S1E30: Transit Data Accessibility

Nov 18, 2021

Andy Keeton

VP Global Strategy

Between the Lines S1E30: Transit Data Accessibility

with Andrew Salzberg


Why will Transit Data Accessibility help save the planet?

On this week's episode of Between the Lines, we chat with Andrew Salzberg.

Andrew is the Head of Policy at Transit, the largest public transportation app in North America.

From 2019-2020, Andrew was a Loeb fellow at Harvard, where he created the Decarbonizing Transportation newsletter. Before the Loeb fellowship, he created and held a unique executive role at Uber, where he created the first teams focused on partnerships with public transportation agencies and environmental sustainability. And prior to joining Uber, Andrew worked at the World Bank on urban and transport development in China. He teaches at MIT and Columbia, holds a Bachelor of Civil Engineering from McGill University, and a Master’s in Urban Planning from Harvard.

And check out our exclusive commuter playlists on Spotify!


Episode Transcript

-(Intro): Commutifi presents "Between the lines" with Andy Keeton. Each week, we explore the challenging issues transportation demand management professionals face on their journey to transition commuters from driving alone to more sustainable, shared and active commuting habits. Be sure to subscribe to hear next week's episode, and check out our exclusive commuter playlists on Spotify. This is "Between the lines" with Andy Keeton.

-(Andy Keaton): Hi, everyone and welcome aboard to this week's "Between the lines" podcast. I'm Andy Keeton, and today, we're joined by Andrew Salzberg who's the Head of Policy at Transit, the largest public transportation app in North America. From 2019 to 2020, Andrew was a LOEB Fellow at Harvard where he created the decarbonizing transportation newsletter. And before that he created and held a unique executive role at Uber where he created the first teams focused on partnerships with public transportation agencies and environmental sustainability. And prior to all of that, he worked at the World Bank on urban and transport development in China. He teaches at MIT in Colombia, holds a bachelor of civil engineering from McGill University, here in Montréal, and a Master's in Urban Planning from Harvard. So, I think it's safe to say we're pretty excited to have you on today, Andrew. Thanks for being here.

-(Andrew Salzberg): Thanks for having me.

-(Andy Keeton): And today, we're talking about an interesting concept, an interesting topic. We've talked a little bit in the past about a kind of data. We're diving more into that again today, to talk about why Transit data accessibility will help save the planet. And obviously you are working at Transit. Data is a big thing that you all do. Can you first tell us, and those of us that don't know what Transit is, what is Transit and what do you do with data.

-(Andrew Salzberg): Yeah, great. Like you said, Transit is the biggest public transportation app in North America. There are navigation apps like Google Maps that are larger, but both Google and Apple, their primary use case is people driving. So, Transit is everything but driving. If you open the app, depending on where you are, you're definitely going to see public transportation directions, but you might also find your local bike share system, on demand scooters, car share, ride hail... kind of every way you can imagine to get around that's not your own car. That's what Transit is if you open it up as a product across North America, increasingly in Europe and Latin America as well. We help you find the best way to get from A to B, no matter what it is as long as it's not driving yourself. When it comes to data, just to answer your data question, we start diving into that, I mean, we consume, as you might imagine. To get that stuff into your hands, we have to consume data from all those different operators. That might be your public transportation provider in your local area, it might be the bike-share company that operates bike share on your corner, it could be a car share operator... It's a whole wide variety of different services that we have to interact with to make that bundle of options available in a way that actually works for people. So that's a pretty complicated thing, you can imagine, right? How are you getting data in a way that you can use it and consume it, and provide it to someone conveniently in cities all across the world from dozens of different operators and hundreds of different transit agencies? That's kind of the core of the work that happens within Transit.

-(Andy Keeton): Yeah, I think that it gets into this whole idea and some of our listeners may know of terms GTFS, GBFS: General Transit Feed Specification and biking. I think there are now thoughts of G like OFS, or something similar. Can you talk a little bit about that and how... What is all this stuff? What are these acronyms? How does that help?

