[Transports] Transition: Simulating and Optimizing a Public Transit Network – Y.Brosseau & P.L Bourbonnais

The Transition platform with Yannick Brosseau and Pierre-Léo Bourbonnais

Walid: welcome everyone to this new episode of Projets Libres!. Today I’m very happy, it’s the first episode of a new series. If you’ve followed the previous episodes, you know that one of my passions is transport, transport networks. And today, we’re going to talk about a project called Transition. It is a project that comes to us from Quebec. It so happens that I met one of the two interviewees of the day, Yannick, at FOSDEM in 2023. He was giving a conference to present this Transition tool and I was interested. We stayed in touch.

To talk about this subject, I have two people, so two guests. The first is Yannick Brosseau and the second is Pierre-Léo Bourbonnais. They both work at the Chaire Mobilité at Polytechnique Montréal. They will explain to us a little bit later what it is. Pierre-Léo and Yannick, delighted to have you on the podcast, we’re going to say today, here we are, on the podcast. Welcome to you and I hope you are well.

Yannick: Thank you, very good. Thank you for having us with us and for allowing us to talk a little bit about what we’re doing here in the pulpit.

Presentation of the guests

Walid: Well, to start, I’m going to ask each of you to introduce yourself, Yannick. Is it your honor that you can introduce yourself, please?

Yannick: yes, so Yannick Brosseau, I’m a computer engineer. I have a career worked in many different fields. I am mainly specialized in the development of free software and in infrastructure support, which led me, a few years ago, to join the Chaire Mobilité and develop transportation tools.

Walid: ok, great. And you, Pierre-Léo?

Pierre-Léo: Yes, I’m a research associate, so it’s a bit like a researcher under the supervision of a professor, Professor Catherine Morency, Polytechnique Montréal at the Chaire Mobilité. I work mainly in data analysis. I work a lot with OpenStreetMap and spatial data, but I’ve also been programming for a while. And that’s pretty much it. I teach a course on the theory of public transit at Polytechnique, then a course on innovations and technologies in transportation. So, it’s one course per session. There you go.

Walid: ok, great. So, for the listeners, we’re going to talk about OpenStreetMap. So, I refer you to the previous episode, which is episode 5 of season 3 (Editor’s note: episode 6), in which we do an introduction to OpenStreetMap with Christian Quest, which can be a good start if you want to understand some of the things of the discussion we’re going to have afterwards.

Presentation of the Polytechnique Montréal Chaire Mobilité

Walid : So now, what I’d like is for you to introduce us a little bit, before we enter Transition, the Chaire Mobilité. Could you explain to us what the Chaire Mobilité is?

Pierre-Léo: The Chaire Mobilité was officially created in 2010, but it started in 2009. It is under the direction of Catherine Morancy, who has been an expert in transportation and mobility in Quebec for quite a long time, and therefore of Polytechnique as well. At the beginning, I was a student at the time with Catherine. I was finishing my bachelor’s degree at the time in mechanical engineering, but I branched off into civil engineering in transportation, because at Polytechnique Montréal, transportation is given in civil engineering at the graduate level. So, I think there is a course on transport at the baccalaureate level (Editor’s note: bachelor’s degree in Europe), but otherwise, it’s in higher education. So, I got there, under Catherine’s direction, at that time. Then there was already, I already had one of our colleagues who is still here today, who started with her. So, at the beginning, the team was two: Catherine is an employee. Now, we’re 7-8 with the whole development team. It has grown in 13-14 years. The idea, at the beginning, was to make the link between the academic, the academic and the transit authorities in Quebec. At first, it was more Montreal, but now it’s all of Quebec.

So, so that there are links and that we can apply things and that there is feedback, so that we can help them with tools, models, simulations, and so that they can send us more recent or more reliable or official data on all transportation in Montreal and Quebec. So, it’s really that connection. And there are also international experts who come to present every year and who come to provide information on what transport experts are doing elsewhere. And it is the partners, the transit companies, the cities and the Quebec Ministry of Transport that fund the Chaire Mobilité , the base. After that, we have piecemeal projects to add funding.

But the basis of the Chaire Mobilité is financed by these transport companies. For all intents and purposes, 90% of Quebec’s public transit operators are present in one way or another in the pulpit. Then some cities. There is the City of Montreal, which works a lot on mobility, cycling and all that. And so does the Ministry of Transport, which is an official partner of the chair.

Pierre-Léo Bourbonnais

Walid: Is it played out at the level of the province of Quebec?

Pierre-Léo: Yes. We will sometimes make contracts with other people elsewhere in Canada, but at least 90% of it is really in Quebec. Although recently, we have new projects with Canada, with Infrastructure Canada. So, it’s getting bigger. We have started to do investigations for all of Canada. So, we’ll see how it goes at that level.

Transition Platform Overview

Walid: Okay, fine. I didn’t say that. I didn’t say it in the introduction, can you tell us what this Transition platform is for?

Yannick: It’s actually the result of a project, precisely Pierre-Léo’s thesis and work during his studies.

It is a platform for planning public transit networks. So we’re talking about buses, metros, trains. So it’s a whole set of that. It is a platform that is really intended to allow operators to essentially analyze networks, finally to carry out simulations on populations and really analyze and optimize transport networks.

Yannick Brosseau

We have optimization and simulation modules at certain levels. We can make comparisons with networks, walking, cycling. We can really analyze the impact of people’s mobility on the networks according to whether there are changes to be made and see where the best performance is. That’s really the Transition platform we’ve developed over the years.

Pierre-Léo: It started, in fact, at the beginning, we had a lot of contracts, precisely, analysis and optimization with the transit companies at the chair. Then well, I started working on tools, prototypes to do certain tasks. Then at some point, we said to ourselves: “well, it would be fun to have a platform where we put all these modules together”. Then that’s when it began, putting together the pieces of the puzzle. Then, after that, well, the team grew. Then we start, we put it open source, then we want to make it more robust, easier to implement with other companies, and then elsewhere in the world. So, that’s the process right now.

Walid: So, the users of the platform are the transport companies, in fact?

Yannick: Essentially, for the moment, yes, the transit companies, the cities, the people who have to plan the networks.

We have a long-term vision to make the tool so easy to use that anyone could use it to try to understand the transportation issues in their region. We’re not really there yet, but the tool is currently aimed at transportation professionals, people who know a little bit well.

Yannick Brosseau

Pierre-Léo: Students too, there are many students who use it.

Yannick: A whole component, indeed, research for the student in our laboratory, but also in other teams, they are starting to use the tool at the research level to understand the different issues.

Walid: Okay, yes. So, it’s a tool that, basically, comes from research?

Pierre-Léo: yes, 100%.

