
The study was conducted in a pedestrian tunnel near Preverenges, Switzerland, approximately 98 feet long and 10 feet in diameter. The following equipment was used for data collection:
· iPhone 14 Pro smartphone;
· Emlid Reach RX RTK receiver;
· PIX4Dcatch mobile application with AutoTags enabled;
· external lighting to enhance image quality in the dark environment.
Underground mapping workflow
During the tunnel scan, 10 AutoTags were placed — targets that are automatically detected in real time and used to improve model accuracy in environments without GNSS signal.
Methodology
To evaluate accuracy and repeatability, four scanning patterns were tested:
· single direction with straight pattern;
· single direction with spiral pattern;
· double direction with straight pattern;
· double direction with spiral pattern.
Visual comparison of the fused point clouds (dense point cloud + LiDAR) for the four scanning patterns
The data was processed in PIX4Dmatic to generate a dense point cloud, Digital Surface Model (DSM), 3D mesh, and orthomosaic. The processing used the Geofusion algorithm, which combines data from GNSS, ARKit sensors, and AutoTags to build a precisely georeferenced model—even in cases of signal loss.
Results
The double-direction spiral pattern gave the most complete results, creating a clean point cloud with no visible artifacts. The absolute error was always under 0.9 inches, and the standard deviation was 0.5 inches. Considering the object scale (~6.5 feet), this means the relative accuracy was better than 1%.
The AutoTags system provided:
- autotags are detected in real-time during data acquisition, enabling efficient and immediate feedback.
- the 3D position of the Autotags does not need to be known in advance, simplifying setup.
- drift compensation is performed automatically on your smartphone, with corrections applied within seconds after data collection is complete.
- autotags are uniquely coded and can be used to align multiple scans with each other, ensuring consistency and accuracy across different scans or sections of the scanned scene.
Conclusions
The study confirms the feasibility of accurate 3D mapping of underground structures using a smartphone, RTK receiver, and AutoTags. This approach significantly reduces project costs, lowers the need for expert photogrammetry skills, and enables mapping to be integrated into regular workflows without requiring specialized personnel.
The updated PIX4Dcatch toolset with AutoTags technology opens up new possibilities for inspecting tunnels, utilities, underpasses, and other infrastructure—even in environments with challenging lighting conditions and no GNSS signal.
Futurology is the official distributor of Pix4D software. The brands' products are available through the Futurology dealer network in the United States of America. For more information, please contact [email protected].
*This article is based on a research paper by Christoph Strecha, Martin Rehak, and Davide Antonio Cucci.
