WEBVTT 00:00.000 --> 00:14.400 Hi, everyone. I don't have an intro slide, so I'll just do it like that. My name is Kat. I'm from Berlin, 00:14.400 --> 00:25.000 and I'm a freelance data software, or whatever person, and also I like to travel by train, and I guess everybody here as well in the room does. 00:25.000 --> 00:34.000 And in particular, I like to take those long distance, cross-country, multi-day far away places, type of train trips. 00:34.000 --> 00:47.000 And when I talk to people about them, oh, I've heard it all. It's just too slow, it's too expensive, it's too complicated, and honestly, yeah, fair. 00:48.000 --> 01:00.000 It's often, it's mostly slower, it's often more expensive, but in this room, and outside of it, I have met so many people who still want to do it regardless. 01:00.000 --> 01:08.000 Many of which who embrace this loneliness, it can be nice to be in a way somewhere, not just arrive there, right? 01:08.000 --> 01:24.000 But complicated, complicated, that's the killer that is where we are, are losing people. Again, heard from so many about train trips that never actually got through the planning stage, 01:24.000 --> 01:40.000 because just trying to figure out the right way for me, the right a tenery that works for me and the people I'm traveling with, had already been a such a frustrating experience that it completely killed their momentum, 01:40.000 --> 01:48.000 and added something else, and honestly, sometimes that was just simply too fly there. 01:48.000 --> 02:03.000 So I, I totally get it. I guess I'll repeat the people in this room get it to, right? Who here has had a similar frustrating experience planning schedule trip? Okay, that's that is. 02:03.000 --> 02:17.000 So here's mine, the original story of this project basically, so I was in October, 23, I was trying to plan a trip to Corsica, and I hope for an optimistic, I had to multiple, 02:17.000 --> 02:36.000 like of these cross-country train journey planners, and I put in my storage, and my destination, and my when, and then boom, long list of options, and so hard to grasp what is actually going on. 02:36.000 --> 02:44.000 It's like, it's sorted by departure time, which at this point of the planning process, I'm not particularly interested in. 02:44.000 --> 02:58.000 What I really want to see is basically an idea of, I could go this way, I could go that way, and all the alternatives I can do here, these are all the alternatives I can, I can do here. 02:58.000 --> 03:09.000 But instead, I get this, this unstructured list that that hides all of that information, and basically I need to click at every one of them in the video and find out, 03:09.000 --> 03:18.000 I, that's that rule, that's that rule, and then if by miracle, I happen to find an option, that kind of works for me. 03:18.000 --> 03:27.000 I just want to make a few tweaks, well what can I do? The only chance I have is to go back into the search interface, and maybe say, 03:27.000 --> 03:38.000 oh, many increase the transfer time, here a little bit, and then hope, hope, that whatever my result comes back actually still works, 03:38.000 --> 03:46.000 and that not some algorithm has decided that now there's some more optimal way to go and my whole idea is lost. 03:46.000 --> 03:50.000 So I gave up, I went off the internet and I went analog. 03:50.000 --> 04:11.000 I cut out little pieces of paper one hour is one centimeter, and then I pushed those boxes around on the paper calendar, and this is the trip I ended actually up doing, which was lovely by the way, totally recommend doing that one. 04:11.000 --> 04:16.000 It doesn't have to be that way, why can't travel planning simply be fun? 04:16.000 --> 04:27.000 Why can't it be kind of about exploring ideas of just trying things, oh, that could be nice, and there I say it. 04:27.000 --> 04:36.000 Why can't it even be about kind of sparking imagination, sparking data gaming, imagining already while you're sitting there in front of your computer? 04:36.000 --> 05:05.000 Oh, it will be so nice, and I finally arrived from Great Berlin in that warm city, step outside of the train station and take a coffee outside in the sun, or it will be lovely passing through the Alps looking at the mountains through the train window, or along the Mediterranean looking over the water, because I think if people can make that emotional connection already during the planning process. 05:05.000 --> 05:15.000 Quite more likely that they actually follow through, look the trip and hopefully go in it and have a lot of fun. 05:15.000 --> 05:34.000 So I believe one of the issues is that, one of the issues is that criteria when planning such long trips are quite different from the one in full short local trips, and that's why perhaps the standard user interface for joining planning isn't as well suited. 