The transit stack

The 'transit stack' is a term to describe the organization of transit technologies and their data for an agency. The number of tools and resources employed by transit service providers to plan, schedule, test, deploy, track, and review operations can be complex and stretched across multiple platforms, vendors, and users.

Learning to describe your transit stack will make it easier to develop RFPs for procurement of new technologies, ensure efforts are not wasted with redundant entry, and reduce the risk errors will crop up between technologies and data sources.

Sample transit stack

How to build your transit stack

1. Make a list of all the software tools used by your agency, from planning all the way through reviewing services with riders.

2. Organize the software tools by when they are used in the operations of your agency. Consider the journey a new bus stop would go through to get disseminated throughout your different systems and tools and documents.

3. Identify which tools 'ingest' data automatically (a person does not enter it manually) and which tools have people entering it manually.

Places to keep an eye on

1. Any time a manual entry is involved, the possibility of errors is introduced. As much as possible, look for solutions and software that can generate and consume data automatically (or at least, with a file upload over manual entry.)

2. Any places where the same data gets entered again, having already been entered somewhere else. A common example is in generating print and web material for riders. It's important to keep operational and rider-facing information in tight sync - if printed schedules are entered manually while the schedule data goes to a tool like CAD/AVL, data conflicts (or rider frustration) are much more likely.

3. Any tools or software that are completely isolated from the rest of the stack. If the headsign software for bus scrolls has no connection to the GTFS, there is no way to leverage the GTFS' headsign values and ensure riders get consistent information. It also means this information has to be kept in two separate locations, and if it is updated in one place, it has to be updated in the other as well.