With the rapid development of artificial intelligence (AI) technologies, deep learning-based structural health monitoring (DeepSHM) methods have gained significant attention. However, their black box nature often limits interpretability and trust. The field of Explainable AI (XAI) aims to address this by enhancing model transparency and reliability through human-comprehensible explanations. This study investigates the use of XAI algorithms in interpreting a 1D convolutional neural network (1D CNN) developed for Lamb wave monitoring of bolt-loosening detection in multi-bolted double-layer aluminum plates under varying temperatures. Four existing XAI algorithms were employed, including Sensitivity Analysis, Deep Taylor, Gradient-weighted Class Activation Mapping (Grad CAM) and Guided Grad CAM. In addition, this paper introduces two new XAI methods, Smooth Simple Taylor and Deep Grad CAM as an enhancement of the Simple Taylor and Grad CAM methods, respectively. These six XAI algorithms were used to establish the relation between the 1D CNN model parameters and the input vector. The results were evaluated for their effectiveness in comparison to the physical insights of the input vector using two proposed methods, namely the Correlation Coefficient with Residual Signal and the Residual Signal Weighted Importance Score Ratio. The results of the evaluation methods, in conjunction with Infidelity, Sense sum, and Sanity check, were utilized to rank the performance of the six XAI algorithms. The rankings were consistent in both simulation and experiment data sets, and the newly proposed XAI algorithm, Smooth Simple Taylor, appeared to be the best in both data sets. Overall, this research establishes a novel approach to using XAI algorithms to enhance the explainability of AI in practical engineering applications.

Abstract: New manufacturing techniques like 3D printing are under development, and they need monitoring methods to ensure the quality of the manufactured parts. Artificial Intelligence has outperformed traditional methods in the monitoring process and has shown high potential in recent years. New approaches in Artificial Intelligence, particularly Neural Architecture Search (NAS), have unlocked the potential for automated design of high-performance and resource-efficient deep learning models. In this work, we propose a training-based, low-fidelity NAS framework to systematically discover optimal architectures for regression tasks. Leveraging 8,610 candidate topologies, we trained models on only of the data for 10 epochs, enabling faster execution and selection of the architecture using low-fidelity information. The dataset belongs to Laser Powder Bed Fusion (LPBF), which is a manufacturing method that is still not well mastered and requires many trials before obtaining a satisfactory result. To cope with this issue, we developed a NAS algorithm to design a lightweight AI model (an architecture with a low number of parameters) to predict the process parameters from video information to ensure having the same printing parameters in action. The ultimate goal is then to embed the AI model in a low-latency feedback control loop that enables on-the-fly supervision of the printing process. The final designed architecture is based on 3-dimensional convolutional neural networks. The final AI models are 3–30 times lighter than off-the-shelf ones, while maintaining almost the same accuracy. This shows the potential of our methods when dealing with regression tasks in an industrial case study.

This review focuses on the application of computational fluid dynamics (CFD) in pulmonary drug delivery, particularly for treating asthma and COPD with pharmaceutical aerosols via dry powder inhalers (DPIs). Aerosol drug delivery effectiveness relies on accurate assessment and prediction of particle deposition in the respiratory system. This method is crucial due to the high number of pulmonary disease cases, efficient lung absorption capabilities, lower dosage required, and reduced systemic side effects compared to oral medications. Given the limits of in vivo and in vitro methods, CFD modeling has advanced rapidly over 20 years, especially when combined with discrete phase model (DPM) and discrete element method (DEM) approaches. CFD simulations can correctly account for a wide range of realistic or idealized parameters using numerical solutions of particle and airflow transport equations. Achieving accurate simulations requires avoiding simplifications and approximating real-world conditions, which will soon be more possible with advancing computing tools. The research aims to review numerical modeling of pulmonary drug delivery along the mouth-to-lung pathway, encompassing governing equations, forces, boundary conditions, the influence of lung geometry on CFD modeling, the effects of powder characteristics on aerosolization and pulmonary deposition, validation of computational results with in vitro/in vivo data, and a discussion of current challenges and future prospects.

Machine Learning (ML) has revolutionized various fields, enabling the development of intelligent systems capable of solving complex problems. However, the process of manually designing and optimizing ML models is often time-consuming, labor-intensive, and requires specialized expertise. To address these challenges, Automatic Machine Learning (AutoML) has emerged as a promising approach that automates the process of selecting and optimizing ML models. Within the realm of AutoML, Neural Architecture Search (NAS) has emerged as a powerful technique that automates the design of neural network architectures, the core components of ML models. It has recently gained significant attraction due to its capability to discover novel and efficient architectures that surpass human-designed counterparts. This manuscript aims to present a systematic review of the literature on this topic published between 2017 and 2023 to identify, analyze, and classify the different types of algorithms developed for NAS. The methodology follows the guidelines of Systematic Literature Review (SLR) methods. Consequently, this study identified 160 articles that provide a comprehensive overview of the field of NAS, encompassing discussion on current works, their purposes, conclusions, and predictions of the direction of this science branch in its main core pillars: Search Space (SSp), Search Strategy (SSt), and Validation Strategy (VSt). Subsequently, the key milestones and advancements that have shaped the field are highlighted. Moreover, we discuss the challenges and open issues that remain in the field. We envision that NAS will continue to play a pivotal role in the advancement of ML, enabling the development of more intelligent and efficient ML models for a wide range of applications.