Figure 8 - uploaded by Guo Dong Goh
Content may be subject to copyright.
Source publication
Additive manufacturing (AM) or 3D printing is growing rapidly in the manufacturing industry and has gained a lot of attention from various fields owing to its ability to fabricate parts with complex features. The reliability of the 3D printed parts has been the focus of the researchers to realize AM as an end-part production tool. Machine learning...
Context in source publication
Context 1
... use of ML in 3D printing covers a wide spectrum of applications, ranging from design for 3D printing, to process optimization, to in-situ monitoring. ML has demonstrated to be a powerful tool to perform data-driven numerical simulation, design features recommendation, real-time anomaly detection, and cybersecurity (Fig. 8). The ML has shown to outperform conventional optimization methods such as second-order polynomial regression especially dealing with high dimensionality data. More advanced ML algorithm techniques and higher computational power in the future would see realtime in-situ monitoring and feedback control to be realized. Potential ...
Similar publications
The need for a vehicle to perceive information about the external environmental as an independent intelligent individual has grown with the progress of intelligent driving from primary driver assistance to high-level autonomous driving. The ability of a common independent sensing unit to sense the external environment is limited by the sensor’s own...
Citations
... AI is poised to have a significant impact on data access and analysis. This impact will extend beyond material selection and optimization to include process improvement and mistake elimination in bio printing [202]. AI will create a lot of chances of 5D printing, where the fifth dimension will play a role in enhancing the efficacy of applications of technology for additive manufacturing in biomedicine [203]. ...
Background
Nattokinase is pivotal in managing cardiovascular disease (CVD) and it serves as a natural adjunct in treating circulatory conditions, offering reduced risks and enhanced cost-effectiveness. This critical review explores the wide-ranging therapeutic effects of Nattokinase, particularly its capabilities in fibrinolysis, blood pressure regulation, atherosclerosis prevention, and decreasing lipid levels but multiple challenges still linger regarding its stability, bioavailability, and the processes for clinical applications.
Main text
The primary challenges in utilizing microbial therapeutic enzymes include instability, toxicity, and in vivo retention. Consequently, we propose enzyme immobilization of Nattokinase as a viable solution. Approaches including adsorption, encapsulation, and cross-linking can boost the strength, stability, reusability, output, and catalytic function of these enzymes. Our review outlines the current landscape of fibrinolytic enzymes, particularly nattokinase, as therapeutic agents for cardiovascular diseases, alongside the advantages and disadvantages of various smart Nano carriers for enzyme distribution.
Conclusion
The advancements in immobilized Nattokinase formulations for CVD treatment and the utilization of sophisticated computational methodologies to address current challenges and future prospects in this domain are discussed. Furthermore, studies conducted on humans show that larger doses of Nattokinase produce better results in treating atherosclerosis and high lipid levels. This review focus on the beneficial effects of Nattokinase, highlighting the shortcomings in its application, and advocates for more extensive studies to fully leverage its treatment potential and to set forth evidence-informed protocols for its use in cardiovascular health.
... Machine learning (ML) has emerged as a transformative tool for addressing such challenges (Alizamir et al., 2025;Van-Canh et al., 2025;Abdellatief et al., 2025;Hu et al., 2025). By analyzing large datasets, ML techniques can identify intricate patterns and relationships that are difficult to capture using traditional approaches (Goh et al., 2021). For predicting CS and FS of 3DP-FRC, Kang et al. (2021) developed eleven ML models for predicting CS and FS of steel fiber-reinforced concrete (SFRC). ...
... While FEM is grounded in physics-based equations and allows mechanistic interpretation of stress and deformation, it requires detailed knowledge of material parameters and can be computationally expensive. In contrast, ML approaches rely on large datasets to recognize patterns and make predictions without explicit physical modeling, enabling rapid optimization once trained [177]. However, ML models are often less interpretable and depend heavily on data quality and quantity. ...
As an emerging additive manufacturing technique, three-dimensional bioprinting enables precise control over the fabrication of tissue replacements, surpassing the limitations of conventional biofabrication methods. However, the successful production of functional bioprinted constructs remains challenging due to the complex interplay of numerous process parameters. The finite element method (FEM) has proven to be a powerful computational tool in biomedical research, offering a means to simulate and optimize various aspects of the bioprinting process. This review systematically examines the diverse applications of FEM across the three key stages of extrusion-based bioprinting—pre-printing, printing, and post-printing—one of the most widely adopted bioprinting technologies. FEM enables the prediction and optimization of tissue construct properties before fabrication by simulating both in vitro and in vivo loading conditions, providing valuable insights into critical yet experimentally inaccessible parameters, such as internal stress distributions and mechanical deformations. By enhancing the understanding of these factors, FEM contributes to the development of mechanically stable and biologically functional bioprinted structures. Additionally, FEM-driven simulations facilitate the optimization of bioprinting parameters, reducing material consumption, improving reproducibility, and accelerating the design process. Despite its significant contributions, existing FEM tools remain constrained in their ability to capture the highly dynamic and multi-scale nature of bioprinting completely. Future advancements should enhance the accurate representation of real-time cell-matrix interactions, bioink dynamics, and the progressive maturation of bioprinted constructs. By refining FEM simulations and embedding them into adaptive bioprinting workflows, this computational approach has the potential to drive transformative innovations in tissue engineering, regenerative medicine, and organ fabrication.
