Lab Description
This module presents an experimental digital twin of a hybrid-ventilated office environment, designed to optimize both occupant productivity and energy efficiency. The central objective of the experiment is to develop a high-level control strategy that dynamically switches between natural ventilation (e.g., automated window operation) and mechanical air conditioning, while maintaining temperature setpoints in alignment with ASHRAE standards.
The digital twin integrates real-time sensor networks, machine learning-based forecasting, and predictive control algorithms to establish a continuously updated virtual representation of the workspace. This virtual replica enables the prediction of indoor thermal conditions, energy demand, and potential productivity impacts across different ventilation modes. Based on these forecasts, the system selects and implements the optimal ventilation strategy for the upcoming 20-minute interval, minimizing productivity losses, energy consumption, and long-term operational costs.
The module further provides learners with a structured pathway for developing such systems, including the design and deployment of sensing–actuation networks, the construction of robust data pipelines, the training and integration of machine learning models, and the development of a custom control framework, culminating in an on-site implementation.
