Bus Arrival Time Prediction: A Comprehensive Review

Authors:
Publisher website
Google Scholar
Abstract
Bus arrival prediction systems play a crucial role in enhancing the efficiency and reliability of public transportation networks. Existing reviews in this area mainly focus on travel time prediction, without paying enough attention to the data, the patterns in the data, which can help in identifying the regressors for the models, various implementation strategies adopted and on performance evaluation techniques. To address these shortcomings, this paper aims to provide a comprehensive overview of 1) methods available to collect and analyse the travel time data; 2) modelling of bus travel time using naive approaches to physics based and data driven approaches; 3) pattern identification techniques that can identify inputs for the modelling; 4) various implementation strategies including route discretization, multi-step prediction and utilization of data obtained from single route and multiple routes; 5) evaluation metrics and performance evaluation strategies adopted by bus transit agencies followed by challenges faced and way forward. Thus, this paper aims to guide researchers toward developing more accurate, reliable, and user-centric bus arrival time prediction systems, which are crucial for enhancing public transportation efficiency and user experience.

This publication has 153 references indexed in Scilit: