Page 33 - IJAMD-1-2

International Journal of AI for
Materials and Design
AI-driven quality assurance in AM

• Benefit: This enhances traceability and makes it variability in AM processes, leading to more uniform
easier to demonstrate compliance with regulatory and reliable parts.
standards, as well as to conduct audits and • Faster certification: Automated documentation
inspections. and predictive compliance analytics streamline the
certification process, enabling faster approval of AM
(ii) Predictive analytics for compliance parts for use in safety-critical applications.
• Technique: Predictive analytics models • Increased trust and adoption: Demonstrating reliable
can forecast potential compliance issues compliance with regulatory standards through
based on historical data and current process AI-driven systems builds trust among stakeholders
conditions. 17,18,23 By identifying trends and and accelerates the adoption of AM technologies
deviations early, these models can alert operators across industries.
to potential regulatory non-conformities.
• Example: In aerospace manufacturing, predictive Quality assurance in AM processes faces a myriad of
models can ensure that parts meet stringent challenges arising from geometric complexity, material
safety and performance criteria by continuously variability, build orientation effects, process-induced
monitoring and adjusting process parameters to defects, and regulatory considerations. 27,28 Addressing
stay within regulatory limits. these challenges requires a multidisciplinary approach
integrating advanced materials science, process
(iii) Standardization and validation protocols engineering, computational modeling, metrology, and
• Collaboration: AI can assist in the development regulatory expertise. AI technologies, particularly ML,
of standardized validation protocols and offer powerful tools to enhance compliance, traceability,
qualification procedures. 13,14,19 ML models can and overall quality in AM, thereby bridging the gap
be trained on industry-wide data to identify best between emerging AM technologies and regulatory
practices and establish benchmarks for quality requirements. 22,23 By overcoming these challenges through
and compliance. strategic AI integration, the AM industry can unlock
• Impact: This collaboration between AI the full potential of AM technologies and accelerate the
technologies and regulatory bodies can lead to the adoption of innovative, high-performance AM products
creation of clearer guidelines and more effective across diverse application domains.
validation protocols tailored to the unique aspects
of AM. Each AI-driven approach for quality assurance in AM
has its unique strengths and limitations. CNNs excel in
(iv) Real-time quality control defect detection from image data, offering high accuracy
• Application: AI-driven real-time quality control but requiring large datasets. SVMs provide clear decision
systems can monitor and analyze the AM process boundaries and perform well with smaller datasets but may
to ensure that parts are being produced according require extensive feature engineering. RL offers dynamic
to specified standards. 25,26 Techniques such as process optimization capabilities, adapting in real-time but
computer vision and neural networks can detect often requiring significant computational resources and
defects and deviations from quality standards as careful implementation.
they occur.
• Advantage: Immediate detection and correction of By understanding the comparative advantages
defects help maintain consistent quality, reducing and challenges of these techniques, researchers and
the risk of non-compliance and enhancing the practitioners can make informed decisions on the most
reliability of AM parts. suitable AI-driven approaches for their specific AM
quality assurance needs. Future research should focus
The integration of AI into regulatory and certification on hybrid models that leverage the strengths of multiple
processes directly impacts the overall quality of AM parts techniques and on improving data collection and model
in several ways: interpretability to further enhance the effectiveness of AI
• Enhanced accuracy and precision: AI models improve in AM.
the accuracy and precision of quality assurance
processes by continuously learning from data and 3. Role of AI in quality assurance
refining their predictions. This results in higher- AI has emerged as a transformative tool for enhancing
27
quality parts that meet stringent regulatory standards. quality assurance in AM processes. By leveraging ML,
• Reduced variability: By optimizing process parameters deep learning, and computer vision techniques, AI
and ensuring consistent monitoring, AI reduces enables real-time monitoring, defect detection, process

Volume 1 Issue 2 (2024) 27 doi: 10.36922/ijamd.3455