March 16, 2009
|IN DEPTH: FARE COLLECTION TECHNOLOGIES
Breaking Down the ‘Silos’ That Isolate Ridership Data
By CURTIS PIERCE, Associate, Booz Allen Hamilton, San Francisco, CA
As multiple elements of public transit operations become more technically sophisticated, transit agencies have access to more and more operational data. As separate data streams, these elements are useful, providing data for planning, reporting, and analysis.
Some common examples are:
* Farebox data including fares paid and revenue collected;
* Ticket Vending Machine (TVM) and Point of Sale (POS) system data including ticket and pass sales by time and location;
* Automatic People Counter (APC) data that shows boardings and alightings, usually encoded for location; and
* CAD/AVL and SCADA data streams that show vehicle location, route assignments, and timing.
However, too often the data streams exist in functional silos, used only within the departments that generate them. Thus the farebox data never makes it beyond the revenue department, the CAD/AVL and SCADA data is only used by operations, and the APC data is only used by planning.
Fare collection data provides a good example. A recent survey by Booz Allen Hamilton of 20 top operators in the U.S. revealed that, for ridership reporting, half of the operators use farebox data alone.
Keeping data in silos means simply that not all parts of the organization have access to all available data. Integrating data from different sources can actually greatly increase the value of the data. Here are a few examples of current best practices.
By merging CAD/AVL data with farebox data, a large bus operator can eliminate the problem of revenue that could not be assigned to a specific route because of login errors. Additionally, the geocoding in the CAD/AVL stream means that the system’s farebox data is now geocoded, providing much richer data for planning and reporting.
For operators that don’t have APCs on all vehicles, an integration of farebox data with APC data allows for the farebox data to extend and reinforce the APC data.
Several operators are moving toward the principle of having a single source for all data. In other words, planners would use the same numbers that are reported to the board, which would be identical to the numbers filed in the National Transit Database. One operator has consolidated this data mart under its Information Technology department, which has the responsibility of eliminating any alternate data sources—internally called “spreadmarts” because they usually take the form of spreadsheets.
Transit operators looking to improve the integration of their data can take some immediate steps as well as some important considerations for future planning.
The first step for any operator is to inventory all of the available data and map its use. Where is the data generated? In what form? Who manipulates it and how? Where is it ultimately used? This type of survey almost always yields surprises and can often improve operations simply by improving communications between previously separate functions.
To build a good foundation for further integration, operators need to guarantee that the plan for future data integration is part of any procurements of systems that generate data. The key element of this is avoiding any proprietary systems that cannot be accessed from a centralized application or merged with other data streams. Specifying that data be held in an open database with a published schema is an important safeguard against these concerns.
A final best practice to be considered is the incremental integration of data. By breaking the data integration into manageable pieces, operators have been able to provide immediate benefit with relatively moderate cost while building a platform for the future.
By taking stock of how data is currently used and developing an incremental and forward looking plan to integrate the disparate sources, an operator can get significantly more value out of the operational data that already exists within their organization.