-(Andrew Salzberg): Yeah, they roll off the tongue first of all, right? And what's catchier than GTFS and GBFS? No one knows how to pronounce them and what they are. The GTFS is the oldest one. And it... The G was originally for Google: the Google Transit Feed Specification, but now it's broadened beyond Google. It first started in Portland with TriMet, one of the most formal looking agencies we got in North America, which were looking at Google Maps, which was, yeah, this is 2005, so still relatively new and kind of getting more and more adapted, and seeing lots of driving directions, but no training directions. And so the dream has been how can transit agencies share information so that people, navigation apps like Transit app, can actually consume it and show it to you. And that sounds simple. We kind of take it for granted, that you can find that information, but it really wasn't. And not just because of the data standards. The GTFS has really helped because now, if you're a small transit agency anywhere in the world, you can publish your schedule information, meaning: where are the stops, when does the bus come, where do they go, what are the routes, in ways that are easy to consume. But also, consumers can take that on without having to build any new code essentially to make that possible. That has been great. But it also took a bit of political will for transit agencies to actually want to share that information. We think now that's obvious, but if you go back even 10 years go, the New York MTA, there's a headline they were suing app developers who were trying to use their data to make schedule information available to people. Because they wanted to build the app and they didn't want to be competing with somebody else. So it's partly like technical standards have made the data sharing possible, but also, I think people have decided, generally speaking, that it's good from a transit agency perspective to make that freely available, and then, have a whole bunch of apps, whether it's Transit or Google or somebody else, kind of consume that and make it available. So yeah, data standards and I think just changes in how people think about what data should be accessible and free over the last 15 years have really changed things a lot. And so I would say on the public transport side, it's kind of standard policy now that you want to make that data available freely, without really restrictive licenses, in standard formats available as widely as you can, and not just schedule, but now we have real-time information, obviously, in a lot of places. And now like you mentioned, the question is... Like we said at the beginning, we're not just Transit information, we're also bikes and scooters and everything else. So there's... GBFS is about six years old as a data standard, and very much modelled on GTFS, works originally for dock bike share and for dockless bike share, scooters... That's pretty widely adopted, but it's far from universal and is much less universal than the transit side. And then like you said, for things like on-demand, taxis, ride hail, micro transit, that stuff, we're working right now with a whole group of people within mobility data which is a non-profit based in Montréal, where we're both sitting, to try and build those data standards for some of the services that don't exist. There's a lot of places that's gone really well. I think public transport's an obvious case where data is really widely available, bike share and scooters, pretty good, but not great and lots of places to improve. And then, more on-demand rides, kind of not even a data standard and not a ton of transparency on the data there.

-(Andy Keeton): I think... I was reading something on your website that kind of maybe put this in perspective in a way people can think about it. You have something like a messaging app, where like if I go on Facebook messenger, I can't message someone on, I don't know, WhatsApp or something. Maybe I can because they're owned by the same company, but... Bad example. But something like email. I can use Gmail, someone else can use Yahoo or Outlook and we can email each other, that's the idea, right? It doesn't matter what your back-end technology is or front-end technology, they're all going to speak to each other.

- Andrew Salzberg): That's the idea, I mean I think it's funny I think mostly people talk about email to complain about email, but in practice, we don't notice how well it works. Like you just mentioned, if I have a Gmail account and I email your corporate account or somebody's university account. Those just connect to each other, and that's so boring it's not even worth mentioning most of the time. But it's really not... Most of the new technology platforms like Messenger in the case of Facebook or other similar messaging apps, they want to be communicating amongst the people within the network, but not more widely. And part of the reason for that is obviously if you're building that service, you want people to come in there and be stuck and see all their friends are in there and kind of join that network. We probably don't want, if we have the choice, our mobility system to work like that: where you have to be in one particular system to have access to all the modes of transportation in your city. Ideally, people would be able to find the services in whatever way it works for them, wherever they want to book it, however they want to book it. I think there's a lot of reasons that's good for you as a consumer, but probably also good, you know, I think the premise of the podcast is how we are making things better for people in a lot of ways. And at the systems level, I think there's a lot of reasons to be more excited about a mobility system, as things go more and more digital, more online, that looks like the flexible system that email has built, and not like what messenger has. And the last thing I'll say is that email is that way not because of random chance, but because literally it's 50 years ago now, in the early 70s that were data standards built out for how message exchange should happen by email. And obviously, there's a lot more people emailing now than there were 50 years ago, but the standard is still there. And so that basic idea of how the message gets sent from one account to another kind of helped underpin that, and that's where we think the kind of unwieldy acronyms like GTFS and GBFS we're just talking about. Those can be the kind of building blocks for the email system we're hoping to have for mobility in the future.