05:34.000 --> 05:45.000 So there's very, very, the experience people that have done these trips, you want to look at buffers, right? You don't want to be missing important trains. 05:45.000 --> 05:55.000 You don't, probably don't want to depend on the last train of the day, because if you missed that one, you'll stuck wherever you are right now, but it's also about opportunities. 05:55.000 --> 06:11.000 Like, oh, I see my train goes, my connection goes through Zurich. I might just want to meet my friend that I have there. Maybe I can even spend the night, then I don't need a hotel room, it could be perfect, right? 06:11.000 --> 06:33.000 So I tried to kind of distribute this a little bit into geographically related things and time related things, but I mean they are, they are like this in a way, they kind of depend on each other and you kind of always have to have both in view at the same time. 06:33.000 --> 06:46.000 Problem is, like this criteria there, obviously, always very personal, but I also think you don't know them when you start traveling, planning your trip. 06:46.000 --> 06:58.000 Some of them, of course, with the buffers and then not depending on the last train, yes, you do know when you could trace them, but many of them are ideas opportunities that just come up while you're planning. 06:58.000 --> 07:07.000 And some of them probably never even to a death far, it's just an intuition, intuition, you kind of look at the trip and all that. 07:07.000 --> 07:20.000 That's not quite right, but I can't even pinpoint what it is that makes it a good itinerary and a bad itinerary, it's just an intuition. 07:20.000 --> 07:29.000 And any and I mean four different people might come up with for the same trip with wildly different itineraries. 07:29.000 --> 07:36.000 I guess you yourself might in different situations come up with four wildly different itineraries. 07:36.000 --> 07:39.000 So here is going from Berlin to Napoli. 07:39.000 --> 07:45.000 There's the first two are going through Munich and Rona. 07:45.000 --> 07:52.000 The other two are going through Zurich and the first one is like the one busy day and you just get there. 07:52.000 --> 07:55.000 For me that bit, this one is the horror. 07:55.000 --> 08:00.000 But I've done these kinds of trips before sometimes you really just need to go somewhere, right? 08:00.000 --> 08:03.000 And I'm sure there's people out there who like these trips. 08:03.000 --> 08:17.000 For me, the second one is the one that leaves me, it takes a bit slower, it disperses it a little bit and gives me time to explore a city I've never been in. 08:17.000 --> 08:21.000 The third one I would love to do if I actually had that friend in Zurich. 08:21.000 --> 08:26.000 For any volunteers. 08:26.000 --> 08:31.000 And this one is actually you can't do Zurich road from Berlin in one day. 08:31.000 --> 08:35.000 So it shouldn't, it probably wouldn't in any journey planning tools it wouldn't fill up. 08:35.000 --> 08:39.000 Because it's worse than the first one because you can't do it in one day. 08:39.000 --> 08:45.000 But if you happen to have that friend in Zurich, why not do that one and why not perhaps even stay a day longer? 08:45.000 --> 08:49.000 Could be fun. 08:49.000 --> 09:01.000 And like I said, it's intuition, it's exploration and that kind of exploration is really hard to do any kind of rigid from to when user interfaces. 09:01.000 --> 09:12.000 Because you'll always be in this situation where you have to change the search input through trial and error to hopefully get the search results you want. 09:12.000 --> 09:17.000 When really, what you want is to just change the search results, right? 09:17.000 --> 09:26.000 To modify it so that it works for you. 09:26.000 --> 09:35.000 So there's a prototype of a user interface that I've been working on running online under this URL. 09:35.000 --> 09:38.000 I also have it on the last slide again. 09:38.000 --> 09:43.000 I'll have some screenshots in here, but it's like just try it out. 09:43.000 --> 09:46.000 It's much more fun that way. 09:46.000 --> 09:54.000 You get more out of it on desktop, but yeah, everybody's on mobile anyway, so yeah, just go over it. 09:54.000 --> 10:02.000 So the question is how could the more flexible more depth of user interface for joining planning look like? 10:02.000 --> 10:17.000 I'm very deliberately not focusing on data just trying to get a prototype to explore some kind of UX things with it, but more on data later. 10:17.000 --> 10:29.000 So you have not unusual, you have like the time view and the geographical view side by side so that you can do time exploration and geographic exploration at the same time, 10:29.000 --> 10:34.000 which, like I said earlier, kind of they are a little bit like this, right? 10:34.000 --> 10:43.000 For geographic exploration, what you get is a map and on the map under the dots, 10:43.000 --> 10:49.000 and these were the representative places that you can go to, so you can click on them and you already get some context information. 