Walid: ok.

Yannick: It’s a bit interesting, precisely, I add the model we have with the Chaire Mobilité, it’s the result of research, but really in collaboration with the end users, the transport companies. So, we really have a feedback loop of, yes, we do research analyses, we have students who do theoretical analyses or analyses with real data, but after that, we can put the tools directly… Well, these analyses, once they are implemented, can be used by direct operators.

Examples of how to use the Transition platform

Walid: Do you have any concrete examples to give to explain a little bit how the transport companies… how they optimize their transportation networks, in fact. Is it variable enough? Are they buses? Is it more with a bike? Finally, who does what with the platform?

Pierre-Léo: Of course, for the moment, Transition is for public transit. So there’s no bike management yet. We have bike-sharing called Bixi and all that. It’s planned, we want to integrate it, but for the moment, it’s strictly public transit. And the transit companies, an example of what they’re going to do, there was recently the implementation of a new rail system called the REM, the Réseau express métropolitain (Editor’s note: see RMTransit’s videos on the subject, for example this one, or this video from Railways Explained), which adds a lot of services in the Montreal area. And then, the transport companies that operate bus networks have to fall back on this new network. So, they redesigned their network around the first stations that were inaugurated two years ago.

So, an example of what they’re going to do is they’re going to overhaul the bus network using Transition. There are other tools too, but they used Transition for that. So they will review each of the lines, change the routes, reduce the number of detours, often, too, that they try to make. To have a more European network, more based on demand. So, they’re going to use the platform to do that. We have also had partnerships with the ministry and with cities to create a network from scratch or to optimize a network that is very fragmentary. So, we used a genetic algorithm to propose a network. For example, we will propose thousands of possible routes and then the algorithm will choose the set of routes that will be optimal according to a given fleet of buses. For example, I’ll say, “Well, I have 50 buses in such and such a city in my fleet budget, what can you offer as a service that would be optimal with that fleet”, using data on people’s movements, where the businesses are, where the points of interest are, where the homes are, and then try to minimize travel times, community, with the public transit system.

So, we did that, they were prototypes, but now, we’re trying to operationalize it so that it’s usable in real life. They have nevertheless used, two or three transit companies have used pieces of what had been proposed to review their network, especially in the suburbs of Montreal, because we are not yet making genetic algae for Montreal Centre, because computing capacity is limited for the moment, but there is nothing that will prevent us from doing so in the future.

The Transition Team

Walid: ok. There is something that I would also like to address while we are on the presentation, it is the presentation of the project, it is the presentation of the team. I would like you to tell us a little bit about the profiles of the people, since you both have quite different profiles. Can you tell us a little bit about the profiles of the people in the team?

Yannick: In the team, if we look at the broad (Editor’s note: in general), we have several software developers, people who have a little different backgrounds too. I do a bit of back-end, we have people who are a bit hybrid, we have very front-end people in development. But many people who have also learned on their own in the open source field. There are a few of us who really have a background in computer science, but Pierre-Léo has a background in mechanical engineering, civil engineering. We have another colleague who was a physical education teacher who taught himself to be self-taught, but who came to us because it is a project of interest. But we also have several professionals from, I would say, analysis transport who work more as customers, as users, who give us feedback on the tools, on use.

So right now, in the team, there are now five people in software development. We are often substituted, sometimes we have interns, sometimes students who contribute bits of tools or even end-of-baccalaureate projects, for example, at the end of engineering training, that people will contribute a module to the tool. That’s the advantage of open source, contributions can come from just about anywhere.

Pierre-Léo: We also have an economist who works part-time, but she still works with us a lot. Because we also find it important to have a foot in everything economic, budgeting, prediction, etc. So, at the statistical level, he is someone who is very strong in statistics, in economics. And after that, we do a lot of IT. I also do data analysis. So, that’s pretty much it.

The project’s MIT license

Walid: ok. And so, the fact that it’s a research project at the base, does that explain the very permissive MIT license that you chose at the base of the project? Was it obvious?

Yannick: Not necessarily, it was still a good discussion that we had.

We wanted to have open source, we wanted to go free, it was obvious, it came from public funds and everything, it must remain. There are very, very expensive commercial tools, but we still wanted to have a tool that was accessible to everyone. But there has been a lot of discussion about the question between LPG and MIT. At first, we leaned more towards the GPL, but in the end, we opted for a permissive license like MIT, precisely to facilitate collaboration with everyone.

Yannick Brosseau

In all circles, sometimes the GPL, in some companies, if we even wanted companies to be able to use it, is going to be a bit off-putting. A GPL license, we can go in a way, a license, one of the most free possible by saying “here’s some code, you can use it in the context you want” by saying, we’ll be able to facilitate as much collaboration as possible this way.

Walid: Does the fact that it’s financed by public funds have to be public code?

Yannick:

Yes, so unfortunately, there is no obligation here in the public research funds to publish the code freely. I have worked a lot in several research groups. I have always said to myself that if the funding is public, the code should remain public. It’s a bit like the evangelism that we sometimes do when we speak, why do you do free software? Yes, it is interesting, but since the funding was paid for by all Quebec taxpayers sometimes, or sometimes all Canadian taxpayers. We want to make sure that there is still a return on investment by making the platform available.

Yannick Brosseau

Walid: Has this project allowed you to have contacts with other organizations?

Yannick: A few, not yet huge collaborations. Indeed, when we go to talk about the project in free places like FOSDEM, it allows me to establish contacts. We’re having discussions with a project called MOTIS, managed by a team in Germany mainly who do things very similar to what we do. And see if there are places where we can collaborate. We didn’t get very far because everyone lacks time at that level, but we can just go and find free tracks everywhere. We don’t want to reinvent the wheel if the wheel has already been invented. So, it allows you to collaborate like that by talking about it everywhere and saying, ah, well, that’s interesting, I can try it. Then, we presented it at conferences such as WCTR, which is one of the major transportation research conferences. People, “it’s like, oh yes, it’s interesting as an access tool, relatively simple”. People look at “OK, how could we use it and everything?”. We are still facing a difficulty. Not all people are computer scientists. Not all transport practitioners are necessarily computer scientists. They are theoreticians, precisely, as at the Polytechnique, from civil engineering. Sometimes it’s other departments, but it’s not necessarily programmers. Our goal of making the platform as easy to use as possible remains quite paramount.

The platform, its possible financing and legal forms in the future

Walid: And now, when you talk about platform, we’ll come back to it in a little more detail later on the technical components. Transition, is it an online platform or is it a platform that transport companies can install in their homes?

