The overarching objective of the IAI Lab is to pioneer AI-driven advancements by leveraging the strength of engineering background. By integrating the knowledge of diverse engineering disciplines ranging from mechanical engineering, physics, and computer science, we harness the full potential of artificial intelligence (AI) in redefining engineering processes and outcomes. We are dedicated to developing advanced AI methods based on physics-guided and data-driven insights, enabling us to model the complexities of various engineering systems well. Our strategies extend beyond conventional and existing approaches by empowering AI-aided engineering, thereby facilitating unprecedented perception, decision-making, optimization, and so on. Through this synergy of engineering and AI, the IAI Lab strives to pioneer a new wave of AI innovation, elevating the engineering landscape and further making the future industry smarter and more efficient.
Prof. Sooyoung Lee gives an invited talk (신진연구자 초청발표) at the Korean Society of Mechanical Engineers (KSME), CAE and Applied Mechanics Division.
Title: "Engineering Meets AI: Physics-guided AI for Smart Design and Manufacturing"
Abstract: This seminar delves into the advancement of physics-guided artificial intelligence (AI), which integrates engineering knowledge with cutting-edge AI methodologies to enhance the predictive performance, fidelity, and interpretability of neural computational methods. The presentation will primarily focus on the development of novel deep neural networks that are inspired by specific engineering mechanisms, significantly improving our understanding of physical phenomena. Additionally, it will cover recent breakthroughs in revolutionizing digital twin applications, offering real-time, accurate, interactive, and customizable solutions across various scientific and engineering fields. Our approach also facilitates the creation of diverse design recommendations through engineering-tailored generative AI to meet specific engineering requirements and address aspects of design such as generalizability and manufacturability. Our studies potentially suggest hybridizing AI methods and domain knowledge to broaden the scope for exploring innovative and efficient solutions in engineering challenges.