The aerospace industry is on a constant quest to adopt new technologies that offer competitive advantage and step-change capability. Artificial intelligence (AI) and machine learning (ML) approaches in additive manufacturing (AM) offer substantial merit in meeting the industry’s needs. Research insights suggest that combining these advanced technologies could streamline existing research and development (R&D) efforts, subsequently improving part quality, reducing costs, and enhancing overall system-level efficiency. Despite these benefits, this integration presents unique challenges, such as the need for robust data management systems and the development of reliable and accurate AI algorithms. Additionally, validating the repeatability and reliability of AI-powered systems in complex AM aerospace applications raises more inquiries than concrete answers, dictating the need for further exploration for widespread adoption.

America Makes, the national additive manufacturing innovation institute, has captured the market’s needs, identifying a brewing interest across the sector to further explore the opportunities and challenges of integrating AI and ML approaches in AM for aerospace applications.

Integration across materials and process qualifications

Predicting outcomes in AM can become obscured due to the variation of machines, materials, and printing parameters. Developing AI and ML models that can be applied across different processes offers substantial benefits for producing high-quality parts. However, the complexity and variability of AM processes make it difficult to predict outcomes accurately, especially for high-criticality applications. To ensure platform and user safety, rigorous testing, validation, and necessary equipment, methods, personnel qualification, and certification are required. Despite the significant cost and complexity involved, developing robust and accurate AI and ML models for AM has demonstrated the potential to improve economic and production efficiency for high-quality AM applications.

As a result, ML is a major opportunity for the AM industry to correlate materials with specific parameter sets, achieving a consistent and reliable material output, and validating a robust process. Manually analyzing real-time data for large data sets is a time-consuming, complex, and potentially confounding process. However, delegating data analysis to computers for concurrent processing can lead to accelerated industry progress and statistically validated results.

In recognition of this issue, directed by the Department of Defense (DoD), the Institute and the National Center for Defense Manufacturing and Machining (NCDMM) launched a $3.2M project titled Demonstration of Novel Methods for Effective AM Process Qualification/Re-Qualification – Delta Qualification. The project’s overall goal was to demonstrate an AM process that could provide an efficient and cost-effective means to incorporate changes in essential processes, post-processing techniques, and material feedstock variables while ensuring qualified AM material validation through statistical analysis.

America Makes member Senvol was awarded a topic within the project focusing on leveraging ML to accelerate the delta qualification process. Senvol, based in New York, devised an innovative approach, using ML algorithms, to calculate statistically-based material property predictions analogous to material allowables.

Within the topic area’s scope, the Senvol team is analyzing numerous changing parameters simultaneously, noting the marginal contribution of each. The research indicates a viable alternative to conventional “point solution” methods, opening the door for a more economically feasible and flexible approach, substantiating that data-driven ML algorithms could significantly reduce the cost of material allowables development. This approach leads to a greater understanding of optimal solutions to overcome the qualification and re-qualification challenges that slow the expansion of this innovative industry.

Quality control and assurance

AI and ML algorithms rely heavily on data, and the quality of such data directly influences the accuracy of the algorithms’ predictions and decisions. AI-powered systems are increasingly being explored to automate various aspects of AM, making it crucial to ensure that the end products meet the required quality standards.

Historically, manufacturing quality control and assurance practices heavily relied upon the ability of human inspectors to scrutinize products for defects and specification aberrations—a tedious, expensive process often prone to errors. Leveraging knowledge gained from increased R&D efforts, industries have better established a new understanding of AI and ML capabilities when paired with AM to address these technological gaps. This has led to significant improvements in AI automation being incorporated at various production stages. Today, AI algorithms can help detect potential failures in aircraft components in advance by analyzing data collected from sensors. These algorithms can adjust in real time, enabling proactive maintenance and preventing costly downtimes. This approach also reduces the chances of human error, enhancing printed part accuracy.

However, adopting AI can also pose new challenges to quality assurance and control, such as the need for specialized knowledge and skills to operate and maintain AI-powered machines. Furthermore, the complexity of aerospace parts demands high accuracy and precision, which can be challenging to achieve within AM processes.

Despite these challenges, the integration of AI holds great potential for improving the quality and efficiency of aerospace part production, provided that appropriate quality assurance and control measures are implemented. As such, there is a need to establish robust quality assurance and control frameworks that can adequately address the challenges that arise. These frameworks should incorporate rigorous testing and verification procedures that ensure the accuracy and reliability of the final products.

Additionally, human involvement cannot be understated as an essential element in

navigating the complex interplay between AI, ML, and AM. Effectively integrating these technologies requires highly skilled professionals with expertise in multiple disciplines, including computer science, material science, physics, and aerospace engineering. The challenge can be met by investing in training and education programs focused on developing interdisciplinary skill sets.

Exploring the possibilities of advanced technology 

From a broader perspective, AM presents an advantage over conventional manufacturing due to its digital nature, enabling an optimized output. Its ability to reduce assembly and labor by allowing the production of complex components in one piece leads to cost savings. With the help of AI and ML, AM data can be leveraged to potentially increase productivity, yield, and quality control, reducing the need for costly and prolonged inspections and post-processes.

Regarding the successful implementation of AI, ML, and AM in aerospace applications, the existing reality is fraught with complexity and obstacles that require careful consideration. Overcoming these challenges requires the deployment of ongoing efforts dedicated to developing robust and scalable models that can be applied across diverse AM processes to realize the development of a strong, efficient, cost-effective manufacturing process. Regardless of the technological hurdles, the possibilities afforded by such advanced technologies warrant interdisciplinary collaboration to foster innovative solutions to propel U.S. manufacturing forward.