-(Andy Keeton): Right, and obviously to the Transit rider, to the user of the system, they don't need to even know that GTFS exists because you at Transit, elsewhere other apps, are creating this easy-to-use system. So, let's actually dive into that. What is it that a user sees when they go on Transit? What are you doing with data to make it easier for users to actually work with it?

-(Andrew Salzberg): The main thing you see when you open the app is the bus stop next to you, when is the next bus arriving. That's the primary use case as people who are opening it up to get real-time information on where they're going to go, or what their next arrival time for the bus or metro, or whatever system they use. That's the bread and butter of Transit, that's what it started with. But increasingly you're also seeing ways to plan a trip across town, not just to know when the bus on the corner arrives but to say: "Okay, I'm to get here," and "What's the best combination of modes I can use to get there?" So, more trip planning as we've gone further along. And then obviously beyond just Transit which tends to be ranked right at the top of the results when you open the app, you're seeing things like I would hear when I open it at Montréal. I see BIXI, our local bike share system, right in there. You might also find on-demand rides from different providers as well. So yeah trip planning, schedule information, real-time data on where you are, or where you want to go. And then, the next frontier we haven't talked about as much is how you actually pay for your ride, or how you unlock your bikes. All the data standards we've been talking about are either schedule or real-time information, but GTFS data from a transit agency doesn't let you actually pay for your ticket, and that's been a much more complicated problem to solve and a more complex solution to build. And there are not really data standards to speak of, widely speaking, to do that. For the case of... In BIXI Montréal, that's one of the places where I actually use transit to unlock the bike itself, and I can do that without leaving the app and it gives me the code, I can punch it and hop on a bike. And so we have... That's happening in more places you can buy your ticket in the app for dozens of transit agencies, etc. But that's still much more limited than the sort of information finding, so we'd love to build more of that. There are a lot of obstacles both technical and otherwise to doing that, so there is some payment and unlocking but a lot of it is finding your next bus, or finding you right across town and using whatever combination of modes makes sense.

- Andy Keeton): I'm obviously very excited about the idea of being able to have one app to pay for, book, do everything, right. That sounds great. In the meantime, while you all are working to get there, and I can understand that maybe the political issues at hand may be bigger than the technical ones, not sure but... Just having the data itself seems to be particularly valuable. So, how does having this data right in the palm of someone's hand? How does this help someone ditch their car? That's what you're talking about. You can do everything except driving. So, how does this actually help you do that?

-(Andrew Salzberg): Yeah, I think we are in a competition, we being everything that's not the car with people hanging on and using a car, right. I think different people have referred to the mix of car share, bike share, transit as the rebel Alliance and owning your own car as the Empire. So it's a scrappy band of people who, in a lot of places, North American context... You know, Montréal is one of the better places for transit ridership, but it's still a minority most places outside of the core of most cities. So, we're trying to grow that mix, and I think to do that. You know, people ultimately make choices that are most convenient for them. And if you have a car parked in your driveway, you know where it is, you have a sense of how long it takes to get across town, you don't have any kind of uncertainty whether it's going to deliver you where you're going, so it's a pretty convenient product. We all know... I think if you listen to this podcast, or you're a transit user, you know about some of the reasons that car use is not so great from a city design perspective and a whole bunch of other reasons, but we shouldn't delude ourselves that it's not pretty convenient and reliable for those people using it. If you want to get somebody off that behavior, you have to offer them something pretty good, and if you want to do that, it really helps to be able to combine different modes you. I will occasionally take a bike to a transit stop to save myself a transfer and save myself a long walk. And because the data for the bike share system and the transit system are publicly available, transit can suggest that kind of thing. And having really reliable real-time information, to me, is a game changer psychologically for taking something like a bus in particular. Now, I'm old enough to know when they were bus schedules that were printed on the bus stop and maybe they were reliable, but, you know, you couldn't tell and you're always like craning your neck to see if the bus was coming, and being able to watch it come on an app is a pretty important feature. I think particularly for transit agencies to have that same level of service digitally, that some of the newer modes. I used to work at Uber as you mentioned. Obviously, Uber has a pretty slick way to get you a car, so how do we make sure that transit services can kind of compete with what people have grown to expect from understanding where the service is, how long it's going to take, etc. So I think it's really important to have data be not just publicly available but high quality and reliable for all the modes that are not your own car. So, we actually have a hope of beating the car on convenience and a few other music measures so we can actually get people to change their behavior in more places.