... Machine learning, in particular, can be utilized to enhance the predictive capabilities of 3D printing systems, providing critical insights into potential failures before they occur [193]. Predictive maintenance is one area where AI has shown significant promise, as AI algorithms can analyze historical data from printers to predict when a component may fail or when a maintenance procedure is due [194]. ...
The integration of 3D printing technologies in automated manufacturing systems marks a significant progression in the manufacturing industry, enabling elevated degrees of customization, efficiency, and sustainability. This paper explores the synergy between 3D printing and automation by conducting a critical literature review combined with case study analysis, focusing on their roles in enhancing production lines within the framework of Industry 4.0 and smart factories. Key opportunities presented by this integration include mass customization at scale, reduced material waste, and improved just-in-time manufacturing processes. However, challenges related to quality control, scalability, and workforce adaptation remain critical issues that require careful consideration. The study also examines the emerging role of hybrid manufacturing systems that combine additive and subtractive processes, alongside the growing need for standardized regulations and frameworks to ensure consistency and safety. Case studies are highlighted, showcasing real-world applications of automated 3D printing technologies and AI-driven print optimization techniques. In conclusion, this paper contributes to advancing the scholarly understanding of automated 3D printing by synthesizing technical, organizational, and regulatory insights and outlining future trajectories for sustainable and agile production ecosystems.
... In this context, artificial intelligence (AI) and deep learning have emerged as powerful tools to automate and optimize this process [15]. Instead of manually testing each parameter, machine learning models can learn from previous data to predict the optimal printing parameters based on the desired mechanical properties [15][16][17]. ...
... In this context, artificial intelligence (AI) and deep learning have emerged as powerful tools to automate and optimize this process [15]. Instead of manually testing each parameter, machine learning models can learn from previous data to predict the optimal printing parameters based on the desired mechanical properties [15][16][17]. This not only saves time and resources but also improves prediction accuracy, thereby enhancing the quality of printed products [6]. ...
This research focuses on optimizing the 3D printing process by adjusting key printing parameters, including temperature, speed, and layer thickness, to achieve desired mechanical properties such as yield strain, Young’s modulus, and peak load. A variant of the autoencoder (AE) network is applied to capture the relationship between these printing parameters and the mechanical properties. The proposed model demonstrates accurate bidirectional prediction capabilities, allowing for the estimation of mechanical properties based on printing parameters and vice versa. In particular, by incorporating a guided encoder-decoder training approach, the model prediction accuracy improved significantly, with the R² score increasing from 90.87 to 93.56%. Furthermore, this study addresses challenges in inverse problems, particularly cases where multiple configurations of mechanical properties correspond to the same set of printing parameters. By utilizing a guided model, ambiguity in parameter prediction is effectively reduced, enhancing the reliability of machine learning-based optimization. These findings contribute to the advancement of additive manufacturing by providing a robust framework for improving the precision and efficiency of 3D printing processes.
... With the advancements in artificial intelligence, machine learning techniques have also been widely utilised in enhancing 3D printing processes and 3D-printed parts [10,74,75]. Machine learning techniques can aid in determining optimal printing parameters such as layer thickness and printing speed that enhance mechanical properties, which can be incorporated into finite element models to improve prediction accuracy. ...
... CI models have been used to predict drug release based on input variables and to design geometries that achieve desired release rates. While there are existing reviews on the integration of CI and 3D printing, they broadly cover CI's role in various stages of the 3D printing process, such as material selection, printability optimization, quality control measures, and structural design [39][40][41][42]. This review specifically emphasizes on CI's roles in customizing drug release profiles. ...
... Semi-supervised learning combines labeled and unlabeled data, which is practical when labeling large datasets is costly. This approach achieves higher accuracy than unsupervised learning alone as there is a small amount of labeled data [36,41]. ...
... The input layer's neurons correspond to the number of process parameters in the study, while the output layer's neurons represent the number of properties to be optimized, usually one or two. The hidden layer, positioned between the input and output layers, typically has more neurons than the input layer to enable effective pattern recognition and learning [41]. ...
Computational intelligence (CI) mimics human intelligence by expanding the capabilities of machines in data analysis, pattern recognition, and making informed decisions. CI has shown promising contributions to advancements in drug discovery, formulation, and manufacturing. Its ability to analyze vast amounts of patient data and optimize drug formulations by predicting pharmacokinetic and pharmacodynamic responses makes it a very useful platform for personalized medicine. The integration of CI with 3D printing further strengthens this potential, as 3D printing enables the fabrication of personalized medicines with precise doses, controlled-release profiles, and complex formulations. Furthermore, the automated and digital capabilities of 3D printing make it suitable for integration with CI. CI has proven useful in predicting material printability, optimizing drug release rates, designing complex structures, ensuring quality control, and improving manufacturing processes in 3D printing. In the context of customizing drug release from 3D-printed products, CI techniques have been applied to predict drug release from input variables and to design geometries that achieve the desired release profile. This review explores the role of CI in customizing drug release from 3D-printed formulations. It provides overview of limitations of 3D printing; how CI can overcome these challenges, and its potential in customizing drug release; a comparison of CI with other methods of optimization; and real-world examples of CI integration in 3D printing.