Yannick: It’s a web-based thing, it’s an online platform. Theoretically, they could install it in their local infrastructure, but currently, in our research project, we are the ones who actually host the platform and therefore users come to our platform right now. It’s something that leads us to think about how to deploy the user platform. We think it’s open source, people will just download it and install it in their platform and everything. But since it’s not just a desktop software, it’s a web platform, it’s a database, there are still several components, we’ll come back to that later, it’s more difficult. So we are currently thinking about how to structure this. For example, do we create an NPO, a non-profit organization, to offer the platform in SaaS mode, for example, in exchange for a reasonable monetary contribution? It’s really a challenge of free software, to say, we write free code, people will use it, but no, there’s still a step of how to bring it to the end user, which is not necessarily easy when you come out of the field of computer science. I was in the field of computer infrastructure, we use free software all the time because we know about it, but when we get out of the field, it leads to a question of how we get the tool into people’s hands. This is not as easy a question as one might think.

Walid: yes, I find it interesting that you mention this because one of the questions I had afterwards was to know currently, it is financed by public funds, but it was also to understand are these public funds sufficient to sustain your work or if not, what structure and how do you imagine that you will be able to organize your project to be able to have the necessary money to be able to continue working on it in fact?

Yannick: That’s a good question. Right now, that’s it. We are looking at, among other things, future research funds. There is still money available in some areas, because we still had funding for a while to develop the first prototype. Here, we want to go further. But yes, as you say, if we want to watch, then make it available too. Currently, the research fund aspect comes from Quebec, comes from Canada, so we are a little limited geographically, but if there are people in Europe, people elsewhere in America or Asia who want to use it, how can we support this platform? So that’s where the thinking comes to be said, maybe we can rent the platform in SaaS mode, that is to say a monthly subscription that would finance part of the development without leading to exorbitant costs, but just to say, we can maybe finance financing in that way. Or even with subsidies, sometimes end-users who could perhaps finance features specific to the needs. We are thinking about that. This is really the avenue that is the most interesting. Either there’s a foundation, a non-profit behind it, or we collect donations directly a bit like Kickstarter, if you want, to say “Ah, we have features, we’re financing them this way” or some amendments. There is a question that we have in mind. When it’s research, it’s quite easy, but it’s a one-off event. We have a project for 3-4 years, but after 4 years, we have to renew it all the time, see if there are partners, let’s say, we have partners here who are public-public, so we have the transit authorities, but there is also funding from the federal and provincial granting agencies. It’s partners like that, but will the partnership still be there in 5 years? This is always a question that is more problematic. That’s a bit of a difficulty.

Walid: Yes, it’s interesting because you talk about foundation. What would push you, for example, to create a foundation rather than, for example, a company?

Yannick: It could be an option, but in our reflections that we are going there, we want to keep the public aspect. We sell… The tool is used by cities, by transport companies, and by public bodies essentially. We don’t want to make a profit on this platform, this tool. It would be very, very easy, to start a company and then support it in that way. And you can see it a lot in the open source model too. There are still several companies that do it. But often, I find that there is a model that catches the eye. Sometimes, there is a conflict between profitability and public service.

And given that the tool really wants to be there to help develop public services, it can also be private services, there can be private transit companies, urban indoor buses that could use it for their network. But the main thing is to have a public need. We say to ourselves… “Let’s keep it in the public domain, it can be more interesting.” There are other private transport companies that will make similar tools. We don’t necessarily want to compete directly in this area.

Yannick Brosseau

Initial funding

Walid: There, you answered the question a little. One of my questions was the relationship with what we call the transport organising authorities, so the transport companies. What I understand is that, in fact, as everyone is associated with the Chaire Mobilité, the needs naturally come up. It is not you who market them or they who approach you. It’s rather a bit collaborative, you know.

Pierre-Léo: Yes, well, the first time we had funding for three years was before the pandemic. It came from the federal government, so the Canadian government. It’s the research funding to develop this tool. Then right away, we went to see the transport companies that were partners, because each one gave a certain amount. Then the granting agency tripled the amount that had been donated by the transit companies. So, we managed to raise a kitty like that. Then, right away, we went to see the workers in the transit companies who were going to use Transition to find out what their needs were. Then we made the list. Then we manage that in the management of the development of Transition as well. Then we also have communication channels where users in the transport companies can write to us directly and we can respond to them. Sometimes, they have special needs, like in Montreal, the Société de transport de Montréal. Sometimes, we need a more demanding calculation. So there, they will contact us. Then Yannick will be able to open machines that are more powerful temporarily so that it doesn’t cost too much in the cloud. So we still have communications that are done like that. Of course we want to continue, but it would be fun to have other countries or other companies too, and that’s what we want to develop in the future.

Why transport companies are interested in Transition

Walid: If I take for example, now I am a transport company. Already, the ones you talk to, they already had something computerized, they had nothing, in fact. What ultimately drives them to collaborate with you on this tool?

Pierre-Léo: Initially, they used most of the software that had also been developed at Polytechnique by another professor since the 90s. There, unfortunately, he died a few years ago. So now, it’s like we took up the torch, but we didn’t necessarily want to take the software and then continue to develop. Because we had developed ours at the beginning in parallel, so we continued with ours. It doesn’t do exactly the same things, but in the long run, it will make it possible to meet the same needs, and more. So, we developed that.

Also, transport companies use paid, commercial software, such as Astus, then the entire M2 suite, etc. These are sequels that are quite well known, which cost a lot of money. That depends. They all have special agreements with the transit companies. So, what is interesting is that they just have to pay normally… Eventually, they’re going to pay just for hosting, and then for particular services, but there’s not going to be an expensive license, something they have to pay for every year.

Pierre-Léo Bourbonnais

Yannick: I can give another example. I had, at a conference, I was talking to people in the city of Los Angeles, California, and they were saying, yes, they have access, there are some teams that have access to big commercial software that do a lot of things, and then a lot of planning, but sometimes, there are teams that are not in the planning team, who are in other groups, who would like to have, just do smaller, simpler analyses, but who don’t have access because they don’t have enough licenses for all the city’s employees. But by having free software like that, which doesn’t necessarily do 100% of the work of commercial software, but which can do most of what people need to develop smaller analyses, they find it really interesting in complementarity with certain platforms that have more specific needs.

Student training with Transition

Pierre-Léo: There is the whole aspect of the training that is really different from commercial software. And commercial software makes a lot of money by sending trainers to their customers on a regular basis. Instead, we use Transition in courses at Polytechnique, transportation courses. So, the students who took the courses at Poly in transportation, public transit, and then transportation data management, will have used Transition as part of their courses, will have done simulations, etc. So, when they arrive at the transit companies afterwards, I would say that there are easily 20-25% of the employees, new employees of the transit companies who have taken at least one or two courses at Polytechnique in transportation. So they already know how to use it, they already know how to set things up. So, it helps a lot, and it costs a lot less in training. All the documentation is public, we are trying to integrate it more and more into the software. We started this month. Each of the elements, the definitions, the symbols, all the operations to calculate things will be integrated and documented directly in the interface.

