The rapid increase in unmanned aerial vehicle (UAV) usage has introduced significant safety challenges, including issues such as system failure, loss of control, transmission failures, and collisions. Analyzing these incidents has been challenging due to the absence of a dedicated category field in the National Transportation Safety Board (NTSB) data. This research tackles this problem by utilizing artificial intelligence (AI) to automate the classification of UAV accident reports collected between 2006 and 2023. Using natural language processing techniques, we categorize NTSB reports to improve the analysis and interpretation of incident data. We also employ advanced data visualization tools to reveal geographic and temporal patterns, offering a detailed view of UAV accident trends. The results indicate that system and component failures unrelated to propulsion systems (system/component failure or malfunction [non-powerplant]) and abnormal contact upon landing (abnormal runway contact) are predicted as the primary categories (37%) of UAV accidents for the period. These insights suggest the potential value of AI-driven categorization and visualization techniques in enhancing UAV safety standards and supporting policy development. Initial results provide promising insight into the use of language models for text classification in aviation safety problems.

Due to the demand for high reliability and thermal conductivity of high-power modules operating at high temperatures, sintered nano-silver (Ag) has garnered significant attention as an excellent interconnect and heat transfer layer, particularly for its thermal conductivity and other reliability research. Since the mechanical behavior and heat conduction capacity of sintered Ag is generally regulated by changes in temperature, its microstructure will change accordingly, affecting its performance. In this study, a machine learning model was used to evaluate and predict the thermal conductivity of sintered Ag, providing an effective method to analyze the influence of microstructural characteristics on its heat transfer properties. Image processing and model simulation of scanning electron microscopy images of sintered nano-Ag nanostructures were performed using MATLAB and Ansys software. A batch calculation of the thermal conductivity of 2D images of sintered nano-Ag nanostructures was performed to obtain sufficient data sets. Based on the artificial neural network model of Bayesian optimization, the equivalent thermal conductivity of different sintered nano-Ag microstructures was predicted with high accuracy using the microstructure image and characteristic parameters of sintered nano-Ag. The proposed method enables rapid, effective, and accurate evaluation and prediction of the thermal conductivity of sintered nano-Ag, contributing significantly to the reliability of power modules.

Developing refractory high-entropy superalloys (RSAs) with performance advantages over nickel-based alloys is a critical frontier in materials science. Body-centered cubic (bcc)-based RSAs have attracted significant attention, with ruthenium (Ru) playing a key role in forming two-phase regions of A2 (disordered bcc) + B2 (ordered bcc), which could lead to superalloy-like microstructures. This study introduces the application of the Kolmogorov-Arnold Network (KAN) model to predict the mechanical and thermodynamic properties of Ru while comparing its performance against other commonly used machine-learned models. Utilizing density functional theory calculations as training data, the KAN model demonstrates superior accuracy and computational efficiency compared to conventional methods, while reducing descriptor complexity. The model accurately predicts a range of properties, including elastic constants, thermal expansion coefficients, and various moduli, with discrepancies within 6% of experimental reference data. Molecular dynamics simulations further validate the model’s efficacy, accurately capturing Ru’s phase transitions from hexagonal close-packed (hcp) to face-centered cubic structure and the melting point. This work presents the first application of KAN in materials science, demonstrating how its balanced performance and efficiency provide a new pathway for designing advanced materials, with unique advantages over conventional machine learning approaches in predicting material properties.

In Singapore’s hot and humid climate, air-conditioning and mechanical ventilation (ACMV) systems account for over 60% of commercial building energy consumption, driving efforts to enhance energy efficiency through predictive control strategies such as model predictive control (MPC) to overcome the limitations of conventional reactive building automation systems. This paper presents a multizone MPC system designed to optimize energy consumption and thermal comfort in a commercial building’s ACMV system in Singapore. The system was implemented in a multi-use test building with real occupancy and a deployment area of approximately 850 m², partitioned into six learning zones, two office spaces, and three open spaces. The ACMV system serving the deployment area consisted of two primary air-handling units and 16 fan coil units, where chilled water was supplied to the cooling coils, and conditioned air was distributed through motorized diffusers. To facilitate predictive control, data-driven thermal prediction models were developed for each zone using a non-linear autoregressive exogenous network with exogenous inputs trained on historical data and disturbances. Thermal comfort optimization was guided by the predictive mean vote, which was targeted at 0, representing thermal neutrality (as per ASHRAE 55 standards), and constrained within a range of −0.5 - 0.5. Performance comparisons demonstrated that the MPC system achieved over 42% energy savings compared to the original thermostat-based control while enhancing thermal comfort. Despite its advantageous control performances, challenges for large-scale deployment remain, including implementation costs, scalability, and model accuracy. Future work can address these challenges by developing comfort models that leverage existing building sensors.

Accurate prediction of multiaxial fatigue life was crucial for structural integrity assessment, yet the variability in material responses under complex loading paths made it challenging for both classical and data-driven models to achieve high accuracy. To address this issue, a contrastive learning-based framework was proposed in this study, enabling the construction of more generalized low-dimensional feature representations across different loading paths. This framework enhanced the robustness of fatigue life prediction without relying on mechanical assumptions. Experimental validation demonstrated that, compared to existing methods, the contrastive learning model learned more suitable feature encodings, significantly improving prediction performance. This framework provided a reference solution for engineering applications requiring reliability assessment under multiaxial stress conditions.