... Although AI integration offers significant potential in enhancing macroscale 3D printing techniques, [43,241,242] its exploitation at the nanoscale level remains in its infancy. Such transformative opportunities address critical challenges, including improving process control, print quality, and scalability. ...
3D nanoprinting presents a fundamentally different approach (bottom‐up) compared to traditional nanolithography (top‐down), enabling the fabrication of nanostructures with greater material versatility and complex spatial geometries. Initially developed for macroscopic devices fabrication, 3D printing is now progressing toward the nanodevices production with active functionalities. This review explores cutting‐edge 3D printing technologies for nanoscale materials, emphasizing key achievements, foundamental principles, and persisting technological challenges. This review discusses potential opportunities in material selection, electronic co‐design, device integration, scalability, and essential steps toward commercialization. Among the numerous insightful reviews on 3D printing, this review aims to provide a more detailed discussion of the perspective and existing gaps in practical implementation, grounded in current technological capabilities. Furthermore, the future impact of 3D nanoprinting on academia and industry is explored.
... Globally, 3D printing development has been growing rapidly in recent years, and its global market value increased to USD 14 billion in 2020 from USD 7.4 billion in 2017 [1]. This figure is predicted to reach USD 56 billion in 2027 [2,3]. These trends point toward the increasing importance of 3D printing technology in our daily lives, which has been widely used in many industries-for example, manufacturing, building, aerospace and medical devices, and environmental research, as well as common applications in offices, small workplaces, and homes [4][5][6][7]. ...
The utilization of 3D printing releases a multitude of harmful gas pollutants, posing potential health risks to operators. Materials extrusion (ME; also known as fused deposition modeling (FDM)), a widely adopted 3D printing technology, predominantly employs acrylonitrile–butadiene–styrene (ABS) and polylactic acid (PLA) as printing materials, with the respective market shares of these materials reaching approximately 75%. The extensive usage of ABS and PLA during the ME process leads to significant volatile organic compound (VOC) emissions, thereby deteriorating the quality of indoor air. Nevertheless, information regarding the emission characteristics of VOCs and their influencing factors, as well as the toxicological impacts of the printing processes, remains largely unknown. Herein, we thoroughly reviewed the emission characteristics of VOCs released during ME printing processes using ABS and PLA in various printing environments, such as chambers, laboratories, and workplaces, as well as their potential influencing factors under different environmental conditions. A total of 62 VOC substances were identified in chamber studies using ABS and PLA filaments; for example, styrene had an emission rate of 0.29–113.10 μg/min, and isopropyl alcohol had an emission rate of 3.55–56.53 μg/min. Emission rates vary depending on the composition of the filament’s raw materials, additives (such as dyes and stabilizers), printing conditions (temperature), the printer’s condition (whether it has closure), and other factors. Additionally, we reviewed the toxicological concerns associated with hazardous VOC species commonly detected during the ME printing process and estimated cancer and non-cancer risks for users after long-term inhalation exposure. Potential health hazards associated with inhalation exposure to benzene, formaldehyde, acetaldehyde, styrene, and other substances were identified, which were calculated based on concentrations measured in real indoor environments. This study provides valuable insights for future research on the development of ME printing technologies and offers suggestions to reduce VOC emissions to protect users.
... A pioneering example for facilitating a data-driven approach to quantifying the quality of the surfaces of building elements, i.e. roughness, smoothness, etc., with expert-informed feature generation can be seen in [11,12]. ML-based approaches have the potential to predict and visualize what we can actually build, and existing literature points towards the potentials and constraints [13,14,15]. However, they mainly target the usage of modular systems such as wood assembly [16] or measurement on a series of 3D printed panels to define the optimal surfaces for acoustics [17]. ...
This work proposes a Graph Neural Network (GNN) modeling approach to predict the resulting surface from a particle based fabrication process. The latter consists of spray-based printing of cementitious plaster on a wall and is facilitated with the use of a robotic arm. The predictions are computed using the robotic arm trajectory features, such as position, velocity and direction, as well as the printing process parameters. The proposed approach, based on a particle representation of the wall domain and the end effector, allows for the adoption of a graph-based solution. The GNN model consists of an encoder-processor-decoder architecture and is trained using data from laboratory tests, while the hyperparameters are optimized by means of a Bayesian scheme. The aim of this model is to act as a simulator of the printing process, and ultimately used for the generation of the robotic arm trajectory and the optimization of the printing parameters, towards the materialization of an autonomous plastering process. The performance of the proposed model is assessed in terms of the prediction error against unseen ground truth data, which shows its generality in varied scenarios, as well as in comparison with the performance of an existing benchmark model. The results demonstrate a significant improvement over the benchmark model, with notably better performance and enhanced error scaling across prediction steps.
