Unmanned aerial vehicles (UAVs) extend geophysical measurements into terrain where ground access is difficult and crewed surveys are costly or operationally constrained. Their practical value, however, differs substantially among sensing methods and platform configurations. This narrative review compares UAV magnetometry, gravimetry, radiometric surveying, electromagnetic surveying, ground-penetrating radar (UAV-GPR), seismic surveying, and geological remote sensing. UAV magnetometry and geological surface mapping have the strongest evidence of established practice for defined tasks. Radiometric surveying, UAV-GPR, and several electromagnetic configurations have demonstrated field capability, but their performance depends on counting statistics, coupling, source geometry, and site conditions. UAV gravimetry has actual flight demonstrations, although large-aircraft results do not establish equivalent capability on compact platforms; UAV-assisted seismic systems remain application-specific. Across methods, useful performance depends on the complete measurement chain, particularly sensor clearance, motion and interference correction, timing, calibration, and independent validation. We identify requirements for multi-sensor interpretation and distinguish the roles and limitations of artificial intelligence in acquisition, processing, inversion, and quality control. Progress toward wider adoption should be assessed through reproducible survey performance and uncertainty, rather than payload integration alone.
European corn borer (ECB), Ostrinia nubilalis Hübner, 1796 (Lepidoptera: Pyralidae), is an important maize pest in Serbia and the wider Western Balkan region, where chemical control remains a common management practice. This study evaluated the effects of insecticide treatments applied by an unmanned aerial vehicle (UAV) under field conditions during a single growing season, with emphasis on maize yield, mycotoxin occurrence and the activity of naturally occurring Trichogramma spp. Field trials were conducted in 2023 using four maize hybrids and three treatments: a selective insecticide (spinetoram), a broad-spectrum insecticide (deltamethrin), and a water-only treatment, all applied using the same UAV-based application system. Spray deposition was assessed using water-sensitive papers, while ECB egg masses and Trichogramma parasitism were monitored during the growing season. The results showed that maize yield was primarily influenced by hybrid performance, with insecticide effects depending on treatment selectivity and hybrid response. Under the tested conditions, spinetoram resulted in higher yields than deltamethrin and the water treatment, whereas deltamethrin did not provide a yield advantage over the water-only treatment. Selective insecticide application was also associated with higher observed Trichogramma parasitism compared with the broad-spectrum treatment. Treatment-related differences were observed for selected mycotoxin parameters, including lower fumonisin B1 concentrations in the spinetoram treatment, although responses were not consistent across all analyzed mycotoxins. For HT-2 toxin and T-2 toxin, treatment-related differences were observed in their occurrence, but not in their concentrations. The findings provide field-based evidence on the performance of UAV-assisted insecticide application under the specific climatic, agronomic and pest-pressure conditions of the experiment. Under these conditions, the results indicate that insecticide selectivity and hybrid-specific responses are relevant considerations when evaluating IPM-oriented maize protection strategies, while further multi-location and multi-season studies are required to determine the stability and broader applicability of the observed effects.
During hover, DEP unmanned aerial vehicles (UAVs) suffer severe aerodynamic interference between rotor downwash and the wing, causing thrust loss. High-fidelity CFD can capture the interference physics but is too costly for iterative design. Existing reduced-order models (ROMs) are fast, but their predictions of wing download have not been adequately validated across variations in wing chord, nacelle offset, and rotor-to-wing height. To address this, we propose an integrated framework combining a validated overset-grid CFD model, a parametric reduced-order model that decomposes wing download into an equivalent coefficient multiplied by the slipstream-impinged area, and Latin hypercube sampling for design-space exploration. The equivalent coefficient is represented via Gaussian process regression over three geometric parameters using CFD-calibrated data. The reduced-order model predicts wing download in seconds (less than 10 s per evaluation on a standard desktop workstation), whereas the corresponding high-fidelity CFD simulation typically requires approximately 2 to 4 h per case on a 256 core HPC cluster. The reduced-order model predicts wing download in seconds with a maximum validation error of 7.6% and successfully classifies the design space into acceptable and high interference regimes. Nine geometrically distinct validation points across both interference regimes show correct categorization with no misclassification. The framework provides a lightweight screening tool for isolated rotor–wing interaction analysis, which is a fundamental building block for understanding more complex DEP configurations. The overall modeling framework is methodologically extendable to other configurations through recalibration, though such extensions are not pursued in the present work.
Wing icing alters the aerodynamic shape of lifting surfaces and threatens aviation safety. Existing icing-image datasets come mainly from wind tunnels or simulations, and image-level random splits overestimate generalisation for video data. We build a real UAV flight-test dataset of 13,890 pixel-level annotated wing images from backward-view videos over the Qinghai–Tibet Plateau, with flight-wise splitting and a workflow extending to onboard prototype verification. For boundary under-segmentation in small-area visible icing we introduce an ROI boundary-weighted loss (ROI masking plus boundary-pixel weighting) with optional optical flow inter-frame smoothing. Against a SENet+CBAM baseline it raises recall for Rice<3% by up to 6.03 percentage points in the best run (mean +2.27 pp; not resolved with three seeds), but at matched precision the recall margin over that baseline is 0.11 pp and the Dice difference is +0.02 pp (95% interval −0.10 to +0.13), whereas the gain over plain U-Net is 0.47 pp Dice (95% interval 0.36–0.58). The final scheme reaches 91.34% Dice and 84.07% IoU (three-run means); an embedded-GPU prototype supports a 2 s/frame cycle. No annotated frame is ice-free, so specificity was measured on 1547 frames of two non-icing sorties (false positive rate 2.37×10−3, specificity 0.99763); the plain U-Net baseline is better there, and 95.3% of frames still show a small false positive on the clean leading edge, so the detections are appearance-driven rather than a physical measurement of ice extent. The results describe one platform, camera, flight region and annotation protocol.
The growing demand for safe autonomous landing of unmanned aerial vehicles (UAVs) in complex terrain requires reliable guidance throughout the final descent, even under degraded Global Navigation Satellite System (GNSS) conditions, where intentional interference, signal obstruction, and residual positioning errors reduce navigation continuity. This paper presents a complementary navigation system combining a ground-based and an onboard MIMO Frequency-Modulated Continuous-Wave (FMCW) radar, providing estimation of slant range, relative velocity, and angular deviation independently of GNSS signal quality. Navigation errors and trajectory corrections are modeled in simulation, and the radar signal processing and localization algorithms are supported by laboratory measurements in an anechoic chamber and by UAV flight tests. The results show that the range, velocity and angular parameters of the radar are sufficient for guidance throughout the descent and that the slant range and height to the landing site are obtained directly, without reliance on satellite geometry. The system is designed around commercially available automotive radar modules, enabling cost-effective implementation with technologically mature hardware. The proposed concept is designed to support autonomous guidance of multirotor UAVs, fixed-wing platforms, and helicopters to a predefined landing site from stand-off distances of several hundred meters and establishes a foundation for mobile take-off and landing stations and hybrid navigation systems.
