Additive Manufacturing (AM) has been in use among several industries, including aerospace and automotive industries, long before it was adopted in healthcare manufacturing. Its adoption in healthcare manufacturing has witnessed several challenges, including timeliness, amount of resource use, and customization challenges, not forgetting the respective challenges seen in fabricating medical devices and printing tissue organs. The study identifies that the use of parallel computing techniques provides solutions to most of these challenges. The proposed similar computing solutions include parallel algorithms for slicing 3D medical models, real-time monitoring of 3D printing processes, and optimizing support structures for complex medical device designs. Besides, there are advanced parallel computing techniques such as Graphics Processing Unit (GPU), cloud, and edge computing that can be used to address some of these challenges.
Introduction
Technological inventions and innovations are considerably controlling the systems and procedures applied in different sectors around the globe. The healthcare sector is one of those that have embraced the intense use of technological inventions and innovations, and one of the goals is achieving more for less. The use of the Advanced Manufacturing technique (AM) has demonstrated the ability to revolutionize the whole healthcare manufacturing and help achieve more for less (Memeti et al., 2019). Compared to the traditional subtractive machine technique, AM has demonstrated the capacity to print metals and produce more complex designs as well as ensure layer-by-layer building of materials while at the same time ensuring less waste on materials. AM is promising many revolutions within the healthcare sector, and some of these include customized products for joint implants, hearing aids, and dental works.
Additionally, Duraes et al. (2020) observe that AM can be critical in printing customized medication on demand for individual patients. If well exploited, it can also help in producing body parts on the market. In light of these, the current study is set to explore the application of parallel computing in healthcare additive manufacturing. It will explore the various similar computing techniques applicable in AM and how well each of these corresponds to AM’s goal of eliminating time and resource-efficiency challenges witnessed in contemporary healthcare manufacturing.
AM allows doctors to use CAD files in prescribing medications, a technological innovation that, if well exploited, can allow patients to print such at the pharmacy shop. In the future, this will enable the printing of medicines at the pharmaceutical shops in accordance with what the customer demands. The customers will also be able to print their treatment by adding the right ingredients and inserting the command to print. 3D printing is also the last step in medical imaging (Maksum et al., 2022). These steps include image acquisition, image post-processing, and 3D printing. In image acquisition, data is collected from Computer Technology (CT) or Magnetic Resonance Imaging (MRI) and stored in Digital Imaging and Communications in Medicine (DICOM) data. In image post-processing, the DICOM data is converted into STL file format In the form of thin layers. The final step is sending the file to a printer for printing, 3D printing (Zhou et al., 2019). It is also possible to fabricate customized prosthetics using CT-scanned data. However, children always reject such prosthetics due to their size and discomfort, a problem that can be solved using AM. Figure 1 highlights the various applications of additive manufacturing in healthcare.
Figure 1 Areas of application of additive manufacturing in healthcare adopted from Ramola Yadav & Jain (2019).
Specification of the Problem
In a world where all stakeholders in the healthcare sector are grappling to help realize quality healthcare provision in a manner that meets equitability and universality principles. Compared to the traditional subtractive manufacturing technique, additive manufacturing is a promising option for implementing lean production. However, in healthcare additive manufacturing, the primary problem is the time and resource-intensive nature of producing custom implants, prosthetics, and medical devices using additive manufacturing techniques (Permann et al., 2020). Additive Manufacturing (AM) is currently confronted with difficulties because of its resource- and time-intensive supply chain. Reaching the full potential of AM requires addressing these inefficiencies. AM may be made more efficient by significantly reducing the time and resources needed by streamlining the supply chain operations.
On the other hand, moving from traditional pharmaceutical frameworks to self-service technologies and on-demand drug printing within the AM domain may provide early financial hurdles. For AM to be widely adopted, cost dynamics must be balanced with supply chain optimization. In addition to improving AM’s speed and resource efficiency, overcoming these obstacles will open the door for creative and economical industrial and healthcare solutions.
On the same note, patients often require personalized medical solutions, and traditional manufacturing processes are slow and costly for producing such items. However, the use of AM faces specific limitations. AM is currently utilized in surgery with a considerable reduction in printing costs, involving almost 95% reduction in the cost (Jihong et al., 2021). The inconsistency between 3D-printed models and actual anatomical features is a significant disadvantage, even though these models improve pre-surgical training. This discrepancy may restrict the efficacy of training by affecting simulation accuracy. Moreover, the two- to three-day turnaround time for surgery models produced using additive manufacturing (AM) presents a problem for emergencies where prompt decision-making is essential. One of the most critical factors in promoting the use of 3D-printed models in surgical education and emergency preparation is striking a balance between the advantages of realistic training and the real-world limitations of time and precision.
There are also challenges emerging from the use of AM in printing tissues and organs. At the moment, AM is not capable of replicating the complex nature of vascular networks in tissues and organs. Besides, AM can only print tissues in millimeters. Increasing the size results in an exponential increase in cost and time required. Bajaj et al. (2020) also note some limitations in using AM in medical imaging. Making a 3D model is step-by-step, with the likelihood of errors at each step. The final picture’s accuracy is greatly affected by mistakes made during the 3D printing process, which causes the image to deviate from the original 3D model. These differences depend on factors including the printer’s characteristics, the scale of the model, and the software used. Larger models may have more apparent faults, and software selection can create algorithmic problems. Differences in the features of the printer may also cause distortions. In order to achieve an accurate representation of the desired design and ensure precision in 3D printing, it is imperative to address and minimize these issues.
The application of Additive Manufacturing (AM) in the production of medical equipment poses issues about adherence to critical specifications. According to Vaneker et al. (2020), overcoming these obstacles calls for the use of reliable software systems to improve the accuracy and precision of device development. Reaching a higher level of uniformity is necessary in order to reduce possible differences that may result from the variety of 3D-printed products. Achieving a balance between innovation and minimal standards compliance is essential to guarantee the effectiveness and safety of medical equipment manufactured using additive manufacturing. This changing environment emphasizes the necessity of ongoing improvements in software functionality and standardization procedures in order to successfully negotiate the complexities of using AM in the field of medical device manufacture.
