WEBVTT 00:00.000 --> 00:15.000 Okay, thank you everyone for passionately waiting for this thing to work, I just go quick. 00:15.000 --> 00:19.000 So I started this project back in 2009. 00:19.000 --> 00:23.000 These me with a friend pretending to have a stunt in a big conference. 00:23.000 --> 00:32.000 The idea of open business at the time was to create something to use the bike sharing network in Barcelona. 00:32.000 --> 00:40.000 And Jan, this was one slide from the first time I did about the project. 00:40.000 --> 00:51.000 But the main thing was about building a free and open source application for using the business system in Barcelona. 00:51.000 --> 00:56.000 This project just needs to be fixed. 00:56.000 --> 01:15.000 Okay, so what started here in Barcelona ended up covering the entire world? 01:15.000 --> 01:22.000 Okay, good. 01:22.000 --> 01:24.000 Thank you. 01:24.000 --> 01:29.000 So the project started really started as an Android app. 01:29.000 --> 01:32.000 So this was the thing that I was trying to build. 01:32.000 --> 01:39.000 During building this app, I realized that accessing this information was more important that actually providing an app. 01:39.000 --> 01:49.000 So in the end providing data to other people to create apps was more fulfilling that just creating yet another app that probably will end up failing anyway. 01:49.000 --> 02:00.000 And so the project kind of focused on pivix and a public API pivix is the Python library that contains all the scraping logic for bike sharing information. 02:00.000 --> 02:06.000 And then the public API is where people can go and request bike sharing information. 02:06.000 --> 02:14.000 When this kind of what the project has been doing, this is a bit of a graph of the number of networks supported in pivix over time. 02:14.000 --> 02:22.000 And as you can see over here, there is kind of a big slope like I started a lot of networks for a bit. 02:22.000 --> 02:26.000 And then suddenly I kind of stopped carrying that much. 02:26.000 --> 02:36.000 And at that time it's where and this free floating systems started appearing like were lots of systems that appeared somewhere and then they were shutting down after some months. 02:36.000 --> 02:44.000 And I really didn't know if I wanted to have support for this weird systems. 02:44.000 --> 02:47.000 At the same time I had been running the project for a bit. 02:47.000 --> 02:51.000 And also something else happened. 02:51.000 --> 02:57.000 So at the start most of the data I was taking was it was coming from regular web scraping. 02:57.000 --> 03:07.000 But over time we can see that dvfs, which is a standard for bike sharing, kind of started taking over. 03:07.000 --> 03:13.000 And then I was kind of in this little valley of despair like what is this project doing? 03:13.000 --> 03:15.000 What I mean, I can keep it running. 03:15.000 --> 03:19.000 Some people the band on it, there's lots of ups that are using this. 03:19.000 --> 03:23.000 Ah, sorry, yeah, really. What I'm doing? 03:23.000 --> 03:27.000 Like I am meeting a debt course with this project. 03:27.000 --> 03:35.000 And then I started thinking why I created the project and where we are nowadays. 03:35.000 --> 03:41.000 And I think that none of the things that I wanted to initially solve are really solved. 03:41.000 --> 03:47.000 Like most of the applications that people are using to use the systems nowadays are closed source. 03:47.000 --> 03:52.000 At the time people prefer to use open source versions because they were better. 03:52.000 --> 03:54.000 Or user continuity versions. 03:54.000 --> 04:02.000 Nowadays the official apps are good enough that people do not use user continuity applications. 04:02.000 --> 04:08.000 I'm using the word user continuity because not all the applications are open source. 04:08.000 --> 04:15.000 But there's like this kind of thing that was for a while in the Android market and everything that there were lots of different apps. 04:15.000 --> 04:18.000 And it was such a fun time to be using apps. 04:18.000 --> 04:27.000 Nowadays it's like three apps and everyone is using either these three apps or the official one that while it doesn't suck so much as it used to suck. 04:27.000 --> 04:32.000 It still sucks and it still does not respect your privacy in any way. 04:32.000 --> 04:45.000 So yeah we need to do something the project is still relevant as a way of making sure that we can keep building stuff. 04:45.000 --> 04:56.000 But luckily I want to acknowledge all the contributors that help keep the project at least kind of maintain and stable during this period of value of this pair. 04:56.000 --> 05:05.000 And then a huge knowledge to internet I had a discussion with them about the problems that the project had or what I wanted the project to be. 05:05.000 --> 05:13.000 And we found a set of tasks that we found that we're setting things for a good for the future. 05:13.000 --> 05:19.000 So this is a name of the real tag 10 minutes, the new stuff. 05:20.000 --> 05:24.000 Why did I need to build such a complicated or fun thing? 05:24.000 --> 05:27.000 I will just jump over this. 05:27.000 --> 05:31.000 But the thing is we have now historical information. 05:31.000 --> 05:42.000 So now that if we assume that kind of open APIs are going to be a reality then we need to go for the next thing which is historical black share data. 05:42.000 --> 05:44.000 So what is historical black share data? 05:44.000 --> 05:47.000 And this is the kind of information that currently we have. 05:47.000 --> 05:50.000 It's like availability data, availability data means this point. 05:50.000 --> 05:56.000 Because let you do the longitudinal and then some bicycles and then some slots. 