All the things that I show in the public transit course, on the theory of public transit, are integrated directly into Transition with the same definitions. So, it also facilitates the training of future students in this. It also allows us to show them all the theory behind it, which is rarely the case, I think, with commercial software that doesn’t necessarily have an interest in giving all the details of how it’s calculated.

Pierre-Léo Bourbonnais

The problem of standards and closed data

Walid: Before we move on to the next point, which is how do you start a project from scratch, I remember in one of the conferences at FOSDEM in 2023, there was someone from the Swiss railways who came and he talked about a free project that they had developed, so at the railway level, to actually be able to store and analyze the data of the measuring trains that run on the network. And actually, one of the things that he was explaining is that one of the most complex things was that proprietary tools, formats are proprietary and you have to pay for the format and you have to pay for the viewer. And so, in fact, you’re a little… You’re a little annoyed because you can’t share your data and if you don’t pay anymore, you can’t see it either. The thing I was also wondering is, through the proprietary solutions on the market, are there also these problems of closed format or is there not this kind of problem on transport data because we use free tools and free formats?

Pierre-Léo: me,

I find the biggest problem with proprietary software is that we don’t have the methodology. How is it calculated? I don’t think we’re going to do reverse engineering to find out how it’s calculated. And that’s always been a problem. It’s also a problem with Google Maps, Google calculators, places of interest. We don’t know how it’s calculated, what their methodology is, what data they decided to remove, what data they kept, why. Whereas in Transition, in everything we do, everything is transparent.

Pierre-Léo Bourbonnais

If you want to do a certain calculation, on the one hand, you have access to the code, but in addition, you document all the equations that are used underneath to calculate things. So that’s for me who is essential at the academic level. I don’t want a student to start doing calculations and not know how it’s calculated and just take the results without question. There can always be errors, there can be data that is bad when it arrives, when there have been problems, just the passenger counters on the buses. It’s never 100% reliable. So, at the end of a day, you can be your counter and count minus 4 because at some point, there were 4 people missing or he missed… So, of course, we have to be able to analyze this data correctly, to know what the method is to correct or accept errors, etc. Whereas in proprietary software, if you don’t know someone who works there, you’ll never really have a clear answer. And even when you have an answer, I always have doubts about whether this is really how it’s done, you never know.

How do I start a project with Transition?

Walid: If now, for example, I am a transport company and I want to start a project on Transition, what do you need to start a project with this transport company?

Yannick: Essentially, the data we need is at three levels to use Transition. The first is that we need a transportation network. So, we use OpenStreetMap to have the base layer of the network, the entire road network, the walking networks and everything, to make the calculation base. So, you need to have a region in which OpenStreetMap data is available. Fortunately, and listeners may have heard about it in previous episodes, this is information that is not everywhere in the world, but it is quite available in the majority of places. This is information that is available.

Walid: In Quebec, it’s very well mapped, I guess?

Yannick: The basic mapping map is not the level of quality that Pierre-Léo would like. Pierre-Léo works a lot with the OpenStreetMap community to correct the data himself or collaboratively. We will have the most accurate data possible. But let’s say, the basic road networks are there, but for example, I’ll give you an example, we are currently insuring in the Montreal area, are all the pedestrian crossings well documented? Because it can affect the calculation of travel times to access the public transit network. It’s really quality levels that we’re going for. The higher the quality of the OpenStreetMap network, the better the computation results will be.

Walid: So, at the very least, you need the road networks.

Yannick: If it’s going to be trains, it takes a rail network, in theory, but we’re working a lot with buses right now, so it takes a road network to be able to travel. Then there is a road network too, including if we assume that there are sidewalks everywhere, so that people can access the network. So, first step, a network of some kind. Second level, if you start from scratch, you can just take that and start drawing lines. Most users, partners, already have a bus network, so they will often import, we will use the GTFS format, which is now a relatively universal standard format, to define public transit networks. So we’ll be able to import the existing network and be able to do calculations on it.

Walid: When you say import, does it mean that this network, you are going to enrich OpenStreetMap?

Yannick: No, the… It’s imported into the tool, into Transition, directly. So, we’re going to display this as a layer on top of the OpenStreetMap data in the tool visualization.

Pierre-Léo:

In OpenStreetMap, let’s put the public transport networks, often, we just have the line with the stops, but we don’t have the schedules, because these are things that change often, and OpenStreetMap is more for something stable. So, in Transition, you’re going to import all the lines, the stops, but in addition, you’re going to have all the schedules for a certain period of the year.

Pierre-Léo Bourbonnais

So there, it allows you to analyze. Then often, you will import the services: weekdays, weekends, it can be Saturday, Sunday, I don’t know, or the same, if the schedule doesn’t change. Then you’ll be able to analyze that and add a line, remove a line, change the services, etc. So, you can edit your entire network and then simulate changes.

Yannick: The calculations in Transition are really done according to the actual schedule that you put in, in a new line that you put in, but it’s really the calculation with the existing schedule, telling yourself, it’s not just theoretical. You take a line, and then there will really be a bus there. It’s like, “OK, you start from a place at a specific time, you walk, how long does it take?” “Will the bus run at a reasonable time?” “If you have a connection to make, what are the reasonable connection times?” So, we’re really going to calculate with good precision real cases of what is being done in the network.

Walid: ok, so you have… On the one hand, the first is the basic data. They must be up to date in OpenStreetMap. And then, afterwards, you add additional data such as hourly data. For example, where is it managed all the notions of road traffic?

Yannick: Unfortunately, this is one of the aspects that we don’t yet have in the tool and that we would like to integrate, that we are thinking about how to integrate.

And the big difficulty with congestion data is that there is very little public data on the issue. We’re used to going to Google Maps and seeing the data, but it’s proprietary data. Here, if we look at Quebec, the Ministry of Transport has some data, street counts on highways and major roads. There is a way to access this data. Then we have other research projects, for example, to use taxi data to have inferred a certain congestion. But these are really major difficulties.

Yannick Brosseau

We are thinking about options. We had discussions with operators, among others. They have real bus movement data. Then, you have real-time data. You have the GTFS, you have the planned data. But you also have the GTFS Real-Time which allows you to have real data. There would also be a way to infer travel times, congestion with it, but it’s still a project, let’s say, to come, in the short or medium term in the tool. But indeed, if there were public data sources on this information or if there are regions that have public data sources on congestion, it would be easy to integrate into the tool and to adapt the path calculation.