Justifications for Using Parallel Processing To Solve the Problem
According to Wei et al. (2021), parallel computing is a technique in which significant problems are divided into smaller, independent parts. These parts, similar and separate, allow for simultaneous operation by using shared memory across several CPUs. The main goal is to increase processing power, which will improve the ability to process applications and solve problems. This method works incredibly well for solving Additive Manufacturing (AM) obstacles. Parallel computing, in particular, improves customization by processing complex geometries and patient-specific data effectively. Parallel computing has become a transformational force in addressing the challenges posed by AM. It improves efficiency and scalability by dividing work across processors effectively, which helps to overcome the challenges presented by complicated geometric structures and specially designed healthcare applications. This discovery establishes parallel computing as a robust framework that opens up new opportunities for additive manufacturing (AM). The technique is driven forward by its capacity to address AM-related difficulties, which holds promise for improvements in accuracy, speed, and flexibility. Consequently, parallel computing opens up previously unimaginable possibilities for intricate manufacturing and customized healthcare solutions while also optimizing existing additive manufacturing processes and stimulating creativity. An age of unprecedented capabilities and increased efficiency is being heralded by the industry’s paradigm change marked by the synergy between AM and parallel computing.
Personalization
Healthcare often requires custom-designed implants or prosthetics. Parallel processing can speed up the production of personalized medical devices. Data flow and communication are at the core of efficiencies and effectiveness realized from parallel computing. In the functionality of AM, it is imperative to meticulously consider data transfer protocols and communication methods to facilitate seamless data exchange and coordination among processing units. Real-time modifications are critical in the dynamic field of healthcare additive manufacturing, which makes compelling data flow imperative. This need also includes the smooth transfer of complex model data, real-time monitoring data, and printing process feedback—all coordinated within the architectural framework (Tartici et al., 2019). It becomes critical to optimize these channels of communication since it enables quick and coherent data processing. This optimization improves the overall quality and productivity of the manufacture of customized medical products in addition to guaranteeing a prompt response to changing needs. Here, the union of state-of-the-art technology and efficient data interchange forms the basis of a flexible and responsive additive manufacturing ecosystem in healthcare, fostering innovations that have a direct bearing on treatment outcomes and patient care.
Similarly, distributed processing is another pivotal facet of architectural design in parallel computing for healthcare additive manufacturing. It entails deploying multiple computing nodes or processing units across a network to enhance efficiency. In this context, each node undertakes defined tasks, like rendering segments of intricate 3D models or simulating real-time printing processes. Architectural design is effective because it strategically optimizes the distribution of processing tasks, minimizing the amount of data transferred and communication overhead. As Rauch et al. (2018) point out, this simplified parallelization of complex calculations becomes essential. The healthcare industry is changing thanks in large part to this dispersed strategy, which makes it possible for it to effectively satisfy the customized needs of the medical product manufacturing process. This guarantees scalability and improves operational efficiency, putting the sector in a position to precisely and nimbly negotiate the challenges of individualized healthcare. Thus, the success of the architecture serves as a pillar for transforming healthcare procedures in preparation for a future that is more flexible and responsive.
Fang et al. (2019) further suggest that the critical nature of this field necessitates the integration of fault tolerance mechanisms within the design. This encompasses redundancy in hardware components, strategic data backup strategies, and robust error recovery protocols. In order to ensure the stability and smooth functioning of the production process, a complete set of procedures must be put in place to reduce the likelihood of faults in medical equipment. When it comes to the production of customized medical products, where accuracy is crucial, a fault-tolerant design is not only required but also a basic need. These steps are combined to provide a strong foundation that is necessary to maintain the exacting quality and safety requirements that are essential to the manufacturing of medical devices. This puts the needs of the patients first by guaranteeing that the final goods fulfill the strict standards of the healthcare sector. Incorporating rigorous quality control and safety procedures together not only promotes regulatory compliance but also develops an excellence culture in the production process. As a result, this coordinated effort enhances the dependability and effectiveness of medical devices, highlighting their critical role in improving healthcare and demonstrating an unrelenting dedication to patient-centric results.
Complex Geometry
Medical devices may have complex geometries that are computationally intensive to produce. Parallel computing can handle these intricate designs more efficiently. Parallel computing in healthcare additive manufacturing necessitates a carefully crafted architectural design to maximize its benefits and ensure optimal performance. It specifically plays a pivotal role in enhancing scalability, speed, and precision. Among the myriad of considerations, high-performance computing clusters top the list. These clusters consist of interconnected computers equipped with multiple processors or cores. Clusters are essential to healthcare additive manufacturing because they distribute and parallelize complex calculations such as the rendering and slicing of 3D models. Scalable clusters are required as production needs rise, and architectural design becomes critical. This guarantees smooth adjustment to changing needs, maximizing effectiveness and output. Scalability is a crucial factor in the design and deployment of additive manufacturing clusters in the healthcare industry since the dynamic nature of the field requires a strong foundation that can adapt to changing needs.
Parallel computing also has in place a load-balancing mechanism. The load balancing mechanism is critical in guaranteeing an even distribution of workload among processing units. Ideally, tasks in healthcare additive manufacturing are diversified, implying that the load can fluctuate significantly. In response to these variations, the designer must pay unique attention to the process and integrate load-balancing algorithms into the architectural design (Umang et al., 2022). Algorithms are essential for effectively utilizing computer resources and planning how best to use them. Algorithms maximize the performance of the entire system by avoiding bottlenecks and promoting unit synergy. Comparable to a conductor arranging a symphony, this strategic optimization makes sure every component plays in unison to attain optimal effectiveness. As a result, the full potential of computer resources is realized, enabling smooth operations. This improves the way each duty is completed individually and simplifies the way the system works as a whole. Algorithms are the maestros of computer processes; they increase computing system capabilities by streamlining operations, optimizing efficiency, and facilitating more efficient use of resources.
Patient-Specific Data
Healthcare data is susceptible and often requires real-time processing. Parallel processing can ensure the security and rapid processing of patient-specific data for customization (Xi et al., 2018). Memory hierarchy and caching are also paramount considerations due to their irreplaceable role in minimizing data access latency and fostering overall computational speed. Considering the fact that healthcare additive manufacturing entails large volumes of data, it becomes crucial to consider and embrace efficient caching mechanisms. Shukur et al. (2020) explain that the cache memory stores frequently accessed data close to processing units, thus reducing the need for repetitive data retrieval from distant memory sources. For computers to process massive datasets effectively, memory hierarchy and caching must be strategically implemented. When applied to medical applications, this method guarantees the smooth management of large datasets, leading to increased accuracy and productivity. The technology can quickly access and alter vital medical data by streamlining data storage and retrieval methods. This improves diagnosis accuracy and streamlines healthcare procedures. This advanced use of memory resources is critical to the development of medical technology, which in turn improves patient care and the effectiveness of the healthcare system as a whole.
