Building Smarter: A Scalable Multi-Project Intelligent Construction Ecosystem for Safety, Quality, and Efficiency Enhancement

The Architecture, Engineering, and Construction (AEC) industry faces mounting challenges in managing large-scale, multi-project projects due to fragmented data systems, isolated AI applications, and inefficient cross-project coordination. This paper proposes Building Smarter, a Multi-Project Intelligent Construction Ecosystem that integrates BIM, 5G, IoT, and multimodal AI into a unified, scalable framework. This system adopts a hierarchical four-layer architecture consists of edge, intelligence, governance, and learning layers, to connect geographically dispersed projects through federated learning and policy enforcement mechanisms. Industrial-grade infrastructure supports real-time data capture and processing, while modular microservices ensure rapid deployment. Core functional innovations include AI-driven on-site management, multimodal LLMs for safety compliance, automated progress tracking, and closed-loop workflows that convert problem reports into systemic improvements. A shared knowledge repository aggregates cross-project multimodal data into reusable insights through domain-specific MLLMs, while scalability mechanisms reduce marginal costs by 40–60% for new projects. Deployed across 273 construction sites in China, the system achieved 78% coverage, reducing safety violations by 30%, accelerating quality control by 50%, and saving 400 person-days annually through remote monitoring. The framework demonstrates that interconnected ecosystems are critical for addressing the complexity of modern mega-projects, particularly for large general contractor companies.