-(Andy Keeton): Yeah, I love that. I mean it makes sense to me. I think I've seen a study, I can't remember the exact numbers, but there was something saying that waiting for a bus is you know... People overestimate that time by like 3x because it's so awful and you're like: you're supposed to be here five minutes ago. And it's five minutes, but it feels like 15 and it's like: "I'm never taking the bus again." So, I love this idea that you can walk out the door, and actually know when you're gonna get there, and not have to wait.

-(Andrew Salzberg): If you look in travel models where people try and predict people's behavior, what you said is true. People will throw away a wait time. There's a... It feels like twice as long or three times as long depending on the study when you're waiting for something. And I think it feels particularly worse if you're waiting longer than you expected, or you don't know how long to expect, so if you can bring that down, it can make the overall ride seem like a more convenient alternative.

-(Andy Keeton): Yeah, that makes a lot of sense. So... Okay so, this is great. So we now have this app in our pocket that allows us to know when the train's coming or when the bus is coming, more particularly, what other options are out there. What are the benefits then? So, what is the impact on society, the greater whole outside of the individual, what do we see from places where Transit is being used a lot?

- (Andrew Salzberg) Yeah, I mean I think the thing is obviously... the big benefit we're hoping for is more people to use more modes than our public transport or other combinations of those things. I think the truth is that people are choosing to take a particular trip, partially because there's data available on it and they can find it. But I think, probably, more fundamentally to be honest, and I'm saying this even though I work at a tech company, is the quality of services underlying that, right. It's one thing to know that your bus only comes every 45 minutes, but that's not as good as a bus that comes every five minutes. So I think the majority of what drives people's decision-making. is genuinely the quality of service and what's available and then the layer on top, that can really work well with that, is the kind of stuff that we're doing, to make it understandable and easy to use. So the impact we're hoping for is, ultimately, more behaviour that gets people out of their cars and the reason that's exciting is probably familiar to people who are listening to this, but particularly in the context of climate change, there is... transportation in the US is the largest source of emissions and the largest piece of that is driving your own car. So if you can get people out of cars that's great. Obviously, there's a lot of move towards electric cars, which I'm excited about. But electric cars still have to be produced and they have to be powered by something, all of which takes quite a lot of energy, even if it's renewable. So really, if you could stop driving or stop using your car, the benefits to that are really huge. It doesn't mean that everybody can do it, but I think, if we can make it more compelling to more people, that's going to be positive. So that's the big picture, right. How does data help the services, that are not your car, become more compelling? To give you some specific examples of where I think data can play a really big role...

-(Andy Keeton) Yeah, definitely.

-(Andrew Salzberg) I think some of the newer modes that are on demand, you can have more impact there from a digital service. So, for example we're working with some smaller suburban jurisdictions that are doing things like on-demand transit or microtransit, you might have heard of. So smaller vehicles that aren't 40-foot long buses and that kind of service people have been toying with that idea and it exists in a lot of forms for a long time, but during Covid there is kind of more cities that were embarking on the idea of saying "Well, we're not going to run a fixed bus every so often. We're going to have a zone where you can request a ride and we'll send you a van." Or something like that. And then one of the challenges for those services has been people knowing about them and knowing how to use them and that's the kind of place we're having it available in Transit App, when you open that up and understanding how it works and how those microtransit services connect to the broader system. That's a place where I think something like our service can really have an impact, so we launched in Durham, Ontario, up here in Canada, we got one out of ten people who are using the new microtransit service, we're finding it through Transit. So there's some pretty significant size impact that people discover in the new service, particularly when it's newer, when schedules have changed, when things are on demand. If you have to just walk to your subway and you do the same thing every day, then maybe changing behaviour is not going to be that important from a service like ours, although the reliability of seeing the data is. But for newer stuff on demand there's some good data, we're seeing people actually being able to find that more easily than before.