Finally, security remains a paramount concern in parallel computing in the healthcare context. Given the involvement of patient-specific data and sensitive medical information, prioritizing security is imperative. The architectural design must incorporate robust security measures, including advanced encryption techniques, stringent access control mechanisms, and secure data transfer protocols. A strong security architecture is essential in the field of parallel computing to protect confidential data from breaches or unwanted access. The security, availability, and integrity of patient data are given top priority by this collective shield, which functions flawlessly throughout the computing process. This strengthened defense goes beyond simple regulatory compliance; it is essential to maintaining the confidence and security of people whose health depends on the accuracy and safety of the creation of customized medical products. The careful planning and implementation of these security measures not only protect against possible threats but also emphasize a dedication to the moral and responsible management of private health information, building a strong and reliable basis for the development of personalized medicine.
Extended discussion about solutions to the challenges faced in AM problem
Parametric computation methods are of the most significant importance when it comes to tackling obstacles in Additive Manufacturing (AM). In the context of intricate printing tasks, parallel algorithms explicitly designed for slicing 3D medical models improve performance by distributing computational burden across multiple processors. The utilization of parallelization in real-time monitoring of 3D printing processes facilitates concurrent data analysis and feedback, thereby ensuring production-level precision and error detection. Furthermore, the optimization of support structures for intricately designed medical devices necessitates the execution of complex computations, which can be accelerated via parallel computing. This, in turn, improves the overall design process. All of these techniques work in tandem to enhance the efficiency of additive manufacturing processes, decrease the duration of processing, and promote progress in medical applications. The adoption of parallel computation in additive manufacturing (AM) represents a significant advancement in terms of performance, scalability, and innovation within the domain of medical 3D printing.
Parallel Algorithms for Slicing 3D Medical Models
The functionality of parallel algorithms for slicing 3D medical models is anchored on a Digital-Light-Processing (DLP) 3D printer. DLP utilizes mask projectors onto photopolymer resin to cure the liquid form resin into a solid model. AlZu’bi et al. (2020) identify that the mask projection is much faster and provides a uniform curing process, given that each layer is being projected at once. Though CAD files in the form of stereolithography (STL) are considered de facto in 3D printing, the STL files must always be sliced into layers of 2D contours before the projection process commences. Initially, there had been suggestions of different methods that can be used in this tiresome pre-processing stage involved in stereolithography, including adaptive slicing and uniform slicing methods. The many methods that are used before the contour projection stage have a significant impact on the final product that is printed. Known as “staircase” or “height” effects, these changes result from differences in the height of the segments. In order to mitigate these impacts and ensure a more exact and refined output in the 3D printing process, processes prior to contour projection must be executed with meticulousness. Therefore, developing these pre-projection techniques is essential to improving the overall precision and quality of the printed products.
STL was developed in 1987 for 3D CAD model conversion. The conversion process involved in STL utilizes the tessellation process responsible for forming multiple triangular facets of XYZ representing the surface feature of the geometry. There are always two types of STL formats, including ASCII STL and binary STL (Oyama et al., 2020). An ASCII STL always begins with a solid name syntax where the name can be regarded as optional and can be omitted with the use of spaces. It then follows a facet syntax with its average vector. The outer loop is essential for organizing vertex data in computer graphics and geometry processing. When the process gets into more complex geometric areas, every set of vertices (P1, P2, and P3) creates a unique facet of the model. The end facet indication indicates where a facet ends. A single file may have several facets; the number depends on the geometry’s intricacy and can frequently exceed thousands of facets (Oyama et al., 2020). As a result, the end of one aspect smoothly transitions into the beginning of the next, preserving the 3D model’s structural integrity. The end of the file signifies the end of sound name syntax, containing all of the geometric representation and creating a complete foundation for image processing and computer analysis.
On the other hand, Binary STL consists of an 80-character header that can be used as a comment. The number of triangles includes a 32-bit little-endian integer and 50 bytes for each triangular facet. A facet is defined in twelve 32-bit floating-point numbers. For each triangle, there are 48 bytes, followed by an unsigned integer in 2 bytes, which is described in the original documentation as the attribute byte count (Kamarianakis & Papagiannakis, 2021). Because STL files are binary, the attribute byte count takes on great significance in the software domain. These files, which are mainly represented as zeros and ones, require the byte count to remain fixed at zero. This constraint results from the software’s intrinsic inability to understand quantities larger than zero in this particular situation. However, because of their zero-centricity, these files are prone to data structure issues. These restrictions are made further worse by the lack of capability for texture, color, or other appearance features, which limits binary STL files to a harsh, monochrome depiction of geometry. Furthermore, the inability to create holes and gaps in layers leads to open loops in the cross-section, which makes it challenging to create complex, detailed structures in the virtual world.
The traditional STL algorithms require a considerable amount of processing time at the pre-processing stage, which makes it inefficient to meet the deformation control required in real-time data adjustments. At the same time, these traditional STL algorithms always have a numerical control system, which always results in the phenomenon of “tool shaking.” This involves the frequent acceleration and deceleration of the motor in the forming process. There have been several suggestions to help in solving these challenges (Oyama et al., 2020). The first of these is the use of a reverse ray tracing algorithm in slicing the 3D model in parallel before adjusting the slice direction of slice position in accordance with the height feedback information in the forming process. The second suggestion involves planning the scanning sequence of the zones in accordance with the area of the forming layer, the thickness of the part, and the number of zones to ensure that the temperature field distribution is uniform. Lai Wei (2021) merged these challenges by advancing a slicing algorithm based on reverse ray tracing. In this new framework, the generation of the STL file involves discretizing the 3D surface into triangular patches in accordance with the available triangle tolerance, and the three vertices of the outer patches are stored in counterclockwise order. The three vertices of the outer patches are also stored in a counterclockwise manner.
Real-Time Monitoring Of 3D Printing Processes
Real-time monitoring involves keeping the system ware and updated incessantly, feeding it with information. Among the activities involved include data collection from the production site, storing and processing the data to detect any anomaly and system performance issues, as well as resolving any of these issues. It provides low or next to zero latency streaming data constantly directly to the desired system. It is from this that the administrator can monitor the system and deduce any emerging problems before giving feedback to the system (Walker et al., 2019). The newly developed architecture enhances mitigation capabilities by enabling the quick distribution of data systems to individuals or automated platforms that are pertinent, all while maintaining continuous real-time monitoring. Its primary characteristic is the regular evaluation of real-time data over time intervals, which is essential for estimating efficiency and predicting trends. With the use of this temporal information, proactive modifications may be made to manufacturing processes to maximize performance and efficiency while reducing downtime. Through a smooth integration into current operations, this framework guarantees a prompt response to new challenges. It fosters a data-driven approach, creating a dynamic environment where performance enhancement and adaptability become fundamental components of the organizational culture.
