Intelligent Automatic Door System based on Supervised Learning

The widespread adoption of automatic sliding doors in both commercial and non-commercial environments globally has necessitated the need to improve their efficiency, safety, and mode of operation. The automatic door gives access to go into or outside a building by sensing the approaching individual using sensors. However, it does not have the intuition to understand when a person is not authorized to go outside based on their age limit, for example, children. To address this problem, researchers have proposed solutions ranging from the use of fuzzy logic to rule-based approaches to make automatic doors better than the previous ones. In this study, an AI-based automatic door system is proposed, which uses a supervised machine learning approach to train classifiers using human body measurement. Our evaluation of different classifiers indicates that SVM is capable of classifying the instances correctly while achieving about 88.9% F-score. Thus, the proposed approach is expected to improve the safety of automatic doors, thereby making them smarter and more intelligent.