Background Conventional engineering workflows rely on rule-based systems, manual analysis, and offline simulations that struggle with high-dimensional data, real-time variability, and complex multi-physics interactions. The proliferation of IoT sensors, high-fidelity simulations, cloud platforms, and cyber-physical systems has generated unprecedented volumes of engineering data, creating both opportunities and challenges for traditional methods.
Objective This article examines how artificial intelligence (AI)—encompassing machine learning (ML), deep learning (DL), generative AI, computer vision, reinforcement learning, and digital twins—drives digital transformation across engineering disciplines. It develops a conceptual multi-layer framework, reviews emerging technologies and domain applications, identifies limitations, and outlines future research directions toward intelligent, autonomous engineering systems.
Methods A systematic analytical and conceptual review framework was employed. Peer-reviewed literature on AI in engineering, Industry 4.0/5.0, digital twins, and related technologies was critically synthesized. No new experimental datasets were generated; findings are literature-derived and theoretical. Evaluation draws on established performance dimensions (prediction accuracy, efficiency, cost, scalability, latency) reported across studies.
Results AI enables superior prediction, optimization, monitoring, and decision-making compared with conventional approaches. Digital twins integrated with AI support real-time synchronization between physical assets and virtual models. Applications span structural health monitoring, generative design, predictive maintenance, smart grids, and sustainable energy systems. Comparative analysis shows clear advantages in automation level, real-time capability, and scalability, tempered by challenges in data quality, explainability, cybersecurity, computational cost, and legacy-system integration.
Conclusion AI-driven digital transformation constitutes a fundamental shift from reactive, manually intensive engineering toward proactive, data-centric, and increasingly autonomous systems. Realizing its full potential requires advances in physics-informed and trustworthy AI, multimodal foundation models, edge intelligence, and human–AI collaboration within Industry 5.0 paradigms.