Transit Hub: A Smart Decision Support System for Public Transit
pdf
Abstract:
Transit hub is a city scale smart phone application that provides real-time, contextual transit travel planning services. It uses a centralized decision support system for integrating the different sensor data streams including the automated vehicle locator, the usage data gathered from the application, and other related data streams that are publically available. During the presentation, we will describe the architecture of this system and describe the core techniques used to create the “lost time” predictive model, which identifies the average lost time on transit trips. This model is then used to present an improved predicted delay information to the end users when they are planning their trips.
Demonstration Abstract:
We will demonstrate the application that the end user uses to plan their trips and receive real-time alerts as the traffic situation changes. In addition, we will show the analytics dashboard of the transit hub project. This dashboard will be used by MTA in future. The source of data used for this analysis includes the Automated Vehicle Locator information from the buses and the static transit schedule we receive from the MTA. During the demo we will demonstrate the capability of the analytics engine to identify the sources of “lost time” on transit trips on our key routes and measure the correlation between them and delays. The lost time is defined as the delay accrued by the vehicle over its scheduled trip. While some of the lost time is due to random variations, often there are correlations between other parameters such as the dwell time at stops for passenger boarding and fare collection, which in turn is correlated to the number of passengers getting on the bus, which often depends on the time of the day. Identifying consistent patterns leading to lost time helps the MTA refine their transit schedule. It helps the transit hub decision support system present an improved predicted delay information to the end users when they are planning their trips. The future capabilities that we have planned for include extending the analysis dashboard to include a ”what-if” simulation analysis for the MTA planners, which will use a predictive simulator to enable them to study the impact of different strategic decisions.
Submitted by Abhishek Dubey
on
Abstract:
Transit hub is a city scale smart phone application that provides real-time, contextual transit travel planning services. It uses a centralized decision support system for integrating the different sensor data streams including the automated vehicle locator, the usage data gathered from the application, and other related data streams that are publically available. During the presentation, we will describe the architecture of this system and describe the core techniques used to create the “lost time” predictive model, which identifies the average lost time on transit trips. This model is then used to present an improved predicted delay information to the end users when they are planning their trips.
Demonstration Abstract:
We will demonstrate the application that the end user uses to plan their trips and receive real-time alerts as the traffic situation changes. In addition, we will show the analytics dashboard of the transit hub project. This dashboard will be used by MTA in future. The source of data used for this analysis includes the Automated Vehicle Locator information from the buses and the static transit schedule we receive from the MTA. During the demo we will demonstrate the capability of the analytics engine to identify the sources of “lost time” on transit trips on our key routes and measure the correlation between them and delays. The lost time is defined as the delay accrued by the vehicle over its scheduled trip. While some of the lost time is due to random variations, often there are correlations between other parameters such as the dwell time at stops for passenger boarding and fare collection, which in turn is correlated to the number of passengers getting on the bus, which often depends on the time of the day. Identifying consistent patterns leading to lost time helps the MTA refine their transit schedule. It helps the transit hub decision support system present an improved predicted delay information to the end users when they are planning their trips. The future capabilities that we have planned for include extending the analysis dashboard to include a ”what-if” simulation analysis for the MTA planners, which will use a predictive simulator to enable them to study the impact of different strategic decisions.