Real-time monitoring has been identified by Paraskevoudis, Karayannis & Koumoulos (2020) to have high-accuracy sensors, cloud storage, optical fiber lanes, and low-cost IoT sensors, which makes it quite suitable for routine check-ups. The quantity and location of sensors in a system must be carefully considered, and the system’s complexity should be taken into consideration. Since the placement of these devices determines how accurate the data collected is, the integration of sensors should be a deliberate procedure. The notion of condition monitoring is introduced by Khan et al. (2021), who emphasize the proactive detection of possible faults prior to their materialization. The paradigm change from reactive to predictive maintenance emphasizes how important process monitoring is for predicting failures and proactively addressing new problems. Optimizing the placement of sensors not only improves system performance but also helps with preventive maintenance, which lowers downtime and eliminates possible hazards. Condition monitoring and selective sensor deployment encourage a proactive approach to system upkeep and fault avoidance.
Generally, the core benefits derived from utilizing real-time monitoring include better optimization of the system through the use of real-time, thereby helping in avoiding defects and errors. Better optimization is also realized by its ability to analyze trends and patterns as well as current events. Real-time monitoring also helps in reducing downtime along the process, together with any delay that may come with it, given that it tracks data for scheduling maintenance effectively (Walker et al., 2019). Besides, it allows for anomaly detection. The efficiency of creating personalized implants, prostheses, and medical equipment via additive manufacturing is greatly enhanced by real-time monitoring. It provides instantaneous insights into the production process, enabling firms to detect and address problems and guarantee smooth operations quickly. In addition to increasing overall productivity, this proactive strategy reduces waste and downtime, two critical elements in the resource-intensive field of medical device production. Accuracy and quality control are facilitated by real-time monitoring, which allows for quick modifications to match precise requirements. Manufacturers are able to provide customized medical solutions more quickly and affordably as a consequence of being able to optimize resource allocation, simplify workflows, and react quickly to deviations. Real-time monitoring integration becomes a vital tool in the dynamic world of additive manufacturing, promoting innovation and raising the bar for customized healthcare solutions.
Optimizing Support Structures for Complex Medical Device Designs
Optimization support structures for complex medical device design are anchored on four main phases, as elaborated in Figure 2. Though the framework is intensely utilized within the building industry, it can still be integrated into the healthcare sector. These four phases include product planning, design optimization, manufacturing optimization, and product validation. The product planning phase involves activities like the definition of objectives and constraints, modeling preparations for base design, feasibility analysis, and finite element analysis (Permann et al., 2020). One of the core features of mathematical-based optimization is the fact that solutions are considerably sensitive to the defined initial domains. This significantly impacts the mechanical performance during the validation process. To maximize the optimization of the planning phase, there is always the need to find the optimal material distribution and define a vast design domain while considering space limitations for part assembly. There is also the need to arrive at a suitable printing direction based on the building chamber size to help reduce the number of required iterations during manufacturing phase optimization.
The design optimization process involves an iterative process of design optimization strategy, design interpretation, and product simulation. The optimization phase is critical in solving the structural optimization process and validating the performance of the proposed design (Jihong et al., 2021). It also helps in finding the optimal material layout necessary for maximizing or minimizing an objective function representing a physical response of the system subjected to the constraints. Generally, natural systems are constantly subjected to several loading conditions that require the definition of multiple objective criterion optimization (Wei et al., 2021). One standard method for solving multi-objective issues is to use a single scalar function. To do this, different loading scenarios are given weighted coefficients, which makes it easier to convert the problematic problem into Pareto optimality. Using weighted integration to combine many criteria, this method simplifies the decision-making process. The Pareto-optimal solutions that emerge offer a wide range of trade-offs, enabling decision-makers to negotiate the complex trade-offs between conflicting goals. This strategy improves the efficiency and efficacy of tackling multifaceted issues in several domains, ranging from engineering to optimization difficulties, by synthesizing different parts into a unique scalar function.
Design interpretation describes the methodology involved in converting optimization results into a parameterized CAD model. It involves a density-filled representation where solid materials and void areas are accepted into the lightweight structure without affecting the structural response (Bajaj et al., 2020). Software platforms are essential for fine-tuning design optimization since they provide essential threshold tools. One of the most critical aspects of creating the border representation of an optimum design is controlling the density element display, which these tools do rather well. The software helps minimize superfluous density interfaces and fine-tune the structural integrity of the design by controlling the percentage of visible pieces. Product simulation also shows up as an essential element, offering a prognostic study of the suggested design’s physical performance. This analytical method explores a number of factors, including natural vibration frequencies, buckling mode, displacements, stresses, and absorbed strain energy. By enabling designers to make well-informed decisions, these simulations help ensure that the finished product satisfies strict requirements for dependability and efficiency.
The manufacturing optimization phase is a complex undertaking that includes critical components like support generation for printing modeling. This stage expands its scope to include additive simulation and the precise optimization of support structures. As emphasized by Vaneker et al. in 2020, the selected building orientation is of the utmost significance during the printing evaluation, as it directs the analysis of the STL design representation and facilitates generation. A multimodal strategy is used for enhancement and optimization in order to improve the complexities of this complicated procedure. Magic and other cutting-edge software solutions are essential for maintaining accuracy and productivity in the production process. In order to further improve the manufacturing process overall, innovative techniques like lattice infill and tree design optimization are used with this. These tactics work in concert to enhance the production optimization stage, leading to improved accuracy, higher output, and smoother incorporation of additive manufacturing innovations. The application of modern technologies not only optimizes the manufacturing process but also promotes flexibility in response to the changing terrain of additive manufacturing. This puts the process at the forefront of technological advancement and guarantees its ongoing significance in the ever-changing industrial landscape.
During the final phase of product development, the validation of the product becomes an essential process. During this stage, comprehensive quality inspection and stringent mechanical testing are conducted to verify the effectiveness of manufacturing techniques, post-processing methods, and performance attributes that were utilized regularly during development. After meticulously devising manufacturing processes and conducting physical simulations, the critical stage entails the printing of a prototype. The physical illustration functions as a concrete embodiment of the conceptual foundation, facilitating practical evaluation and improvement. The product validation phase functions as the final assessment, guaranteeing that the developed product not only fulfills but surpasses the anticipated criteria, signifying the progression from an abstract notion to a physical, superior-quality product.
Figure 2 Optimization support design adopted from Sbrugnera et al. (2021).
Further Discussion on Parallel Solutions to Similar Problems
The concept of additive manufacturing has been in use among several industries, including the automotive and aerospace industries, long before being adopted in the healthcare industries. However, the enhanced precision, efficiency, and effectiveness demanded in the healthcare sector have resulted in a refined framework on how AM can be used in realizing timely and proper usage of the available scarce resources. The parallel computing methods that were first applied in the aerospace and automotive industries to additive manufacturing in healthcare provide a revolutionary prospect for cross-industry applicability. Innovative advances in healthcare may be seamlessly applied to various fields to enable increased productivity, personalization, and quality control in industrial processes. The acceleration of production cycles and complex design simulations may be achieved through the optimization of parallel computing approaches, leading to increased accuracy and dependability. This hybridization of technical developments highlights the mutually reinforcing development of seemingly unrelated sectors and advances a shared trajectory toward advanced production techniques. Parallel computing will be a crucial component of diverse industrial innovation in the future as healthcare advancements echo across the aerospace and automotive industries, where knowledge convergence creates a synergy that redefines productivity norms.
Sharing insights and adapting parallel solutions across industries promotes innovation and accelerates the adoption of parallel computing techniques. Several sectors, including toys, furniture, and footwear, have challenges related to personalization and material utilization. Therefore, adopting creative solutions is crucial. Customers are becoming more demanding of unique product features, especially in the footwear industry. In order to satisfy this need, producers are utilizing cutting-edge technology to personalize patterns while guaranteeing the highest level of comfort in the finished product. These technologies not only increase productivity but also enable companies to meet the specific needs of individual customers. For companies looking to gain a competitive advantage, integrating these technologies is essential as the industry develops. It enables them to strike a delicate balance between providing comfortable, high-quality products and distinctive customization.
Advanced parallel computation techniques are of tremendous significance when it comes to tackling challenges that require more time and resources. The efficient execution of complex tasks is made possible through the efficient utilization of Graphics Processing Units (GPUs). Cloud computing enables the implementation of adaptable and scalable solutions through the utilization of distributed resources. By optimizing processing in closer proximity to data sources, edge computing effectively mitigates delay. Combined, these methodologies comprise an adaptable set of resources, enabling sectors to surmount impediments in computation time, promote groundbreaking concepts, and effectively confront complex challenges in domains including data analysis, artificial intelligence, and scientific simulations.
GPU Computing
According to Nickolls and Dally (2020), GPU computing entails maximizing unique hardware for parallel processing, tailored initially for rendering graphics in video games and other applications. This approach has been leveraged in the healthcare system to conduct intricate simulations and calculations required for additive manufacturing with utmost speed. Graphics Processing Units (GPUs) are essential for decomposing complex modeling and printing jobs into smaller, more manageable parts, which drastically cuts down on production time. This efficiency transcends traditional industries and finds surprising use in the healthcare sector, where it makes it possible to develop and print bespoke items quickly. Healthcare workers use GPUs to accelerate the creation of customized solutions, following the lead of the automotive and aerospace sectors. By accelerating innovation and improving precision, this technological synergy eventually improves patient care by producing tailored medical equipment and solutions more quickly. The ubiquitous integration of GPUs highlights their revolutionary influence across several domains, promoting an era where complex operations are optimized, and efficiency is enhanced.
Cloud Computing
Cloud computing is a unique paradigm that allows access to computational resources over the internet. This proficiency enables users to conduct data processing, storage, and analysis on remote servers. This perspective is a game changer in additive manufacturing as it offers unprecedented accessibility and flexibility in the long run. As such, the application of parallel processing in the cloud helps distribute the computational workload across various virtual servers or machines. This ideation has turned up as the preferred solution for healthcare organizations, especially those with limited on-site infrastructure. Through the system’s effective democratization of access to high-performance parallel computing capabilities, healthcare institutions may be more widely involved in the creation of individualized medicinal goods. This improves productivity and encourages teamwork in research and development projects. Geographically separated teams may collaborate on additive manufacturing projects without difficulty thanks to the democratization of such computer resources, which breaks down geographical constraints. This inclusive strategy fosters innovation and speeds up the development of specialized medical solutions by enabling a wide variety of healthcare institutions to utilize cutting-edge technologies (Butt, 2020). In the end, enhanced parallel computing’s accessibility dramatically aids in the democratization of healthcare innovation by promoting a collaborative environment with enormous potential for game-changing breakthroughs in the industry.
Edge Computing
As a decentralized computing model, edge computing makes processing data closer to the generation source feasible, attaining real-time decision-making and minimizing latency. It plays an invaluable role in healthcare additive manufacturing, especially in scenarios where immediate data analysis and local control are imperative (Nain et al., 2022). A perfect example is the need for real-time adjustments of patient-specific implants during surgery based on progress and emerging needs. By processing medical model data on-site and guaranteeing a smooth alignment with changing surgical needs, edge devices perform a praiseworthy role. The exceptional speed with which they can complete these jobs in real time dramatically increases the safety and efficiency of additive manufacturing in the healthcare industry. This is especially important in applications that require quick responses in order to provide the best possible care for patients. Healthcare workers may count on accurate and prompt medical device creation by utilizing edge computing, which will help them meet the changing needs of surgical operations. Thus, the incorporation of edge devices becomes a crucial development that holds promise for improved safety and effectiveness in the field of additive manufacturing in healthcare.
Reflection on the Architectural Design of Parallel Systems
However, a distributed memory architecture requires communication between processors since each unit has its local memory. In order to maximize performance, the study suggests looking at parallel solutions inside these frameworks. While distributed memory can improve scalability but necessitates effective communication protocols, shared memory permits data access without interruption but may encounter scaling issues. Through an examination of both approaches, this study seeks to find customized solutions depending on particular computational needs, offering a thorough method for handling a variety of parallel computing issues. As a result, it allows multiple processors to operate independently but share the same memory source. Changes in a memory source affected by one processor are visible to all other processors.
On the other hand, distributed memory always relies on a communication network to connect memory among processors. Each processor has its memory, and the memory address in one processor is not linked to another processor. Consequently, each processor operates independently. The programmer is intensely involved in defining how and when data is shared among processors (Hockney & Jesshope, 2019). The programmer is also involved in the synchronization between tasks—all these point out profound effectiveness and efficiency issues.
A programming paradigm that streamlines memory access through the utilization of a global address space, the shared memory architectural design has been subject to criticism due to concerns regarding scalability. Although there are scalability concerns, the efficient and rapid exchange of data between duties in a standardized fashion is enabled by the proximity of memory to the CPU in shared memory systems. Shared memory architectures are advantageous for applications that require rapid data exchange due to the efficient communication that results from this proximity. By facilitating a direct approach to memory management, the collaborative memory environment of programming significantly improves the efficiency of developers. Shared memory architectures are favored by certain applications, especially those that require real-time or low-latency communication, due to their efficient data-sharing capabilities and user-friendliness, notwithstanding their scalability limitations. Through these shared memories, architectural design always ensures timely and less resource-intensive operation. As a result, it is the preferred architectural design for the suggested parallel computing solutions in this study. This design would go a long way in ensuring data security, patient data privacy, scalability, and fault tolerance. There are two basic types of shared memory based on memory access time. These two include Uniform Memory Access (UMA) and Non-Uniform Memory Access (NUMA) (Khan et al., 2020). NUMA operates on a symmetric Multiprocessor (SMP) consisting of multiple identical processors with similar levels of access and access time to the shared memory.
