Industrial product design for end user with digital image processing and artificial intelligence method

Air conditioner is a widely used system for the purpose of climate control. It is designed to improve indoor air quality, provide comfort and create an environment compatible with weather conditions. Artificial intelligence is a technology that enables computer systems to acquire human-like intelligence and learning capabilities to analyse complex data, learn and perform specific tasks. This technology is generally used in industrial areas to improve production processes. In recent years, it has been widely used in many areas of daily life. The developed artificial intelligence-supported air conditioning system aims to optimise the user experience by combining these two important concepts. This system is developed with common and low-cost components such as Raspberry Pi and Pi camera. It works using basic artificial intelligence functions such as face recognition and human detection. The face recognition function is performed by the Haar Cascade algorithm and the air conditioning settings are automatically adjusted according to the preferences of the users who are recognised by the system. The human detection system can detect the number of people in the room and their locations by using the TensorFlow machine learning library and YOLO artificial intelligence algorithm. In this way, air flow is optimised and air flow is made more comfortable without the air blowing from the air conditioner directly on people. In addition, the system automatically sets the air conditioner to energy saving mode when no people are detected in the room for a certain period of time. Thus, it contributes to sustainability by providing energy efficiency. The developed system is an important advance in not only the area of air-conditioning, at the same time saving energy and increasing user comfort. This study highlights that artificial intelligence technology has the potential for practical applications that can be widely used in daily life, even with platforms with limited processing capacity.