Autonomous Vehicles are Coming to Minnesota's SouthWest Transit Prime ServiceLearn More

Go Home

The path to get on-demand services in mobility apps: GTFS-Flex & the help of Spare GTFS-Flex Builder

The introduction of on-demand public transit has repeatedly shown its role and success in addressing the need for mobility in rural areas. Learn how these tools enhance mobility in rural areas, increase service discoverability, and improve public transit accessibility.


Leo Frachet

Why should we have on-demand services in mobility apps?

Mobility is said to be the first of all rights, because you need it to exercise all the others. Until recently, you couldn’t work, access food, study, get medical assistance or vote without first traveling to specific places. Internet has softened that absolute, but mobility endures as a crucial basic need, which too often complicates life and undermines the rights of disadvantaged persons and communities.

In densely populated areas, traditional public transit plays its role to reduce those inequalities, with its limits and challenges. But in rural areas, poverty rates are higher, while at the same time access to public transit is reduced.

The introduction of on-demand public transit has repeatedly shown its role and success in addressing the need for mobility in rural areas. Furthermore, it has the other positive impacts of public transit, like reducing greenhouse gas emissions.

Slowly, the new industry frontier moved, and now the frontier isn’t the existence of those services anymore, but their discoverability. Asking to use a dedicated app for a specific on-demand service requires the rider to know about the app, where to find it, and how to download it.

Another path is possible, where those services are integrated into the widely used apps. This will put in front of the eyes of the community their new options, increasing the discoverability of on-demand transit options, leading to increase of ridership, and a reduction of social inequalities.

This path builds on the historical success of Google Maps and TriMet with fixed-route, with the GTFS standard. And it has been opened by the recent adoption of the extension of GTFS for on-demand, called GTFS-Flex, thanks to MobilityData. The last mile to reach that goal starts with the creation of GTFS-Flex dataset by every public transit agency providing on-demand service, and it is with that goal in mind that Spare launches the first free GTFS-Flex Builder.

How are fixed-route services in mobility apps?

How was that problem solved in the fixed-route world? And how did we get all those apps providing public transit information for fixed-route?

Most mobility apps are privately owned. So, does that mean that transit agencies had to pay them to get their fixed-route services listed in the result?

No.

In the world of fixed-route public transit, most, if not all, of those apps are using publicly available descriptions of the public transit service. This publicly available description takes the specific form on an open dataset following an industry-wide used data format.

Data formats are like languages.

They are a way to describe reality using a specific set of words (like a dictionary) and by combining them together to create meaning (like grammar).

A data format becomes a standard when it’s used by various stakeholders to exchange information, just like a language builds its strength by the community of speakers using it, and exchanging thoughts, ideas and emotions with it.

If every public transit agency is describing its service in its own data format (in its own language), then the mobility apps will have to learn those data formats (every single language), which becomes an expensive and almost impossible endeavor.

Instead, the industry agrees, either formally or informally, on a data format which is used as a standard, and behaves as a shared language. Each agency then describes its information in that data format, and each mobility app learns to ingest data with that data format.

In 2005, TriMet (the transit agency of Portland, Oregon, USA) and Google defined a data format called GTFS, which became the standard for passenger information. It’s thanks to the work of visionary individuals like Bibiana McHugh and Mike Gilligan from TriMet, who got the ball rolling.

Today, over 2,500 agencies worldwide publish GTFS data, from its GTFS-JP version defined by the Japanese government, to the description of informal transit in Nicaragua or Kenya. And over a thousand of which are in the United States. As of Fall 2023, GTFS is now even a reporting requirement of the National Transit Database, administered by the Federal Transit Administration.

All of that means that today, if a fixed-route agency describes its service in GTFS, and makes it publicly available, its likelihood to be integrated by the Big Four of mobility apps — Apple Maps, Citymapper, Google Maps & Transit app — becomes extremely high.

Is the GTFS fixed-route success replicable?

