MACHİNE DESİGN İN THE DİGİTAL ERA: NEXT-GENERATİON APPROACHES AND TRENDS
3rd INTERNATIONAL CONGRESS ON NATURE, ENGINEERING AND SCIENCE, İstanbul, Türkiye, 31 Ekim 2026, ss.1-9, (Tam Metin Bildiri)
- Yayın Türü: Bildiri / Tam Metin Bildiri
- Basıldığı Şehir: İstanbul
- Basıldığı Ülke: Türkiye
- Sayfa Sayıları: ss.1-9
- Dokuz Eylül Üniversitesi Adresli: Evet
Özet
In this study, the fundamental transformation
of engineering design from traditional, linear, and human-experience-dependent
methodologies toward intelligent, data-driven, and autonomous paradigms is
explored. Driven by advancements in computational power and deep learning,
Generative Artificial Intelligence in Design (GAIDe) enables the rapid
exploration of high-dimensional design spaces while optimizing conflicting
objectives such as mass reduction, structural integrity, manufacturability, and
environmental sustainability. By evaluating recent literature, key technical
approaches including surrogate modeling (MLP, CNN) and direct design generation
(GAN, diffusion models) are analyzed. Furthermore, closed-loop framework
architectures incorporating physics-based simulations (FEA, CFD) and
reinforcement learning (RL) are demonstrated to significantly outperform
conventional evolutionary algorithms across metrics such as design accuracy
(97%), material utilization efficiency (92%), and computational speed
improvement (93%). Real-world case studies in aerospace, automotive, and soft
robotics illustrate how topology optimization combined with additive
manufacturing yields over 40% mass savings and enhances mechanical performance.
Additionally, the shift toward "GAIDe for X" integrates downstream
constraints such as manufacturing rules, structural reliability, and lifecycle
considerations directly into early conceptual ideation. Despite current
challenges regarding data generation costs, model interpretability, and
physical accuracy, the fusion of physical laws, domain-specific constraints,
and agentic LLM workflows forms the foundation of next-generation engineering
systems. Ultimately, this paper demonstrates that AI-driven generative design
establishes a collaborative human-AI ecosystem that accelerates innovation,
reduces development cycles, and supports sustainable industrial manufacturing.