Artificial Intelligence in Civil Engineering: A Systematic Review of Transformative Applications, Gaps, and the Imperative for Hybrid Frameworks

https://doi.org/10.46610/JOCBME.2025.v011i03.005

Abstract

The fusion of artificial intelligence (AI) and machine learning (ML) is the most important and revolutionary element in the field of civil engineering. It provides the discipline with innovative, data-driven solutions that are a hundred times more efficient, accurate, and environmentally friendly than traditional ways. Key AI methods such as neural networks, deep learning, and generative AI are being used in a wide range of applications, like SHM and predictive maintenance, which help in early damage detection and the prolongation of the infrastructure lifespan. Besides, AI intervention has been made throughout the construction lifecycle by ameliorating construction management, resource planning, and risk forecasting to deal with the traditional industry problems like cost and time overruns. During the design phase, AI has made it possible for the intelligent structural design pipeline integrated with BIM to automatically produce and optimise structural drawings while at the same time helping the environment by carrying out energy efficiency analysis and impact assessment. Although AI in civil engineering has all the potential to be a game changer, its full-scale use still encounters serious challenges, mainly of a technical and ethical nature. Consistently, the biggest technical hurdle is the need to maintain data quality and availability, since the AI models are dependent on large quantities of uniform, high-quality data, which is very often incomplete or hard to standardise in civil engineering projects. Other factors that are contributing to these difficulties include model interpretability and transparency issues, as well as the requirement for further system integration and the establishment of standard performance benchmarks. A major portion of the future research is directed towards working on the hybrid models that merge AI with classical engineering practices to produce the best outcomes.