Search Results (9)

The rapid advancement in self-driving and autonomous vehicles (AVs) integrated with artificial intelligence (AI) technology demands not only precision but also output transparency. In this paper, we propose a novel hybrid explainable AI (XAI) framework that combines local interpretable model-agnostic explanations (LIME) and Shapley additive explanations (SHAP). Our framework combines the precision and globality of SHAP and low computational requirements of LIME, creating a balanced approach for onboard deployment with enhanced transparency. We evaluate the proposed framework on three different state-of-the-art models: ResNet-18, ResNet-50, and SegNet-50 on the KITTI dataset. The results demonstrate that our hybrid approach consistently outperforms traditional approaches by achieving a fidelity rate of more than 85%, interpretability factor of more than 80%, and consistency of more than 70%, surpassing the conventional methods. Furthermore, the inference time of our proposed framework with ResNet-18 was 0.28 s; for ResNet-50, it was 0.571 s; and that for SegNet was 3.889 s with XAI layers. This is optimal for onboard computations and deployment. This research establishes a strong foundation for the deployment of XAI in safety-critical AV with balanced tradeoffs for real-time decision-making. Full article

Ensuring the safety of autonomous vehicles (AVs) in complex and high-risk traffic scenarios remains a critical unresolved challenge. Current AV planning methods exhibit limitations in generating robust driving trajectories that effectively avoid collisions, highlighting the urgent need for improved planning strategies to address these issues. This paper introduces a novel iterative optimization framework that incorporates safety-critical trajectory generation to enhance AV planning. The use of the HighD dataset, which is collected using the wide-area remote sensing capabilities of unmanned aerial vehicles (UAVs), is fundamental to the framework. Remote sensing enables large-scale real-time observation of traffic conditions, providing precise data on vehicle dynamics, road structures, and surrounding environments. To generate safety-critical trajectories, the decoder within the conditional variational auto-encoder (CVAE) is innovatively designed through a data-mechanism integration method, ensuring that these trajectories strictly adhere to vehicle kinematic constraints. Furthermore, two parallel CVAEs (Dual-CVAE) are trained collaboratively by a shared objective function to implicitly model the multi-vehicle interactions. Inspired by the concept of “learning to collide”, adversarial optimization is integrated into the Dual-CVAE (Adv. Dual-CVAE), facilitating efficient generation from normal to safety-critical trajectories. Building upon this, these generated trajectories are then incorporated into an iterative optimization framework, significantly enhancing the AV’s planning ability to avoid collisions. This framework decomposes the optimization process into stages, initially addressing normal trajectories and progressively tackling more safety-critical and collision trajectories. Finally, comparative case studies of enhancing AV planning are conducted and the simulation results demonstrate that the proposed method can efficiently enhance AV planning by generating safety-critical trajectories. Full article

Source counting is the key procedure of autonomous detection for underwater unmanned platforms. A source counting method with local-confidence-level-enhanced density clustering using a single acoustic vector sensor (AVS) is proposed in this paper. The short-time Fourier transforms (STFT) of the sound pressure and vibration velocity measured by the AVS are first calculated, and a data set is established with the direction of arrivals (DOAs) estimated from all of the time–frequency points. Then, the density clustering algorithm is used to classify the DOAs in the data set, with which the number of the clusters and the cluster centers are obtained as the source number and the DOA estimations, respectively. In particular, the local confidence level is adopted to weigh the density of each DOA data point to highlight samples with the dominant sources and downplay those without, so that the differences in densities for the cluster centers and sidelobes are increased. Therefore, the performance of the density clustering algorithm is improved, leading to an improved source counting accuracy. Experimental results reveal that the enhanced source counting method achieves a better source counting performance than that of basic density clustering. Full article

Agrivoltaics (Agri-PV, AV)—the joint use of land for the generation of agricultural products and energy—has recently been rapidly gaining popularity, as it can significantly increase income per unit of land area. In a broad sense, AV systems can include converters of solar energy, and also energy from any other local renewable source, including bioenergy. Current approaches to AV represent the evolutionary development of agroecology and integrated PV power supply to the grid, and can result in nearly doubled income per unit area. AV could provide a basis for a revolution in large-scale unmanned precision agriculture and smart farming which will be impossible without on-site power supply, reduction of chemical fertiliser and pesticides, and yield processing on site. These approaches could dramatically change the logistics and the added value production chain in agriculture, and so reduce its carbon footprint. Utilisation of decommissioned solar panels in AV could halve the cost of the technology and postpone the need for bulk PV recycling. Unlike the mainstream discourse on the topic, this review feature focuses on the possibilities for AV to become more strongly integrated into agriculture, which could also help in resolution of relevant legal disputes (considered as neither rather than both components). Full article

Three dimensional path planner is crucial for the safe navigation of autonomous vehicles (AV), such as unmanned aerial vehicles or unmanned underwater vehicles, which operate in three dimensions. In this paper, we develop a novel 3D path planner, which is fast in generating a near-optimal solution path. The planner generates the 3D path considering the size of an AV so that as the AV traverses the constructed path, it does not collide with an obstacle. This paper introduces a 3D path planner with novel concepts, such as a virtual agent and virtual sensors. In order to generate a 3D path to the goal as fast as possible, we let the virtual agent deploy virtual sensors iteratively, such that the connected sensor network can be formed. The constructed sensor network serves as a topological map for the AV, and we find a shortest path from the start to the goal utilizing the network. The virtual agent’s maneuver is biased towards the goal, in order to find a path to the goal as fast as possible. Moreover, the size of the agent is set considering the safety margin of the generated path. Through MATLAB simulations, we demonstrate the outperformance (low computational load and short path length) of our 3D path planner by comparing it with the 3D RRT-star algorithm. Full article