- (Andy Keeton) That's really interesting. So I hadn't necessarily thought about this from an on-demand transit kind of system, but it obviously makes a lot of sense. Obviously, you need the data behind it to make it work well. And really what you're saying is like here's this suite of solutions, the TDM, Test Data Management solutions, and we talk about this every week in the podcast. We'll you hone in on a single one, we'll say like this is the one that... here's one solution that could work, here's another solution that can work and this is this overlaying layer that says you need this for all of them to work.

-(Andrew Salzberg) Yeah, yeah -(Andy Keeton) So, if we want people out of their cars, we need to provide a good sense of how these systems work and that is coming from data.

-(Andrew Salzberg) Yeah, and that's... I think, that's the thing two people say, well, if the data is available, what's Transit App doing? And I think the question is you can imagine, you just listed, there's a whole lot of different ways you can get around. So, how are you making that information kind of easy to use, compelling, available and not just overwhelming people by dumping every piece of data on them about all the services at the same time?

-(Andy Keeton) Yeah, yeah.

-(Andrew Salzberg) So, we... I think there's a definitely a precursor to having good data available that's accurate, high-quality, real time and decent service that underlies that data, that you have something that people really want to use, but then combining that into something you can open up on your phone and really find something simply, without being overwhelmed by it, is kind of the core of what we're up to.

-(Andy Keeton) I love it, I mean it makes a lot of sense. I remember, this was years ago. My fiancé was... she's like we were waiting for a bus, I think it was in Boulder and she pulls out her phone and I pull out mine and I'm like I have no idea when this bus is coming and she pulls up this app I've never heard of, this is before I was doing anything in transportation and she's like "oh, this is really cool, it shows when the bus is coming" and I was like "no way, that's so weird." It turns out it was Transit and it was like... it was very cool, it worked well and then we're like "Okay, we can take the bus now even when it's snowing," like we don't have to wait at the stop for five minutes, it makes perfect sense and it makes sense that it can also now go beyond just a fixed route bus, whether it's on demand public transit or something else. This is great and you've talked a little bit about where you see kind of the future of this going, you talked about payments, things like that, but let's just kind of concretely ask this question. Where do you see the future of data accessibility going? And what do we need to do, in your opinion? Maybe we're not going in a direction you think we should be going. What do we need to be doing, that we're not?

-(Andrew Salzberg) Yeah, well I think we just put out a long... if people really want to go into the details on this, we put out a guide called the transit guide to open mobility as a service. So mobility as a service is this idea that I would say is... let's call it slightly overhyped, I think it's been in the past, despite the fact that I work at a company that's building, I think, one of the most concrete examples of it. There's been a lot over the last five, six years people have said, mobility is a service. Basically, the idea that you don't have to own your mobility device, your car, you can rely on it and just get the rides you need and that'll be better for a lot of reasons which I agree with and there's even ideas of saying well, you'll be able to pay a subscription fee that will give you access to all kinds of different modes. So you might have 100 bucks a month and you get not just a transit pass, but a transit plus, other stuff pass.

-(Andy Keeton) It sounds great. Yeah, it's great.