General Discussion on Parallel Programming Using MPI
My thorough investigation of parallel programming utilizing the Message Passing Interface (MPI) has revealed its exceptional capability to improve communication and coordination between parallel processes within complex systems. In the domain of healthcare additive manufacturing, where MPI optimizes the performance and efficacy of parallelized operations, this disclosure is especially significant. The acquired knowledge highlights the critical significance of MPI in further developing the functionalities of intricate systems, thereby making a contribution to the progression of additive manufacturing methodologies and their implementations within the healthcare industry. With its exceptional capability to facilitate the smooth transfer of data between processors, MPI functions as a resilient framework. This characteristic has been demonstrated to be essential in enhancing the efficiency of complex additive manufacturing tasks, such as the precise segmentation of three-dimensional models, the continuous monitoring of manufacturing processes in real time, and the implementation of rigorous quality control protocols. MPI is a cornerstone in the field of high-performance computing for complex and resource-intensive applications, such as additive manufacturing processes in the healthcare industry, due to the parallelization it enables, which not only expedites computational tasks but also improves the scalability and efficiency of parallel applications.
MPI has in place three different communication methods that MPI processes can use to communicate with each other, including point-to-point communication, collective communication, and one-sided communication. MPI point-to-point forms one of the most commonly used, and it uses the same communicator in transferring messages from one process to the following process (Jani, Kumar & Nahata, 2022). This eliminates the possibility of errors occurring along the communication channel. At the same time, it offers a considerable level of flexibility that stakeholders can intensely exploit to achieve their intended goals and objectives. Suppose the objective is to ensure a relatively enhanced precision when there is much time. In that case, an MP blocking communication framework can be used in sending a message to another and waiting until the receiving completely and correctly processes the message before it executes the following function. On the other hand, when there are time constraints, MPI non-blocking communication allows for the execution of the following function even when the receiving processor has not completely and correctly received the message.
Besides, several MPI distribution implementations can be utilized to help in realizing efficiency. However, the most commonly used implementation is the message-passing interface chameleon (MPICH). MPICH supports high-performance, open-source, and portable implementation of MPI for parallel computation (Skjellum et al., 2020). MPI (Message et al.) has been an indispensable component of parallel computation across a variety of systems in my professional endeavors. Using my experience, I have optimized parallel systems by utilizing MPI (Message Passing Interface) capabilities. My background includes optimizing multi-core nodes and clusters for effective data sharing and smooth communication. With this expertise, I am now able to increase the efficiency of intricate, parallel computing systems. The effective incorporation of cutting-edge computers and fast networks is critical, particularly in fields where operational efficiency is of the utmost importance. The capabilities of MPI are particularly evident in tasks that require real-time monitoring, task parallel solution optimization, and 3D model segmentation. In addition to enabling concurrent processing, its robustness serves as the cornerstone for efficient, high-performance computing environments. Within an environment characterized by limited time and computational resources, MPI arises as a fundamental component that stimulates progress in parallel computing across a multitude of domains.
The notable portability and scalability of MPI (Message Passing Interface) are fundamental to its positive impact on efficiency and efficacy. From my personal experience, MPI’s most significant advantage is its flexibility in handling a wide variety of computing environments. Scalability is a critical determinant in guaranteeing optimal performance in shared memory architectures. This is especially beneficial as it circumvents constraints caused by the restricted quantity of processors or cores within a solitary computing node. By virtue of its capacity to scale effortlessly in response to rising computational demands, MPI is a resilient parallel processing solution that empowers applications to exploit the complete capabilities of distributed computing resources. The portability and scalability of MPI remain crucial in propelling progress in parallel computing paradigms as technology evolves. One notable aspect of this flexibility is that it does away with the requirement for a thorough understanding of each component in order to utilize it successfully. This flexibility gives consumers a more simplified experience and lets them take advantage of the system’s capabilities without having to worry about learning every little aspect. This solution’s improved characteristics make it adaptable to a variety of applications, which increases its efficacy. Concurrently, intuitive improvements optimize functionality and increase accessibility. These developments guarantee a more user-friendly and effective experience by increasing its power and lowering the learning curve.
Other Relevant Topics
A critical area of emphasis within the domain of healthcare additive manufacturing is the integration of blockchain technology. By employing this state-of-the-art fusion technique, critical health data is guaranteed to be accurate, thereby enhancing the traceability and security of individualized medical devices. By harnessing the decentralized and tamper-resistant characteristics of blockchain technology, healthcare providers can create an impregnable chain of possession for devices that are specific to each patient. This innovation enhances the security of confidential medical information and strengthens the dependability of additive manufacturing procedures, thereby ushering in a paradigm shift towards more remarkable accuracy and confidence in the healthcare sector. An additional critical aspect pertains to the examination of the influence that 3D printing materials have on the biocompatibility of healthcare applications. This pertains to comprehending the mechanisms by which materials interact within the human body, thereby impacting the effectiveness and safety of manufactured medical constructs.
Furthermore, it is critical to analyze the ethical implications and regulatory obstacles that emerge as a result of the integration of parallel computation in the manufacturing of healthcare products. This entails effectively managing concerns pertaining to confidentiality, data ownership, and compliance with rigorous healthcare regulations. By taking into consideration these supplementary aspects, the domain of healthcare additive manufacturing can progress not solely technologically but also in terms of ethics and security, thereby nurturing advancements that authentically prioritize the welfare of patients.
Currently, blockchain technology provides a framework for safeguarding copyright and industrial property, certification, and sustainable responsibility regarding the replication in AM. Besides, data on blockchain has been identified to be tamper-proof and traceable. Through this, the use of blockchain is recommended in addressing the challenge of mutual distrust among consumers in a multi-user environment. The blockchain ledger’s smooth and automatic updates, made possible by smart contracts, are essential to maintaining the accuracy of transaction records and device upkeep. Safeguarding the legitimacy of past activities requires real-time transparency to prohibit unlawful changes by other parties. The traceability capabilities of blockchain also play a critical role in promptly detecting and resolving unexpected device quality problems. Blockchain technology transforms several industries by increasing efficiency and building trust by creating a decentralized, tamper-resistant structure. Decentralization reduces the possibility of a single point of failure, while its secure design guarantees data integrity. A crucial element, intelligent contracts, automate and bring transparency to processes, simplifying them and creating the framework for quick actions. This helps with preventive troubleshooting as well as expediting operations, especially when it comes to device-related problems. The blockchain has a revolutionary effect that goes beyond security alone. It is causing a paradigm change in the direction of smooth, responsive, and accountable systems, which is changing how businesses function and develop in the digital age.