05:56.000 --> 06:01.000 But then there are systems that are publishing three data for instance. 06:01.000 --> 06:05.000 This is from point A to point B. 06:05.000 --> 06:10.000 We have when this trip happened and then some extra information. 06:10.000 --> 06:12.000 So this very good quality information. 06:12.000 --> 06:20.000 You can do lots of good research with this kind of information but unfortunately it's not common to have this published at all. 06:20.000 --> 06:24.000 So what I came up with is what I call time series of availability data. 06:24.000 --> 06:28.000 It's kind of an hybrid between availability data and trip data. 06:28.000 --> 06:32.000 Of course I cannot get trip data just by serving networks. 06:32.000 --> 06:37.000 So what we have is in different moments of time for a particular station. 06:37.000 --> 06:47.000 This station had zero bikes and it had one bike at this hour and then it jumped again into zero after 10 more minutes. 06:47.000 --> 06:50.000 So this data is now published. 06:50.000 --> 06:53.000 I think this should work as a link. 06:53.000 --> 06:54.000 We press this link. 06:54.000 --> 06:57.000 Like it's published as monthly market files. 06:57.000 --> 07:06.000 This is now loading this market file for the month of December for the local system here in Brussels. 07:06.000 --> 07:16.000 And then we can go through the stations and see how the station kind of moved overtime. 07:16.000 --> 07:23.000 So anyone can just play with these informations nowadays. 07:23.000 --> 07:25.000 Where else is there? 07:25.000 --> 07:35.000 I made a small demo about what else you can build with this information. 07:35.000 --> 07:39.000 It was sorry. 07:39.000 --> 07:45.000 It was in 2020. 07:45.000 --> 07:47.000 Wow. 07:47.000 --> 07:50.000 Yeah, maybe we don't need to see that already. 07:50.000 --> 07:54.000 It's strange to work interesting. 07:54.000 --> 08:01.000 Well, yeah, hookers. 08:01.000 --> 08:06.000 Well, it's an icing. 08:06.000 --> 08:11.000 You can look at it at some point. 08:11.000 --> 08:16.000 Luckily I know some shortcuts. 08:16.000 --> 08:21.000 Okay, so besides historical information that it's very interesting. 08:21.000 --> 08:30.000 If you are interested in statistics or research, I really want all of you to just go there and play with these parking files. 08:30.000 --> 08:33.000 Because it's taking a lot of effort to put them out. 08:33.000 --> 08:40.000 And now that I've been running them for one year, I think there is a lot of information that you can use to create a very cool shit. 08:40.000 --> 08:41.000 Sorry. 08:41.000 --> 08:44.000 Stuff. 08:44.000 --> 08:45.000 What else? 08:45.000 --> 08:48.000 So GVFS is taking over the world. 08:48.000 --> 08:51.000 Yes, but there are some systems that still will not use GVFS. 08:51.000 --> 08:57.000 So what can it be to help these projects move into using GVFS? 08:57.000 --> 09:08.000 And it's creating a GVFS layer that basically it converts traditional systems into my own simplification of GVFS. 09:08.000 --> 09:12.000 Of course, we don't have all the information, but it's kind of like a GVFS. 09:12.000 --> 09:22.000 That means if somebody builds an application that supports GVFS for the feeds that are not in GVFS, they can use the city by example with GVFS. 09:22.000 --> 09:31.000 At the same time, something else interesting happens is that we are also converting feeds that are GVFS into a worse version of GVFS. 09:31.000 --> 09:39.000 And yeah, it's something I need to work on, but at the same time, the good thing is it kind of standardizes multiple versions of GVFS. 09:39.000 --> 09:41.000 Some of them usually have queries. 09:41.000 --> 09:45.000 You know, these companies not always respect standards. 09:46.000 --> 09:49.000 So this kind of kind of polishes a bit. 09:49.000 --> 09:57.000 And then suddenly you have a GVFS that it's a bit lighter on details, but still it's compliant and it will work with most applications. 09:57.000 --> 10:02.000 And we have a manifest, which nowadays supporting nowadays. 10:02.000 --> 10:10.000 Well, as we added support for free floating bikes, this was something that I was very lazy about that in for many years. 10:10.000 --> 10:14.000 And suddenly we got support for free floating bikes and it's nice. 10:14.000 --> 10:23.000 And finally, yeah, because I found that there was no applications that were easy to use and nice. 10:23.000 --> 10:31.000 I decided that it was time to create an application if it loads, it does load. 10:31.000 --> 10:37.000 But basically it's something that you can use on your phone, on your browser. 10:37.000 --> 10:45.000 Usually it's fast, but the reagents that it supports multiple systems at the same time. 10:45.000 --> 10:50.000 And it loads them. So this one system in Barcelona, that this is another. 10:50.000 --> 10:55.000 So for a lot of cities in the world. 10:55.000 --> 11:00.000 They are like they support multiple systems at the same time. 11:00.000 --> 11:08.000 It's like kind of seamlessly including information about multiple systems in the same map. 11:08.000 --> 11:11.000 Which kind of is the future of micromovility. 11:11.000 --> 11:15.000 Like we are going to move from this pattern of just having one provider. 11:15.000 --> 11:17.000 We're going to have in cities five. 11:17.000 --> 11:20.000 I mean, it's already like this, cities have like five or six providers. 11:20.000 --> 11:28.000 So we need to create one app that supports all of them, not just a different app for every system. 11:28.000 --> 11:34.000 And I don't know, think that's all. 11:34.000 --> 11:39.000 Or not, how do I go back? 11:39.000 --> 11:42.000 Yes, that's all, thank you. 11:42.000 --> 11:49.000 Thank you.