Harnessing Machine Learning for Analyzing Key Attributes of MaaS and PID in Inland Water Transport Ridership

Abstract

In today's technology-driven era of smartphones and data-sharing platforms, commuter demand for Mobility as a Service (MaaS) and Passenger Information Displays (PID) is rising. However, ambiguity in planning due to missing real-time information remains a key factor affecting public transit ridership. This study addresses this issue by quantifying user and non-user preferences of information attributes in the context of Inland Water Transport (IWT) in developing countries like India. Detailed literature review revealed 19 information attributes of MaaS and PID. A total of 1,806 responses were collected from captive and choice riders of IWT in Kochi, Kerala, India. A three stage hypothetical attitude intervention methodology using sophisticated clustering techniques like Fuzzy- C means and Self Organizing Map (SOM) was used for data analysis. The study unveiled Occupancy and safety certification as the critical performers. The prioritization of attributes was done and the applicability of similar studies in emerging economies were discussed.