3D printing materials have unquestionably transformed the healthcare industry by increasing customization and precision. This technological advancement not only reduces wastage but also enhances efficiency, promoting rapid manufacturing. Ameta et al. (2022) assert that this technology is indispensable for the cost-effective development and manufacturing of biomedical devices, guaranteeing their precision and accessibility. As a result, the healthcare industry enters a period of profound change characterized by individualized solutions, effective utilization of resources, and expedited manufacturing procedures, all made possible by the revolutionary effects of 3D printing materials. The long-term results achieved from these include better patient outcomes, improvement in the quality of care, cost savings, and a more sustainable healthcare framework that meets the needs of both current and future generations. Additive Manufacturing (AM) has brought about unprecedented changes in design, manufacturing, and use, completely changing the field of medical device development. By integrating AM technology, medical equipment may be made to order and customized to meet specific needs. The core of real-time data collecting and transmission becomes crucial, supporting this operational progress above and beyond aesthetics. In addition to guaranteeing the smooth operation of these cutting-edge gadgets, real-time data opens the door for individualized treatment regimens. AM and data-driven insights work together dynamically to promote medical innovation and alter the paradigm toward more patient-centered, cost-effective healthcare solutions.
There are several potential ethical considerations and regulatory challenges that arise when implementing parallel computing in healthcare manufacturing. There are genuine concerns about AM becoming a new source of inaccuracy and data breaches. In the healthcare environment, mistakes can result in severe consequences to the patient as the primary victim, which prompts the need for a robust framework for protecting patients. When the framework for diagnosis and treatment reaches a high degree of accuracy, the paradox of precision in healthcare emerges. Paradoxically, this kind of achievement can also make healthcare professionals feel complacent, which can eventually cause their skill sets to deteriorate. A decrease in the amount of human interaction in patient care is an unexpected result. Although the use of parallel computing methods to artificial intelligence in medicine (AM) shows encouraging developments, it also brings up moral questions. Strict regulations are required because of ethical concerns, which range from algorithmic biases to privacy concerns. Thankfully, a lot of legal frameworks are proactively tackling these issues and maximizing parallel computing’s potential in the medical field while guaranteeing its ethical and responsible application. In the era of modern medical technology, maintaining the human-centric nature of healthcare requires striking a balance between ethical considerations and scientific progress.
The available regulatory framework for parallel computing is critical in its operability. As 3D printing with parallel computing enables the creation of highly customized medical products, there is a need for regulatory guidelines that specifically address customization. These guidelines can help manufacturers understand how to ensure compliance while tailoring products to individual patients’ needs. Besides, some regulatory bodies are establishing regulatory sandboxes, which are controlled environments for testing innovative technologies and products (Allen, 2019). In healthcare additive manufacturing, a regulatory sandbox can provide a space for manufacturers to test new 3D-printed medical devices under regulatory supervision. This fosters innovation while ensuring compliance.
Moreover, regulatory bodies may offer incentives for manufacturers developing innovative, compliant 3D-printed medical devices. These incentives can take the form of expedited approvals or reduced regulatory fees, encouraging manufacturers to invest in cutting-edge technologies, including parallel computing, while ensuring compliance. Ultimately, regulatory authorities can provide training and certification programs specifically tailored to the unique challenges and opportunities presented by additive manufacturing. This ensures that professionals involved in the industry are well-versed in the latest regulatory requirements.
Conclusion
Though additive manufacturing has been essential in addressing material use, cost-effectiveness, and timeliness issues in healthcare manufacturing compared to traditional techniques, there are still concerns regarding the level of accuracy achieved in prototyping, the amount of material used, and the time involved. The application of parallel computing solutions like parallel algorithms for slicing 3D medical models, real-time monitoring of 3D printing processes, and optimizing support structures for complex medical device designs are critical in realizing these objectives. The use of these solutions has been identified to help speed up the production of personalized medical devices while at the same time offering sensitive and real-time processing of patient-specific data, thereby enhancing customization. These applications also acknowledge the complexity involved in medical geometry. Medical devices may have complex geometries that are computationally intensive to produce. Parallel computing can handle these intricate designs more efficiently. Some of the emerging topics identified by this study include the need to investigate the integration of blockchain technology for enhanced traceability and security of patient-specific medical devices, exploring the impact of 3D printing materials and biocompatibility in healthcare applications, and studying the potential ethical considerations and regulatory challenges that arise when implementing parallel computing in healthcare manufacturing.
References
Allen, H. J. (2019). Regulatory sandboxes. Geo. Wash. L. Rev., 87, 579.
AlZu’bi, S., Shehab, M., Al-Ayyoub, M., Jararweh, Y., & Gupta, B. (2020). Parallel implementation for 3D medical volume fuzzy segmentation. Pattern Recognition Letters, 130, 312-318.
Ameta, K. L., Solanki, V. S., Singh, V., Devi, A. P., Chundawat, R. S., & Haque, S. (2022). Critical appraisal and systematic review of 3D & 4D printing in sustainable and environment-friendly smart manufacturing technologies. Sustainable Materials and Technologies, p. 34, e00481.
Bajaj, P., Hariharan, A., Kini, A., Kürnsteiner, P., Raabe, D., & Jägle, E. A. (2020). Steels in additive manufacturing: A review of their microstructure and properties. Materials Science and Engineering: A, 772, 138633.
Butt, J. (2020). Exploring the interrelationship between additive manufacturing and Industry 4.0. Designs, 4(2), 13. https://doi.org/10.3390/designs4020013
Duraes, T. D. J. O., de Souza, P. S. L., Martins, G., Conte, D. J., Bachiega, N. G., & Bruschi, S. M. (2020, October). Research on Parallel Computing Teaching: state of the art and future directions. In 2020 IEEE Frontiers in Education Conference (FIE) (pp. 1-9). IEEE.
Fang, Y., Chen, Q., & Xiong, N. (2019). A multi-factor monitoring fault tolerance model based on a GPU cluster for big data processing. Information Sciences, 496, 300-316. https://doi.org/10.1016/j.ins.2018.04.053
Hockney, R. W., & Jesshope, C. R. (2019). Parallel Computers 2: architecture, programming and algorithms. CRC Press.
