Drone LiDAR and Ground-Penetrating Radar Simulation for Concrete Inspection in Nuclear Facilities
Conference
Saripalli, HHB, Lagos, L. (2026). Drone LiDAR and Ground-Penetrating Radar Simulation for Concrete Inspection in Nuclear Facilities
. TRANSACTIONS OF THE AMERICAN NUCLEAR SOCIETY, 134 224. 10.13182/T134-11489
Saripalli, HHB, Lagos, L. (2026). Drone LiDAR and Ground-Penetrating Radar Simulation for Concrete Inspection in Nuclear Facilities
. TRANSACTIONS OF THE AMERICAN NUCLEAR SOCIETY, 134 224. 10.13182/T134-11489
The inspection of nuclear infrastructure is a significant challenge because of the limited accessibility, structural aging, and radiation exposure risks. This paper introduces a MATLABbased simulation framework for a drone-enabled Digital Twin system that integrates Ground Penetrating Radar (GPR) and LiDAR sensors to facilitate autonomous concrete structure inspection. To validate multi-modal sensor fusion algorithms, a virtual inspection arena measuring 3×3×2.5m was created, which contained 137 programmatically embedded anomalies across 11 material categories. The simulated quadrotor achieved 98-99% surface coverage by performing wall-following trajectories with sinusoidal scanning patterns. In order to facilitate real-time 40Hz sensor fusion, a novel detection algorithm that utilizes spatial indexing with a 20cm grid resolution attained O(log n) computational efficiency. Through a weighted combination of material reflectivity factors, angle of incidence, and distance attenuation, detection confidence was determined. The results indicate that, through sensor fusion, 99.3% of anomalies are detected, which is a significant improvement over the efficacy of individual sensors (LiDAR: 95.6%, GPR: 82.5%). The dual-sensor approach is validated by the complementary detection capabilities of LiDAR, which excels at surface anomalies and GPR, which penetrates subsurface structures. In order to facilitate the digital transformation of nuclear Non-Destructive Testing (NDT) operations, this simulation framework offers a riskfree environment for algorithm development and optimization prior to field deployment in radiological environments.
