Digital Transformation in Science and Engineering: Trends, Challenges, and Future Directions

Abstract

Digital transformation—the integration of digital technologies, data, and intelligent computation into the core of how work is conceived and conducted—is reshaping both the practice of science and the practice of engineering, and increasingly dissolving the boundary between them. This article provides a comprehensive review of digital transformation across science and engineering, charting its principal trends, the challenges that impede it, and the directions in which it is heading. We first distinguish digital transformation from the narrower notions of digitization and digitalization, then organise the field around four interlocking pillars: a data foundation built on big data, the data-intensive “fourth paradigm,” and the FAIR and open-science movements; a layer of computational and intelligent methods spanning high-performance and cloud computing and the rapid diffusion of artificial intelligence and machine learning; the convergence of the physical and the digital through digital twins, cyber-physical systems, the Internet of Things, and the Industry 4.0 paradigm; and the automation of discovery and production through self-driving laboratories, robotics, and additive manufacturing. Across these pillars we identify a common set of challenges that determine whether transformation succeeds in practice: data quality and interoperability, cybersecurity, the absence of standardization, the reproducibility of computational results, shortages of interdisciplinary skills, and pronounced inequities in access to data and compute. We then survey emerging future directions, including the human-centric and sustainability-oriented vision of Industry 5.0, cognitive digital twins, agentic artificial intelligence, edge intelligence, and the industrial metaverse. We conclude that digital transformation is best understood not as the adoption of any single technology but as a convergent, socio-technical reconfiguration of research and engineering practice, whose ultimate success depends as much on data governance, standards, security, and human capacity as on the underlying technical advances.