Pierre-Léo: There is also a whole methodology. We have a lot of taxi data now, in Quebec. We have access there. Then there is a part that is accessible to the public as well. But with that, taxi data, can a taxi driver make trips, then there are speeds, then ways of getting around that are representative of the rest of the users? Not necessarily. There is a whole methodological treatment that must be done to remove the outliers. Sometimes, when the taxi driver is not in a ride, he may be in a much less hurry. Maybe he’s just going to walk around the streets to do things, we don’t really know what. So, there’s a lot of work we’re doing right now, but we haven’t put it all together to have a congested calculator. But it’s planned. That’s just what is planned. The problem is that, unlike OpenStreetMap, which works all over the world, almost, that’s piece data that needs to be retrieved. Other than Google, there are very few people who have data for just about every city. Then Google, you can’t buy them because it would be too expensive. I don’t even think they sell them on a massive scale like that. So, we have to take it slowly. But yes, with the operators who already use buses, who have travel times with the buses, we will eventually be able to weigh them up.

But then, it’s a problem if a new district is under construction, there are no buses passing there, there are no taxis because it’s not yet built, but if you want to evaluate a bus network, then you’re taken because you’re comparing it with an existing congestion network and a district that doesn’t exist. So that’s really the challenge, how to create things that don’t exist, compare them and make it realistic. That’s really an issue we have right now.

Pierre-Léo Bourbonnais

Walid: That’s a general problem, not only on Transition and not only in Quebec. I guess everyone has the same problem of finding reliable and accessible data sources?

Yannick: It’s often the crux of the matter to make good calculations, to have good data. And that, by the way, puts the third level of data that you need to really do a good analysis with Transition, that’s really what we call demand. So, what are the trips, what are the routes, where people have to leave from, where people have to go from. Most countries conduct surveys, which are called origin-destination surveys. They do big polls to find out where people are going, for what reasons. We have a sister platform of Transition called Evolution, which is used to conduct this kind of online survey, which is also open source. But people do it either on the phone or in person. So, take this data and make simulations for it. We can also do simulations with synthetic populations. If you want, you can just take random points or stuff like that, but if you want to have real data, it’s to try to match as much as possible your real population and the real trips that will be made to be able to do the real analyses and then see the impact of the network.

Pierre-Léo: Because there are two aspects to this: There are residences, where do people live? That’s going quite well because, at least in Quebec, we now have access to public data, the land roll, residential addresses, etc. The part that we don’t have, or very little, is the places of activity: shops, schools, parks, that’s fine, but everything that is, all the businesses, the industries, we have very little information. Google has plenty of them, but we don’t necessarily have access to that. So here, we have to build databases on workplaces, shopping places, leisure places, to be able to send people somewhere, because the origin-destination survey rarely covers enough population to find exact places apart from the big shopping malls and the big schools. But we have to be able to distribute people throughout the country in a simulation from their homes. So that’s an issue we have right now, that we’re working on. The two big issues, I would say, in terms of data, are that. The places of activity that must be acquired, then the data on congestion, on the actual use of the networks, especially for the car, which obviously has an influence most of the time on the speed of the buses.

Walid: These origin-destination surveys, the results, aren’t they public data?

Yannick: It depends on the country. Most countries, you won’t have precise public data, no, indeed. These are generally accessible to cities, to transport operators to be able to carry out precisely these kinds of simulations. Most of the public data is going to be somewhat aggregated, higher-level data. That’s why we’re asking ourselves the question precisely to do simulations with precisely the places of residence, the public data. There is data on OpenStreetMap, but it’s not always complete. In cities where we’ve done very thorough analyses, we’ve often taken the time, when I say we, it’s essentially Pierre-Léo, to really see OpenStreetMap and make sure that all the points of interest, all the businesses have been added, all the schools, all the hospitals, all the specific places. Once we have this data, we can do simulations a little more.

Pierre-Léo: To give you an idea, we collected the statistics.

For a city, for example, with a little less than 100,000 inhabitants, we did a big test in Drummondville , which has reached 80-85,000 inhabitants. It took four months full-time. For me and an intern who is part-time, so, let’s say, let’s say six months full-time, one person, to ensure that the entire city network, the pedestrian networks and places of activity, that everything is checked and validated for a reference year.

Pierre-Léo Bourbonnais

Well, fortunately, after updating, we’re talking about 10% per year of that time. But here, if we wanted to do the whole of Quebec, we are talking about 10 years to one person. 10 people for a year, so that’s big budgets. But we will have no choice because otherwise the data is not reliable enough. So it really requires a lot of work on that front. We also try to get funding for this because transit companies, cities, government agencies, we know that they use Google a lot. Then they pay, it costs them hundreds of thousands of dollars, even millions a year. We’re trying to build partnerships to convince them instead of putting that money in Google, putting it in OpenStreetMap. There, they will be able to control the data. And in the long run, it’s going to cost them a lot less because they’ll just have to make the annual additions with a team that’s going to do it full-time.

Yannick: It’s something where Europe is one step ahead of us. There are several jurisdictions that require cities or transport companies to put information on OpenStreetMap directly. And so, the community keeps the networks up to date. We don’t have that in America yet.

Policies around Open Data

Walid: Precisely, I wanted to make a micro-aside while I have you on hand. In France, we have indeed had legislation for a few years on what we call open data. So, in fact, we have a lot of accessible and public data on transport, our operators, etc. There are lots of tricks. And I wanted to know a little bit if you had similar things or if these are things that were under discussion, precisely, to open up this data. At the beginning, finally, what we said in the episode on OpenStreetMap is that we realize that when we open the data, the first ones who are happy to contribute and especially to use it are the administrations. Because in the end, between them, it’s not always easy. And there, they have a framework where it’s easy to contribute and it’s easy to retrieve the data. So I wanted to know a little bit if there were things in your house…

Yannick: Things are progressing well here on Open Data. There are not really binding legislative frameworks yet, I would say, to force things. But there are a lot of things that are done voluntarily by the different administrations, the different governments. It’s a movement that has been in place for a good fifteen years. So there is still more and more data available. Sometimes, it’s not the data we want. Sometimes, it’s data format that is a little inappropriate or not updated regularly enough. Then there are still silos. We know that there are cities that have all this data in internal databases that could put this data on OpenStreetMap and now share all this data. But it’s not there yet. But there is more and more information. I think there is, for example, if we take Montreal, I think that the city of Montreal makes available on its Open Data site the position of all the trees in the city. There is a lot of information that is a bit fragmentary, but it’s about bringing it all together.

