Predictive Maintenance of Thermal-Energy-Storage Air-Conditioning with Deep Learning : Smart Predictive Facility Management and Maintenance for Charging Load Prediction of Thermal-Energy-Storage Air-Conditioning System with Deep Learning Techniques deployed in a User-Friendly Application.

Climate change is evident throughout the world as heat waves and floods are making the headlines. One of the daily amenities that are used by people as they feel the heat and humidity for a bearable atmosphere is an air-conditioner (AC). However, conventional AC emits greenhouse gases and adds to the heat and humidity in the atmosphere contributing more to global warming. It is time to focus on a greener alternative and yet sustainable form of AC which is a Thermal-Energy-Storage Air-Conditioner (TES-AC). This green alternative stores chilled water in the form of thermal energy at night when energy demand is low and then uses the stored thermal energy to cool the building’s air the next day. It does not just reduce building energy consumption but also reduces building management costs. Nevertheless, there is a reluctance to this shift at a large scale as management and maintenance of TES-AC is complicated due to inappropriate charging load doubling energy consumption. Hence, this research focuses on utilizing deep learning techniques to predict the charging load required in advance, so the facility managers prepare at night and minimize disruption in daily operations. The purpose of this research is to make facility managers more inclined to use the greener and more sustainable TES-AC to reduce energy consumption and management costs while contributing positively to the environment. Furthermore, this research demonstrates the possibility of reducing energy consumption even more and improving efficiency with deep learning and discusses deploying it in a user-friendly application.