Passenger Transportation Analysis Using Smartphone Sensors and Digital Surveys

Abstract

Increasing context awareness plays an essential role in developing intelligent transportation systems. In this paper the focus is on the context of the passenger, using smartphone measurements and available weather data as sensory inputs. One case example is to recognize whether the subject is inside or outside the bus. When this is recognized, the differences between bus types (diesel or electric) from the passengers’ point of view are studied. This is topical as new electric vehicles are rapidly emerging in public transportation. Modern smartphones contain various sensors such as accelerometers, gyroscopes, magnetometers and barometers. They enable access to useful information regarding the mode of transport. We collected data from public buses with smartphones and simultaneously conducted digital passenger survey to merge passengers’ own evaluations for travel conditions to the data. We demonstrate that context recognition using machine learning (ML) algorithms provides useful information for transportation analysis. It can be used together with digital passenger surveys to achieve deeper understanding of dependencies between travel conditions and passenger satisfaction using buses