Jani, K., Kumar, A., & Nahata, R. (2022). Hpcfolder: a simple tool used to parallelize algorithms using the message passing interface (MPI). The Journal of Supercomputing, 78(1), 258-278.
Jihong, Z. H. U., Han, Z. H. O. U., Chuang, W. A. N. G., Lu, Z. H. O. U., Shangqin, Y. U. A. N., & Zhang, W. (2021). A review of topology optimization for additive manufacturing: Status and challenges. Chinese Journal of Aeronautics, 34(1), 91-110.
Jihong, Z. H. U., Han, Z. H. O. U., Chuang, W. A. N. G., Lu, Z. H. O. U., Shangqin, Y. U. A. N., & Zhang, W. (2021). A review of topology optimization for additive manufacturing: Status and challenges. Chinese Journal of Aeronautics, 34(1), 91-110.
Kamarianakis, M., & Papagiannakis, G. (2021). An all-in-one geometric algorithm for cutting, tearing, and drilling deformable models. Advances in Applied Clifford Algebras, 31(3), 58.
Khan, L. U., Yaqoob, I., Imran, M., Han, Z., & Hong, C. S. (2020). 6G wireless systems: A vision, architectural elements, and future directions. IEEE access, 8, 147029-147044.
Khan, M. F., Alam, A., Siddiqui, M. A., Alam, M. S., Rafat, Y., Salik, N., & Al-Saidan, I. (2021). Real-time defect detection in 3D printing using machine learning. Materials Today: Proceedings, 42, 521-528.
Lai, X., & Wei, Z. (2021). Slicing algorithm and partition scanning strategy for 3d printing based on gpu parallel computing. Materials, 14(15), 4297.
Maksum, Y., Amirli, A., Amangeldi, A., Inkarbekov, M., Ding, Y., Romagnoli, A., … & Akhmetov, B. (2022). Computational acceleration of topology optimization using parallel computing and machine learning methods–analysis of research trends. Journal of Industrial Information Integration, 28, 100352.
Memeti, S., Pllana, S., Binotto, A., Kołodziej, J., & Brandic, I. (2019). Using meta-heuristics and machine learning for software optimization of parallel computing systems: a systematic literature review. Computing, 101, 893-936.
Nain, G., Pattanaik, K. K., & Sharma, G. K. (2022). Towards edge computing in intelligent manufacturing: Past, present and future. Journal of Manufacturing Systems, 62, 588-611. https://doi.org/10.1016/j.jmsy.2022.01.010
Nickolls, J., & Dally, W. J. (2020). The GPU computing era. IEEE Micro, 30(2), 56-69. DOI: 10.1109/MM.2010.41
Oyama, Y., Maruyama, N., Dryden, N., McCarthy, E., Harrington, P., Balewski, J., … & Van Essen, B. (2020). The case for strong scaling in deep learning: Training large 3d cnns with hybrid parallelism. IEEE Transactions on Parallel and Distributed Systems, 32(7), 1641-1652.
Paraskevoudis, K., Karayannis, P., & Koumoulos, E. P. (2020). Real-time 3D printing remote defect detection (stringing) with computer vision and artificial intelligence. Processes, 8(11), 1464.
Permann, C. J., Gaston, D. R., Andrš, D., Carlsen, R. W., Kong, F., Lindsay, A. D., … & Martineau, R. C. (2020). MOOSE: Enabling massively parallel multiphysics simulation. SoftwareX, 11, 100430.
Ramola, M., Yadav, V., & Jain, R. (2019). On the adoption of additive manufacturing in healthcare: a literature review. Journal of Manufacturing Technology Management, 30(1), 48-69.
Rauch, E., Unterhofer, M., & Dallasega, P. (2018). Industry sector analysis for the application of additive manufacturing in smart and distributed manufacturing systems. Manufacturing Letters, 15, 126-131. https://doi.org/10.1016/j.mfglet.2017.12.011
Sbrugnera Sotomayor, N. A., Caiazzo, F., & Alfieri, V. (2021). Enhancing Design for Additive Manufacturing Workflow: Optimization, Design and Simulation Tools. Applied Sciences, 11(14), 6628. MDPI AG. Retrieved from http://dx.doi.org/10.3390/app11146628
Shukur, H., Zeebaree, S., Zebari, R., Ahmed, O., Haji, L., & Abdulqader, D. (2020). Cache coherence protocols in distributed systems. Journal of Applied Science and Technology Trends, 1(3), 92-97. https://doi.org/10.38094/jastt1329
Skjellum, A., Rüfenacht, M., Sultana, N., Schafer, D., Laguna, I., & Mohror, K. (2020). Example: A modern design and implementation to accelerate message-passing interface innovation. In High Performance Computing: 6th Latin American Conference, CARLA 2019, Turrialba, Costa Rica, September 25–27, 2019, Revised Selected Papers 6 (pp. 153-169). Springer International Publishing.
Tartici, I., Kilic, Z. M., & Bartolo, P. (2023). A systematic literature review: Industry 4.0 based monitoring and control systems in additive manufacturing. Machines, 11(7), 712. https://doi.org/10.3390/machines11070712
Umang, N., Bollapragada, S., & Hertel, K. (2022). Practice summary: Multilaser load balancing in additive manufacturing. INFORMS Journal on Applied Analytics, 52(4), 390-393. https://doi.org/10.1287/inte.2022.1118
Vaneker, T., Bernard, A., Moroni, G., Gibson, I., & Zhang, Y. (2020). Design for additive manufacturing: Framework and methodology. CIRP Annals, 69(2), 578-599.
Walker, J. M., Prokop, A., Lynagh, C., Vuksanovich, B., Conner, B., Rogers, K., … & MacDonald, E. (2019). Real-time process monitoring of core shifts during metal casting with wireless sensing and 3D sand printing. Additive Manufacturing, 27, 54-60.
Wei, H. L., Mukherjee, T., Zhang, W., Zuback, J. S., Knapp, G. L., De, A., & DebRoy, T. (2021). Mechanistic models for additive manufacturing of metallic components. Progress in Materials Science, 116, 100703.
Xia, T., Dong, Y., Xiao, L., Du, S., Pan, E., & Xi, L. (2018). Recent advances in prognostics and health management for advanced manufacturing paradigms. Reliability Engineering & System Safety, 178, 255-268. https://doi.org/10.1016/j.ress.2018.06.021
Zhou, X., Zhao, J., Wang, J., Huang, X., Li, X., Guo, M., & Xie, P. (2019). Parallel computing-based online geometry triangulation for building information modeling utilizing big data. Automation in Construction, 107, 102942.
write