Using Train Movement Data for Service Management
We led the establishment of a program at a large public transit agency to routinely utilize historical and near-real-time train movement data for a variety of ad-hoc analyses in support routine service management and special projects. At that time, a vendor had just completed a major project to connect local control towers on the newer subway lines to a central operations control center, allowing visibility of movements along with train identification from one single database. Working with the vendor and the local I.T. and Signal departments, we arranged for automated nightly export and processing of the data from the isolated rail control network, making it available for routine analysis. One of the first application of this data was to replace the manual and burdensome random sample observations for auditing en-route service reliability with automatically collected data. We also utilized the data to monitor the day-to-day variability of station dwell times during busy periods, terminal departure timeliness, maintenance track outages, and to monitor the performance impacts of major service changes. This pioneering work prompted the agency to explore more of the passively collected data that had been available within their systems, and eventually led to similar projects that took data directly from relay rooms on the older parts of the system.
Related Publications/Presentations: