Applications of multimodal large language models in construction industry

Document Type

Article

Publication Date

1-2026

Department

Department of Civil, Environmental, and Geospatial Engineering

Abstract

The advancement of transformer-based models, including multimodal large language models (MLLMs), has led to growing interest in their application across diverse industries, including construction. While a few earlier reviews have explored generative artificial intelligence in the construction sector, they are limited in scope— limited coverage of multimodal models, covering a shorter timeline prior to the expansion of MLLMs research, and offering limited emphasis on practical use cases, adaptation strategies, and integration into construction workflows. This study addresses that gap by reviewing 83 peer-reviewed studies published between January 2020 and February 2025, identified using a structured search process guided by the PRISMA framework and focused on academic literature. By synthesizing these studies, this review highlights trends across application domains, model types, adaptation strategies, technical limitations, and performance evaluation practices—offering a comparative analysis across use cases. It concludes with recommendations for future research, underscoring the need for standardized evaluation frameworks, critical limitations related to technical aspects, ethical risks, and regulatory uncertainty, underscoring the need for responsible development and deployment of MLLMs in construction settings.

Publication Title

Advanced Engineering Informatics

Link to Full Text

Share

COinS