-(Andrew Salzberg) I think there's a lot of reasons to be excited about that, but I think if you want to get to a world where it's really seamlessly that you can book every mode of transport whenever you need it, maybe even bundle some of those app and pay for them in a subscription, if that's useful for you, I think to get there, if that's the vision we're hoping for, I think we need a few more things. We need to finish out some of the data standards we talked about all the way at the very beginning, that aren't fully adopted, even just finding the information for things like on-demand taxi ride-hailing, etc, that's not always standardized or easy to find. But then, to your point, I think payment's the other real big piece, so right now it can be a bit of a hassle to actually pay for a ride. In the case of public transport agencies, a lot of times you can't buy a mobile ticket or if you can, you have to download a special app each city you're in, to be able to access that and that app might not have trip planning. So you have to kind of like toggle various things, if you're able to and certainly for bike share, scooters, ride-hail, at the best-case scenario, you can be linked to another app but a lot of cases it's just really hard to find and pay for. So I think more openness there would allow, I think, a level of integration that we haven't seen yet. And as more things are available on demand, including public transportation, I think that kind of integration gets more important and we don't want to end up in a situation where all these services are well off from one another. So our vision is that they're easy to integrate, they're easy to operate, they're easy to pay for. It doesn't mean you have to use all of them, but they're all available that you can use kind of when it makes the most sense for you. So to get there, I think we need data standards, very boring, unsexy but important data standards, I think we need the willingness of public agencies to kind of push for some of the openness, we're looking for and I think we need companies to build on top of that and kind of deliver these things to people, like Transit.

-(Andy Keeton) So, to put you on the spot really quickly, do you think this is something we have in five years, ten years? Like is there a time horizon or is it just hopefully at some point, we'll get it?

- Andrew Salzberg) I mean, there are places you can point to, that have different versions of it already, like here in Montreal, over the summer, I could unlock a bike without leaving Transit and I could buy a bus ticket. I couldn't buy a metro ticket for a bunch of reasons that have to do how the tickets are validated and things like that, but there are places where information is almost universally available and you can pay for a ride in a bunch of different places. I think it's not universal even in Montreal and there's work to do here, but the line that the future is here is just not evenly distributed yet. I think it is true about this stuff, there are places, Denver's a good example. You can buy a Denver RTD ticket from the transit agency in Transit App, but also in Uber and Lyft and a few other places. That ability to connect to transit systems is getting better in a lot of locations. There's another layer too of how we do... now that those prices and tickets are more available, how do you make sure people are being offered incentives to maybe reduce congestion or maybe help people make a choice that's more sustainable? That's one level even beyond just being able to pay for things, but I would say mobility as a service is here, in some isolated examples, but we've got a long way to go to make that super seamless and then start to see if we can deliver some of the results you want to see in terms of behaviour change and different mode adoption. That's going to take a little while, but the technical foundations exist, but not that widely adopted yet.

-(Andy Keeton) So we're getting there, we'll be there soon. It's just getting that collaboration, it's how it always comes down to it, it makes a lot of sense. So okay, we're running short on time here, so we come down to our last question and we ask this every week to every guest. Let us know in a few sentences why will transit data accessibility help save the planet.

-(Andrew Salzberg) Got it. Well, if you look in the United States, car and light truck emissions one third of the global total for those emissions are in the United States, then it's the single largest source of emissions. So, if we can get people to change behaviour onto anything but the car, you get huge benefits and if we're going to do that, we need all those modes to be accessible, easily available and data standards are what makes that competition possible.

-(Andy Keeton) I love it, succinct, perfect, couldn't have said it better myself. Everyone who's listening or watching, thank you for being here again this week. Make sure you subscribe to our podcast, wherever it is that you listen to it, as well as on YouTube and take a look, whenever you're in the office and looking for something to watch, and look like you're productive, this is a great thing. You actually are being productive because you're learning and if you want to actually dive more deeply into any of these conversations, we send out a newsletter every week, you can subscribe to that at betweenthelines.io. We'll certainly be sending out this guide for open data standards that you put together. It's dense, I like it, it's a lot. It's a good place, a good thing to dive into over many, many sittings, really thorough, I like it a lot. So find us at betweenthelines.io and we'll see you all again next week. Andrew, thanks for being on.

- Andrew Salzberg) Thanks for having me.

-(Andy Keeton) Perfect. See you everyone.

-(Outro) Thanks for joining us on this week's episode of Between the Lines with Andy Keaton. Be sure to subscribe to hear next week's episode and check out our exclusive commuter playlists on Spotify.

Better commuting starts here.

Better commuting starts here.

Better commuting starts here.