Pierre-Léo: That’s right, it’s the balance that we have trouble having. You know, like the land role, it’s great. I think it is even in Canada that Quebec is the one that gives the most information online. I know that in Ontario, they do not have access to that easily. But at the same time, you have things, you say to yourself “how come we don’t have access to this?”, and then it’s difficult. And I don’t think it’s bad will, I think it’s just a need for training. To put data on OpenStreetMap, you first have to teach a little OpenStreetMap to those who are going to work in geomatics and then in transport. So that’s why we do it, that we show it to the students as soon as possible, because it’s still complex. There is always a reluctance to say OK, but that’s 100% public data, we don’t control who will change. Then there is anyway… There is sometimes vandalism, although in Canada, it’s quite limited. We don’t talk like in Ukraine or Russia right now, where it’s more problematic. But there still has to be a community that follows up on that. That’s what we’re trying to propose, and then not to build ourselves, but to push behind it to convince the public authorities. I talk to a lot of people at the Department of Transport, and we also learned that the Department of Natural Resources and Land Use Planning are putting them together. Often, it’s funny, at one point I had a meeting with them, and then they never really spoke to each other. Then I asked them, ah, the OpenStreetMap data, could we do this, this, that? Then they discovered that each of their colleagues are from the same government, civil servants, working on similar things, and then it kind of opened up the debate. So we have to open doors like that, for people to talk to each other, and then at some point, after that, we can propose a framework that will force cities and all that to put this data in. But even forcing cities is a big word, because cities, small ones in any case, sometimes they don’t have the expertise. Often, they have some of this work done by private firms. So, yes, in this case, there has to be a law or a regulation that says, “OK, you do business with whomever you want, but it has to follow the framework of OpenStreetMap, it has to be compatible”, etc. So, that’s what we tried to push, but that’s a long-term job.

Walid: Do you do map parties ? Are there any map parties or stuff like that?

Yannick: There aren’t any regularly. I already organized them here in Montreal several years ago. But what we’re doing right now, we’re doing training, I’d say, more. We explain to people who study, among other things, at Polytechnique, we organize training on how to contribute to OpenStreetMap, what the different tools are, how to use the data. So, it’s there, right now. We’re trying to extend that a little bit more. In the case, precisely, that there is a lot of data already available, in the past, the map parties when you are 10-15 years old, you had to draw the paths, draw the roads.

Pierre-Léo: It’s a lot of precise data, the type of pedestrian crossing. And all this, are they shelters? There are plenty of ways to do it. Or sometimes, it’s to check the connectivity. Is the sidewalk that the person has just decided on well connected to the rest of the network? Or if it’s disconnected, then there, it creates problems in the path calculators. At that level, it still takes a basic level of training before people go on it, because it happens sometimes that people will correct things and say, oops, there’s a misunderstanding of network connectivity. We must not go too fast either, there is a parallel work that is being done. But I think it’s promising, we’re moving towards something more robust.

Transition’s software stack

Walid: If we now talk a little about technique, what would interest me is to know a little bit what is the software stack you use, what you are basing yourself on, what you have developed, what have you integrated as open source bricks that already exist, to make this Transition platform?

Yannick: Essentially, Transition is a software written, for the most part, in TypeScript, typed JavaScript. Historically, it was JavaScript, and even before that, the first prototypes developed by Pierre-Léo were in Ruby. So, we have evolved towards that. The whole thing is based on a PostgreSQL database with the PostGIS extension for the graphical extension, which allows us to do geographical calculations more efficiently. We have a part of the backend that is equipped with Rust, currently it’s a very small part, but we’re trying to extend it quietly to have accounts that are a little more robust, a little more efficient than JavaScript code in Node.js. And we have a path calculator that is also open source, which is trRouting, which is a bit of an independent project.

Walid: what’s his name, sorry?

Yannick: trRouting, Transition Routing basically, which is a path calculator, really taking point A to point B, which is used for our calculations. One of the things I looked at is precisely can we take this calculator and replace it with another one and take other external modules. So, it’s a job that we look at a bit in our spare time. These are the essential pieces, so a lot of libraries, of course, in Node.js that we use. We divided the project a little in two. We have what we call the sharelib, so a basic library that is used by the various projects that we develop through the Chaire Mobilité, and the whole part there, if you will, specific to Transition, which is calculation, interface, simulation algorithms.

Walid: From memory, because it’s been a while, but I listened to a talk you gave in Los Angeles, and in it, you were talking, Yannick, about the fact that you had rewritten part of the software to make it more reliable. Because precisely, so I don’t know how you were talking about that at the beginning, but because in fact, at the beginning, the code had not necessarily been written to be, well, was not necessarily very maintainable. And that there was a rewriting that had been done over time to make it more reliable, to make it more maintainable.

Yannick: Yes, indeed, it’s code that came a bit from research at the beginning, so prototyping that was put into production. So, there’s a lot of code that’s been rewritten, more structured. Over time, we have added a lot of unit tests. Often, every time we touch a piece of code to modify it, then we say to ourselves, “well ok, there may not have been a piece of unit test code in that corner, we’re going to add a whole series of tests on top of it”. If you look, for example, I worked a lot on trRouting, which got this on a code that is in C++, that didn’t have a lot of tests, that has code that had been developed a little quickly. So a lot of rewriting to make the code a little bit more solid, a little bit more understandable. Well, unfortunately it reduced the performance because we added checks and everything everywhere. We were working on coming back, we’re going to work on that, to do it a little later. But yes, there is a challenge to take code, sometimes, from students. Like, OK, we don’t want to impose very complicated software development techniques either. He’s a student who makes a prototype quickly. It’s kind of our mandate as research associates working on the project to take these prototypes and make them more robust. So, there are still some less robust pieces, but as we went along, when we made the transition from JavaScript to TypeScript, it still removed a whole class of problems that could potentially have. We are trying to solidify. One of the projects we want to do is to convert part of the back-end a little more into Rust to once again make it more robust, more efficient, in several aspects.

Walid: and where is the code hosted?

Yannick: that’s all on GitHub, chairemobilité/transition. You can find everything, the code. All the development, all the work we do, is done mainly in the public sector. Our PRs (Editor’s note: pull requests) are public, all comments are…

Walid: Do you already have people who have contributed or who have contributed on the code? External, I mean.

Yannick: Not yet. We have former students who have contributed a little, who have worked, studied, did their master’s degree, now they work in transportation companies. Sometimes there are people who have contributed, contributions. But it’s still not huge. It’s still essentially driven by us. We can’t wait to have people. From time to time, we have bug reports from people from outside. It’s always interesting because sometimes, we have out-of-the-view, people we don’t know at all who report a problem: “ah, OK, that’s interesting, we’ll look into it. We hadn’t thought of that.” It’s still a lot of small communities here.

