Internet of Thing-based ambience automation for optimizing patient's comfort

There have been efforts to automate internal temperature inside hospital premises. However, none took a patient-centric approach. Therefore, there has been a significant research gap in automating the ambience within hospital buildings, keeping the patient's comfort level at a focal point. The objective of this proposed system is to utilize a reinforcement learning (R L) approach to maximize a patient's comfort level by regulating the thermal parameters of his/her ambience. A machine learning model is first trained with the statistical data for several unique health parameters and their corresponding thermal comfort points for a wide range of individuals. When it encounters a patient, the system first reads the medical condition and then utilizes its Internet of Things (IoT) devices to control the smart thermostat to start regulating the environmental parameters of the patient's room. After going through frequent attempts to optimize the ambience for a particular patient, the system starts to learn from its error. It capitalizes on the previous decision it made each time. Thus, the system can identify the perfect environment for each patient. As a result, the proposed system will benefit the intended patients and lessen the efforts that the on-duty doctors and nurses should put in. The proposed system is equally effective in a normal room with thermostat facilities, as it is primarily trained with the data from the 'ASHRAE Global Thermal Comfort Database II'.