Smart Campus Ecosystems: Designing Digital Twins for Educational Infrastructure and Strategic Management

Abstract

The explosion of smart technologies has been forcing educational institutions to utilize same or similar intelligent infrastructure for optimal performance and data-driven decisions. This paper describes the design and creation of an intelligent campus ecosystem based on Digital Twin (DT) technology, and the use of machine learning algorithms to model the campus to facilitate simulation, monitoring and optimization. Four algorithms were used in the project - K-Means clustering for occupancy pattern detection; Long Short-Term Memory (LSTM) networks for energy prediction; A* algorithm for real time navigation on campus; Random Forest for Security risk classification. A simulated set of historical data of a size similar to a mid-sized university for a 30-day period was developed to run the project. The results showed significant benefits of a smart campus, such as a decreased energy usage of 9.4% per day; decreased HVAC runtime of 20.9%; response to security alerts improved by 48.9%; improved average navigation efficiency by 22.6%. This study presents, relative to the present models, a more comprehensive and adaptive digital twin model specifically designed for educational settings. This system supports real-time decision making and sustainability outcomes, providing a framework that can be scaled for digital transformation in higher education. This research adds to the existing pool of work emphasizing smart, intelligent and data-informed educational uses, connecting emerging technologies with university strategic planning.