The future of Transition

Walid: There is a part that I do quite regularly in interviews, it’s a bit about the challenges, the future, what you’re working on, what are the issues you have and want to work on in the future on this platform.

Yannick: I would say that our biggest problem is to have more time. Anyway, if we look at our issue tracker on GitHub, I think we’re at 550, 600 open issues. Sometimes bugs, but sometimes it’s just suggestions for improvements, modules to add. It’s really about having the time and, slash (/), the funding to hire more people as well, to be able to contribute. Because we have a lot of songs that we want to work on. One of the big pieces that is coming, that we are starting to work on, is the whole aspect of comparing the calculations. Currently, people can do the calculations, but often the analyses will extract the data and compare it externally. So we really want to integrate this directly into the tool. We discussed a lot about the congestion aspect, which is a piece that is always requested by our partners to have calculation times that are a little more, even more real. Then there’s a bit of work, I’d say, robustness, testing, validation of path calculations to make sure. Because sometimes, there are small artifacts. Our calculator is good, but there are small artifacts. We have one of the bugs that is very, very specific to the Montreal transportation network. We have a few metro lines, but there are two in particular that intersect downtown. And so, there are always two possible paths, but the path calculation, it always gives just one. Sometimes, it doesn’t change anything in the final calculation, but it’s the kind of little bug that we want to fix, to make a little more robust in our test suite. Pierre-Léo, what else is on your wish list?

Pierre-Léo: I think I’ll go more international. This means simplifying the installation or, if not the installation, that people can, with a SaaS, use it directly. I think that could help. That’s where we could find contributors from outside the small Quebec environment, because it’s still limited for the moment. I think that would be it. Well, in terms of funding, I think it would be good to convince the transit authorities to put money in again, but right now, at the political level, the money for public transit in Quebec is getting louder rather than improving.

Walid: That was one of my next questions.

Pierre-Léo: It’s difficult, but maybe we can talk about it right away. At the political level, the government is rather right-wing. It’s not the same right as in the United States, but there is still… The people who vote mainly for this party are mainly from the suburbs, and then at the rural level, from the smallest towns that either have no public transit or so little that they don’t see the point. Of course, there are people who would like to have more, but these voters are reluctant to fund public transit because they think it doesn’t really help them. They tend to want more roads because they still think that adding more lanes will go faster. So that, we hit a wall at that level. At the city level, I think it’s improving. In any case, the vision of the biggest cities like Montreal, and even in Quebec City, there is a more progressive vision, but it’s still long.

Currently, there are many, many investments in cycling, but as much as I’m 100% in favour, I think that public transit is a bit of a poor relation right now. Then, you know, when we install a priority lane for bicycles, it’s great, but often, we’ll remove a car lane. If we take away a car lane, it takes away a bus lane too. Then, sometimes, the buses go slower than before. And here, we say, we have to get the buses out of congestion, add reserved lanes. Then, sometimes, it creates conflicts. So, you know, it would take a better balance in investments. public transit and bicycles, because in the long term, in the short term, cycling is great because we sometimes even have a decrease in the number of car users. But in terms of public transit, it’s also decreasing because the network, we don’t increase budgets, everything costs more, salaries cost more, it costs more to build too. But there is more congestion, so the buses go even slower, so it’s a vicious circle. If we put money in, it’s just to cancel the loss of time we have. There is not even a gain in service.

Pierre-Léo Bourbonnais

Yannick: Compared to more European cities, which have a lot of fairly extensive metro networks, in Montreal, we have a good metro, but the metropolitan area still depends a lot on bus networks. And if we go to the other medium-sized cities in Quebec, it’s just bus networks. We have like four metro lines in Montreal, a few commuter train lines. We now have the new REM network that adds a small high-level service, but it doesn’t cover everything. So we still need a lot of buses to cover all the needs.

Walid: I didn’t say, but the REM, for French listeners, is an RER, in fact. Montreal Metropolitan RER.

Pierre-Léo: We’ll say small REM, there, because you have two floors and more cars. We have four carriages, one floor. So, it’s like a mini-metro.

The cost of building infrastructure and insurance

Walid: He’s a little vague about it, but I’m a… a YouTuber who talks about transportation and who is Canadian, called Reece Martin and who has a very good YouTube channel called RMTransit (Editor’s note: he is involved in Pierre-Léo’s courses on public transport). And what also highlights all the time is the cost of construction in North America compared to… For example, in France, we build all the time, so our costs are quite low, in the end, compared to it. And there, there’s the typical example of the Quebec City tramway, for example, where the costs are indeed inflating and everything, it’s also one of the problems you have at home with the cost of construction (Editor’s note: see this video on RMTransit).

Pierre-Léo:

In my opinion, having seen it over the past few years, the main problem is the way contracts are managed. The government is afraid to manage risks, so it will ask the consortium that it will make a contract, a turnkey contract with insurance, and then it wants to make sure that it will not cost more than what is expected, so they will cap a lot, and sometimes it will even cost more. So, there is a lot of money that goes into insurance, just to manage projects, in project planning, in expropriations. These are things, sometimes, that are not calculated in the same way in Europe too. So there is a lack of a lot of transparency of how money is calculated, how budgets are calculated. Then, of course, there is the lack of expertise. And since we don’t do much, each time, we have to bring people from Europe. And besides, to come and help us, we have to train people. That’s extremely expensive.

Pierre-Léo Bourbonnais

In the long run, it should improve, but for now, it’s very, very expensive. But that’s especially when you look at the insurance costs, just before starting a contract. Let’s say, whoever is going to build the Quebec City tramway has to be insured for billions of dollars. Whereas, I mean, it doesn’t make sense, and in Europe, you often do splitting as well. Not everywhere, but there are places. It’s that now, we’re going to say, “OK, such and such a company, you’re going to travel two kilometers. With another company, you go another mile. This other company, you make the station.” So there, the risks of each bidder are much lower. So their insurance costs less. If one of them misses its cost, we have a station where a 2 km station needs to be replaced instead of the whole system that has been badly done and needs to be corrected. So I would like to know that there is a little more latitude so that we can create small companies that will become experts in public transit, in rail and in all that. And who are able to make small contracts and when you put that end to end, it becomes robust. And if one of them misses its cost, we just have one piece of the puzzle to fix instead of all the shit. Let’s hope it will come. I think we’re starting to become aware of it, but before it’s integrated, and then before there are changes at the legal level. Everything that is legal, and then insurance, is very, very heavy. And lawyers are afraid to change, you know. And it pays off for them. The whole legal system is very, very profitable, these contracts, because they know that there are going to be lots of lawsuits, and then a lot of things.

