Passenger BIBO detection with IoT support and machine learning techniques for intelligent transport systems

The present article discusses the issue of automation of the CICO (Check-In/Check-Out) process for public transport fare collection systems, using modern tools forming part of the Internet of Things, such as Beacon and Smartphone. It describes the concept of an integrated passenger identification model applying machine learning technology in order to reduce or eliminate the risks associated with the incorrect classification of a smartphone user as a vehicle passenger. This will allow for the construction of an intelligent fare collection system, operating in the BIBO (Be-In/Be-Out) model, implementing the "hands-free" and "pay-as-you-go" approach. The article describes the architecture of the research environment, and the implementation of the elaborated model in the Bad.App4 proprietary solution. We also presented the complete process of concept verification under real-life conditions. Research results were described and supplemented with commentary.