Related Experiment Videos
Optimizing the Reliability of Underground Utility Tunnel Localization via Multi-Source Fusion and DVAE-CNN
Shaolong Chang1, Zhiguo Zhang1, Xueliang Gug1
1State Key Laboratory of Information Photonics and Optical Communications, School of Electronic Engineering, Beijing University of Posts and Telecommunications, Beijing 100876, China.
Sensors (Basel, Switzerland)
|July 28, 2026
Summary
This study introduces a new method for reliable underground utility tunnel localization using multi-sensor data fusion and a denoising variational autoencoder (DVAE-CNN). The approach significantly enhances positioning accuracy and reliability in complex subterranean environments.
Area of Science:
- Geospatial engineering
- Artificial intelligence
- Sensor fusion
Background:
- Underground utility tunnels present significant localization challenges due to complex environments and potential data loss.
- Existing localization methods often suffer from low reliability in these conditions.
Purpose of the Study:
- To develop a reliability-optimized localization method for underground utility tunnels.
- To improve the accuracy and robustness of positioning services in subterranean infrastructure.
Main Methods:
- Multi-source sensor data fusion with a unified time reference.
- Utilizing a convolutional neural network-assisted denoising variational autoencoder (DVAE-CNN) for localization.
- Implementing a multi-source heterogeneous data quality assessment model and environmental prior-information-aided weight update strategy.
Main Results:
- The proposed method achieved an average improvement of 73.6% in localization accuracy.
- Demonstrated an 85.2% improvement in localization reliability compared to methods without reliability regulation.
- Experimental validation in an underground utility tunnel confirmed robust and continuous positioning.
Conclusions:
- The DVAE-CNN based multi-source fusion method offers a highly reliable and robust solution for underground utility tunnel localization.
- This approach significantly overcomes limitations of existing methods in complex subterranean environments.
- The method shows strong potential for widespread application in infrastructure management and navigation.