Plans to promote Transition

Walid: Well, we could talk about that for hours. I still have a lot of things to say, but there’s one last subject I wanted to talk about, it was, you started to address it earlier, very quickly, it was, me and Yannick, I saw your conference at FOSDEM. In fact, I’d like to know how, what do you have planned to make Transition and your other projects like the Evolution platform known, but how do you plan to make your work known in Canada, in North America, and then elsewhere, like in Europe?

Yannick: In fact, I think that the crux of the matter, as you sign, is to go and talk about it. So, I try to cover open source conferences, such as FOSDEM or SCALE in the United States and California, and talk about them at that level. We talk about it in the issues, in the places where we are searching. We are starting to talk about it a little more. There is also a Canadian network of transportation researchers that is active. We will try to present in these places.

Eventually, there is a huge conference in the United States called the TRB, which we would like to present, eventually in those places. Speaking of conferences, talking with different collaborators, eventually, for sure, if we have a more commercial offer, we will probably knock on the doors directly of the transport companies and cities, and we will say, “here is a tool, you can use it”. Currently, we are not pushing on a very large scale to try to have too many users because we still have to be able to support them. With the local people, we already have a lot of needs to identify and solve. But if there are people elsewhere in the world, in Europe, who want to try the tool and who want to ask us questions, we remain available to answer and help. We talk about installation problems, but it’s still not very difficult. We have a Docker image that you can simply run in Docker locally and be able to test the tool that way. So, it’s a bit tech-savvy, it’s done relatively well. It’s a bit like the open source philosophy of word of mouth. People will talk about it. We are not looking for growth either. If we were a private company, we might want to look for rapid growth, to have revenues. We want to maintain our projects. We want to develop them in a somewhat sustainable way to be able to withstand and not expose users too much to problems that are too big too quickly. So, we take it easy, talk to people like you on your podcast and then go piece by piece.

Walid: What do you think the feedback from FOSDEM actually means? How do you see this conference and what type of discussion will you have afterwards? What kind of people will you meet? What’s in it for you that you came? Well, you’ve been here twice, I think.

Yannick: I’ve been going to FOSDEM several times. I’ve been there twice with my hat in transit now. Then it’s interesting because now, FOSDEM, since it’s going to be the third year now next February, that there will be a track ( Editor’s note: Railways and Open Transport). For people who don’t know FOSDEM, it’s the Free and Open Source Developer European Meeting, the biggest free software conference in Europe. And it’s all divided into different silos, different tracks, on specific themes, either it’s software like Postgres or areas of operation like source control, stuff like that. So, there is a track now in transport. And what’s interesting, because most of the operators, especially European rail operators, are present, they contribute, they share tools. So, for us, it’s an opportunity to go and talk to them, to talk to people who use computers to say, well, we have a tool that can be useful to you. Or they have tools that take interest. I come back all the time every time with one or two ideas like, ah, they developed this tool. If we could find a way to bridge Transition with their tool, simulation pieces or specific stuff, for example, to the railway. We haven’t tested rail much because in North America, we don’t have many rail companies, whereas in Europe, they may have problems that we should look at a little more. So, it’s a kind of discussion to have other users with a slightly different point of view. And other projects, what I like about open source is that you don’t have to develop everything, I was talking about that earlier. So, are there pieces that we can just take and glue with the rest of the platform to make a kind of ecosystem a little bit broader of transportation planning, but it can be further away. We have a research project that is new to research right now, which touches a little on the intersectionality between housing and transportation and costs. Going to the suburbs would incur higher transportation costs. We start looking at objects. So, we use Transition a little bit to expose this data. So, collaborating with people like that, it allows us to talk about it and to really have a discussion forum on open source software in transportation is still a pretty human opportunity.

Walid: For those who would be interested in coming to the transport track this year, I think you have to arrive well in advance to get a seat in the room. I arrive half an hour early, I sit down and I don’t move until it’s over because places are expensive.

Yannick: The suggestion for any interesting conference at FOSDEM, there are a lot of people, you have to arrive in advance to get a place. The conferences are all recorded, you can find my old conference on Transition online on the FOSDEM website. If you want to be there, you have to be there.

The last word from the guests

Walid: ok, well, we’re coming to the end. A final word, if you want to pass a word to the listeners of Projets Libres!, I’ll let you all have the floor, if you want to say something before we leave.

Pierre-Léo: yes, I’m still quite positive about the future. Even at the level, I’m going to say, I think, I would have thought it would have been more difficult, especially with public servants, and then governments, to get them to talk to each other. I realized that in the end, it’s just that they weren’t doing it because they didn’t really know who was there and then where. So I encourage people to go and see their government and the people who work in the various departments. It’s really surprising because people are curious, really more curious than I would have thought. Then there’s a whole new generation that is more open to free software, to all that, and to making it more transparent, easier to access. So I would say that’s what’s positive. It’s a long time, but there is not necessarily a closed door. You just have to know that the door is there and then go open it. I think that’s positive.

Yannick: I would say, among other things, to the listeners here, if you have a background in computer science and you are interested in the field of transportation, take a look at Transition, just out of curiosity or if you want to help us eventually.

I know a lot of people who are in IT and are looking to, say, make a lasting impact. People are asking themselves how I can reduce the carbon impact, how I can contribute. Developing tools that improve the transportation network is one option. So come and take a look at the platform. Getting involved on OpenStreetMap is also something that helps us all a little bit in this place. So go see these two projects, ask us questions, try it, try your network, and then look at the impacts, it can be interesting.

Yannick Brosseau

There are other projects, there is more than just another open source project in transportation that can be interesting. That’s an interesting aspect, that’s what keeps me interested. I’ve worked in other fields that are a little less socially relevant, but in the field of transportation, it’s still a field that excites me.

Walid: Thank you both very much. It was really exciting. I would certainly have had plenty of other questions to ask you about the transport networks, but at some point, we have to stop. Listen, for the listeners, if you liked it, I’d like you to tell me in the comments on Mastodon. We also have a BlueSky account now, for those who want to go to another platform. So there you have it, don’t hesitate. And then, there will also be other episodes on transport. I think it’s going to become a fairly regular series, because I have a lot of very interesting topics on it that I want to deal with and that are a little different from what I usually do on the podcast. See you soon, be well. And then, Pierre-Léo and Yannick, I hope for a next one. Yannick maybe at FOSDEM 2025.

Yannick: Not probably, I should be there.

Walid: Well, there you go.

Pierre-Léo : thank you, bye bye!

This episode was recorded on November 26, 2024.

License

This podcast is released under the CC BY-SA 4.0 license or later.