10:49.000 --> 10:56.000 I'm not sure if it's readable, but it says here that for example London, from Hamburg, which is our home city right now, 10:56.000 --> 11:00.000 you can reach London with two transfers. 11:00.000 --> 11:09.000 And that already gives you the opportunity to explore a little bit because often you don't know exactly where you're going, right? 11:09.000 --> 11:14.000 You just know, I want to go to the beach there, I want to go into some, I want to go into the mountains. 11:14.000 --> 11:18.000 You don't necessarily have fully formed idea, I want to go there. 11:18.000 --> 11:27.000 So you can explore a bit on the map, which places are easy to get to, and then you click on, show the routes, 11:27.000 --> 11:35.000 and well, it will show you the routes to get there, which you then of course can click on to select, 11:35.000 --> 11:42.000 all pretty usual, pretty common, a user interface everybody knows, but not ready for trains. 11:42.000 --> 11:51.000 Like, for cars, for walking, this is pretty standard, for trains, Google Maps does it, 11:51.000 --> 11:57.000 but most of the times when you see a map, it's for visualization, when we're in journey planners, 11:57.000 --> 12:04.000 like they show you the journey, but you can't really interact with the journey on the map. 12:05.000 --> 12:16.000 There's a few things that still can be added to this, I think the map part of the whole prototype is the much less developed. 12:16.000 --> 12:20.000 So of course you might want to change your journey. 12:20.000 --> 12:27.000 A lot of people have asked for just kind of dragging routes, like, what if you wanted to go through Amsterdam, for example, 12:27.000 --> 12:32.000 where I get to see how that would look like, if that would work for you? 12:32.000 --> 12:38.000 I'm not sure if it's accessible to everybody, people understand that that is possible. 12:38.000 --> 12:43.000 So there's just something to try out, but you can always just make it a click on Amsterdam. 12:43.000 --> 12:49.000 I want to go here, instead, in case this kind of dragging thing doesn't work out. 12:49.000 --> 13:01.000 And another feature, I think, could be going back to what I said earlier about the day dreaming, the thinking of what could be. 13:01.000 --> 13:10.000 Maybe you could even do a little bit to trigger that by displaying this kind of background information, 13:10.000 --> 13:15.000 like little pictures of, here, this is up there, this is what if you take the train to Stockholm, 13:15.000 --> 13:21.000 in the winter, you might, if you're lucky, see a lot of snow outside, it's really beautiful. 13:21.000 --> 13:31.000 And if you take the night train from Milan to Sicily, just below Napoli, you'll wake up and you'll look out of the window, 13:31.000 --> 13:37.000 and you'll see the sea, the Mediterranean for the next three hours. 13:37.000 --> 13:41.000 And who doesn't want to go on that trip right now? 13:41.000 --> 13:59.000 So I think what I think is the real heart of this project is this calendar view, which very obviously is taking what I did on the paper and put it into computer, right? 13:59.000 --> 14:12.000 So when you have clicked in the map on one of the routes, you get a proposed intenary, it's probably not going to be the one that works for you. 14:12.000 --> 14:23.000 So you have a very easy way to change that by just doing drag and drop operations, so you could grab the purple one, when you start moving it, you see these lines appear, 14:23.000 --> 14:29.000 and these lines mean other departures between in this case, it's Paris and London. 14:29.000 --> 14:36.000 Other departures between Paris and London, so you already get an idea of the train run a day, 14:36.000 --> 14:46.000 is that one I need to focus on because it's rare and I really need to plan that one in, or if it's in this, like in this case, where it happens all the time. 14:47.000 --> 15:04.000 So you grab it, you start dragging, the purple lines appear, when you move it into a correct position, you get a preview, and then you can drop it, and then it has, it has moved there, and you have adapted your itinerary. 15:04.000 --> 15:14.000 And you probably get a nice feeling already also for, if that works better for you or not, then there's not, then yeah, just drag and drop it back. 15:14.000 --> 15:30.000 What I also like is that it kind of directly, of course gives you a feeling for the length of the transfer time that you have, and you can directly see also, if you happen to be on the last train of the day. 15:30.000 --> 15:42.000 Like, if there is no line, it'd be below where you are right now, then maybe this is not the one, you want to have, if you as risk, ever as I am, when taking trains. 