Strategic Energy Forecasting and Policy Design for Carbon-Neutral University Campuses in KMUTT

Higher education institutions in developing economies now function as key drivers to help their countries achieve carbon neutrality goals. These institutions as major energy users and innovation centers enable the combination of data forecasting with sustainable policy development for the long-term. This research introduces a strategic energy planning system which merges AI-based forecasting with policy evaluation scenarios to direct university campus decarbonization efforts. The research at King Mongkut’s University of Technology Thonburi (KMUTT) in Thailand implements a deep learning system which combines Convolutional Neural Networks (CNN) with Bidirectional Long Short-Term Memory (BiLSTM) and Transformer-based attention mechanisms. The model achieved R2 = 0.96 predictive accuracy after being trained with high-resolution energy and meteorological data from 2020 to 2024. The model allows to predict three policy-oriented scenarios, including Energy Efficiency (EE), Renewable Integration (RE), and Behavioral Change to determine their extended effects on campus energy usage. This research shows that EE measures will decrease the total energy consumption by 20% until 2040, while RE and behavioral strategies will decrease the energy usage by 15–18% and 8–10%, respectively. The research combines these findings into a strategic roadmap that follows the National Energy Plan (NEP), the Power Development Plan (PDP), and the Energy Efficiency Plan (EEP) of Thailand. The research establishes a transferable method which enables universities across ASEAN to use high-performance AI forecasting for institutional policy design to guide stakeholders in their higher education carbon neutrality transition.