This study investigates the use of unmanned aerial systems (UAS) mapping for monitoring the efficacy of invasive aquatic vegetation (AV) management on a floating-leaved AV species, Nymphoides cristata (CFH). The study site consists of 48 treatment plots (TPs). Based on six unique flights over two days at three different flight altitudes while using both a multispectral and RGB sensor, accuracy assessment of the final object-based image analysis (OBIA)-derived classified images yielded overall accuracies ranging from 89.6% to 95.4%. The multispectral sensor was significantly more accurate than the RGB sensor at measuring CFH areal coverage within each TP only with the highest multispectral, spatial resolution (2.7 cm/pix at 40 m altitude). When measuring response in the AV community area between the day of treatment and two weeks after treatment, there was no significant difference between the temporal area change from the reference datasets and the area changes derived from either the RGB or multispectral sensor. Thus, water resource managers need to weigh small gains in accuracy from using multispectral sensors against other operational considerations such as the additional processing time due to increased file sizes, higher financial costs for equipment procurements, and longer flight durations in the field when operating multispectral sensors. Full article

This article presents the results of a computational investigation of an integrated propeller test case using the HPCMP CREATE${}^{\mathrm{TM}}$-AV Kestrel simulation tools. There is a renewed interest in propeller-driven aircraft for unmanned aerial vehicles, electric aircraft, and flying taxies. Computational resources can significantly accelerate the generation of aerodynamic models for these vehicles and reduce the development cost if the prediction tools can accurately predict the aircraft/propeller aerodynamic interactions. Unfortunately, limited propeller experimental data are available to validate computational methods. An American Institute of Aeronautics and Astronautics (AIAA) workshop was therefore established to address this problem. The objective of this workshop was to generate an open access-powered wind tunnel test database for computational validation of propeller effects on the wing aerodynamics, specifically for wing-tip-mounted propellers. The propeller selected for the workshop has four blades and a diameter of 16.2 in. The wing has a root and tip chord of 11.6 and 8.6 in, respectively. Two different simulation approaches were used: one using a single grid including wind tunnel walls and the second using a subset grid overset to an adaptive Cartesian grid that fills the space between the near-body grid and wind tunnel walls. The predictions of both approaches have been compared with available experimental data from the Lockheed Martin low-speed wind tunnel to investigate the grid resolution required for accurate prediction of flowfield data. The results show a good agreement for all tested conditions. The measured and predicted data show that wing aerodynamic performance is improved by the spinning tip-mounted propeller. Full article

Unmanned aerial systems (UAS) fill a gap in high-resolution observations of meteorological parameters on small scales in the atmospheric boundary layer (ABL). Especially in the remote polar areas, there is a strong need for such detailed observations with different research foci. In this study, three systems are presented which have been adapted to the particular needs for operating in harsh polar environments: The fixed-wing aircraft M ${}^2$ AV with a mass of 6 kg, the quadrocopter ALICE with a mass of 19 kg, and the fixed-wing aircraft ALADINA with a mass of almost 25 kg. For all three systems, their particular modifications for polar operations are documented, in particular the insulation and heating requirements for low temperatures. Each system has completed meteorological observations under challenging conditions, including take-off and landing on the ice surface, low temperatures (down to −28 ${}^{\circ }$ C), icing, and, for the quadrocopter, under the impact of the rotor downwash. The influence on the measured parameters is addressed here in the form of numerical simulations and spectral data analysis. Furthermore, results from several case studies are discussed: With the M ${}^2$ AV, low-level flights above leads in Antarctic sea ice were performed to study the impact of areas of open water within ice surfaces on the ABL, and a comparison with simulations was performed. ALICE was used to study the small-scale structure and short-term variability of the ABL during a cruise of RV Polarstern to the 79 ${}^{\circ }$ N glacier in Greenland. With ALADINA, aerosol measurements of different size classes were performed in Ny-Ålesund, Svalbard, in highly complex terrain. In particular, very small, freshly formed particles are difficult to monitor and require the active control of temperature inside the instruments. The main aim of the article is to demonstrate the potential of UAS for ABL studies in polar environments, and to provide practical advice for future research activities with similar systems. Full article

Advances in the development of micro-electromechanical systems (MEMS) have made possible the fabrication of cheap and small dimension accelerometers and gyroscopes, which are being used in many applications where the global positioning system (GPS) and the inertial navigation system (INS) integration is carried out, i.e., identifying track defects, terrestrial and pedestrian navigation, unmanned aerial vehicles (UAVs), stabilization of many platforms, etc. Although these MEMS sensors are low-cost, they present different errors, which degrade the accuracy of the navigation systems in a short period of time. Therefore, a suitable modeling of these errors is necessary in order to minimize them and, consequently, improve the system performance. In this work, the most used techniques currently to analyze the stochastic errors that affect these sensors are shown and compared: we examine in detail the autocorrelation, the Allan variance (AV) and the power spectral density (PSD) techniques. Subsequently, an analysis and modeling of the inertial sensors, which combines autoregressive (AR) filters and wavelet de-noising, is also achieved. Since a low-cost INS (MEMS grade) presents error sources with short-term (high-frequency) and long-term (low-frequency) components, we introduce a method that compensates for these error terms by doing a complete analysis of Allan variance, wavelet de-nosing and the selection of the level of decomposition for a suitable combination between these techniques. Eventually, in order to assess the stochastic models obtained with these techniques, the Extended Kalman Filter (EKF) of a loosely-coupled GPS/INS integration strategy is augmented with different states. Results show a comparison between the proposed method and the traditional sensor error models under GPS signal blockages using real data collected in urban roadways. Full article