Is the success of GTFS a one-off? A miracle that happens only once, thanks to the unique convergence of interests?

Thankfully no.

History has shown that GTFS success for fixed-route is replicable to other modes of transportations.

In 2014, GBFS was created to replicate the GTFS approach to shared-mobility. At first only for bike-share, but it quickly expanded to scooter and car sharing. This was made possible mostly by the work of Mitch Vars, and with the support of various organizations and individuals, including Michael Frumin and Jess Chan-Norris from Motivate.

This allowed the same Big Four of mobility apps — Apple Maps, Citymapper, Google Maps & Transit app — to easily integrate bike-share into their apps, and display it in their trip planner result, sometimes even alongside fixed-route public transit.

This proved the replicability of the open standard and open data approach, along with its incommensurable impact on getting travelers to know about other service offering.

How far along are we with on-demand? What’s GTFS-Flex?

On the on-demand side, the missing link was the standard. Conversations started as early as 2011 with a first draft by Brian Ferris from Google, which led to a pilot project with Trillium and VTrans in Vermont.

The scope of that data format, called GTFS-Flex, is precisely the information you would put on a poster in a bus terminal.

With GTFS-Flex, you can describe:

  • the geographical scope
  • the timeframes
  • the price
  • the conditions and the process to book (call a number, use an app)
Caption: Comparison between GTFS and GTFS-Flex trip planning applications

Around 2018, I wrote a second version called GTFS-Flex v2 that combined all the feedback that the industry provided on v1.

In 2019, I founded the nonprofit MobilityData to host and foster improvements of GTFS, with the help of Aaron Antrim and RMI. MobilityData then hosted the conversations around it, leading to the adoption of GTFS-Flex in March 2024. This adoption has been supported by the mobility industry, from public transit agencies (LA Metro, SEPTA), software vendors (Trillium, Interline), and mobility apps (Transit).

March 2024 is therefore the starting pistol shot for the creation and the integration of standardized on-demand data. As of today, in May 2024, already six companies are consuming or producing it: IBI, OpenTripPlanner, TfNSW, Transit, Trillium, TfNSW and us, Spare.

This makes Transit, a mobility app used in more than 300 cities worldwide, the first mobility app to support GTFS-Flex. And the three others from the Big Four are likely to follow suit in short or mid term.

What’s next? More agencies producing GTFS-Flex.

The key is to continue to apply the successful blueprint that the GTFS success story created. The next step is to encourage more public transit agencies to describe their service in that format by producing GTFS-Flex datasets and keeping them updated.

The question is – how?

Today, Spare offers to every one of its customers the feature to natively export their service in GTFS-Flex.

However, we believe that we could have an even wider impact…

And that’s where we decided to create the first free GTFS-Flex Builder.

Spare GTFS-Flex Builder & its impact on agencies

The GTFS-Flex Builder is a free online tool that empowers small and mid-size public transit agencies to create, edit and update the description of their on-demand service as GTFS-Flex.

In just a few minutes, you can create and download a GTFS-Flex dataset that can be shared with partners, providers and third parties. You can also use it as a sandbox to create your first GTFS-Flex data and test it in your tools.

Since this is a brand new tool, we are eager for community feedback. You can reach out directly to us at hello@sparelabs.com or you can provide feedback by creating a GitHub Issue in the GitHub community repository.

By making such a tool available, Spare is enabling the creation of more GTFS-Flex datasets, which will contribute to the momentum around GTFS-Flex in the industry, leading to a broader interoperability, an increase in ridership, and finally, an improvement of rider experience and effective service.


Trusted across four continents and counting, Spare Platform enables cities and transportation operators to plan, launch, operate, and analyze the next generation of mobility services, including integrated on-demand microtransit systems. Some of the world's largest cities are using Spare Platform as a core piece of the city’s transportation infrastructure to build next-generation transit and paratransit systems - like Fukuoka and also Dallas, Texas, which was named APTA's most innovative transit system in North America for 2019. To learn more about Spare, visit https://sparelabs.com