15:42.000 --> 15:59.000 If you, ideas that this one is more of a buck really, still open on the calendar, the first is right now, you can do very fun stuff, like you can move things around, and you could take the purple one and put it in front of the green one, and then you have. 16:00.000 --> 16:08.000 Be free, be wild, do you think, but it's probably not going to get you to London. 16:08.000 --> 16:15.000 So that's thinking about how to resolve that, whether to make it illegal. 16:15.000 --> 16:28.000 So there are some people who want to do things like, if I move the orange one, maybe the purple one should move as well, I don't know if that's too complicated, a lot of things left to explore here. 16:28.000 --> 16:36.000 And then I think this could be, this could be super cool. If extra information was available, I don't have it. 16:36.000 --> 16:47.000 But if I knew how expensive different trains were, it would be cool to directly display it here in the calendar. 16:47.000 --> 17:00.000 So I did promise, I talk on routing, and I think in a way all of this is routing, right, but to put it together. 17:00.000 --> 17:10.000 What is routing in this context? Number one, we need to find first the past that can get us from some A to some B. 17:10.000 --> 17:18.000 And right now I'm pre-calculating that and I probably can always be pre-calculated. 17:18.000 --> 17:32.000 And then if the user comes and asks for an attenuary on a specific date, that is when you get into action and fill that route with actual traffic connection data. 17:32.000 --> 17:42.000 And it might happen that this connection just doesn't exist on the day that the user is asking, then the user face obviously needs to handle that gracefully. 17:42.000 --> 17:47.000 It doesn't right now, because right now this situation doesn't arise, see next slide. 17:47.000 --> 18:00.000 And then we let the user do the rest. So we just supply them with some proposal route and then they can start dragging and dropping. 18:00.000 --> 18:15.000 Right now, I am on the tool as only supporting four-star cities and I think 20-ish destination cities, which means I could get away with just brute-forcing. 18:15.000 --> 18:31.000 And I actually did manual post-processing, meaning I pruned and this is stupid route. Let's not take that. So again, lots of thought needed and work needed for doing that. 18:31.000 --> 18:44.000 Right now, the tool is full front end. There's no back end. Meaning the second part needed to be super fast so that it's not lagging when the user is clicking on something. 18:44.000 --> 18:49.000 So it's an extremely simple, it's actually called a create stupid itinerary in the code. 18:49.000 --> 19:01.000 It's an extremely, oh, it's not there, sorry. Extremely simple algorithm where you take your, you want, you go from this city to this city to this city to this city. 19:01.000 --> 19:18.000 And you just fill them with the earliest possible connection times and then it might happen that your proposal actually starts at four in the morning. So you push this a little bit further so that it's at a more reasonable starting out while not trying to 19:18.000 --> 19:36.000 have its swap into the next day. And then again, you let the user to the rest. As for four data, like I said, didn't do much on data, wanted to focus on the user interface for now. 19:36.000 --> 19:48.000 Had a problem with, I mean, he told the data. And just pulling different GTFS files from different countries isn't going to, isn't going to cut it because that if they stop at the border. 19:48.000 --> 20:01.000 If it's a cross border turn and they stop at the border, then what's the, what, what are people supposed to do with that? It's just as a feedback to the data producing entities. 20:01.000 --> 20:17.000 So what I've currently did is I pulled the data for Germany at UFS file once. I did, for the, the connection outside of Germany, I kind of manually pulled some data together. I did this back in November. 20:17.000 --> 20:25.000 I'm pretending that that data is still true that every train runs on every day. None of that is obviously correct. 20:25.000 --> 20:34.000 But I think it is enough of a basis of data to explore and see if this mechanism works. 20:35.000 --> 20:47.000 So I've already had interesting discussions here at this conference about people with people about how to make use of existing efforts of data, routing, et cetera. 20:47.000 --> 20:57.000 But if anybody else just come up to me if you have an idea of what could help me could help the project further. 20:57.000 --> 21:19.000 And that's it from me again, just try out the tool. Everything is on on GitHub as well. And thank you. 21:19.000 --> 21:24.000 Yeah, so questions. 21:24.000 --> 21:30.000 Oh yeah, I saw you first at least. You are realizing a dream. 21:30.000 --> 21:35.000 Okay, so the question was I am realizing a dream. Thank you. 21:35.000 --> 21:49.000 I just, some feedback. You said, yeah, if you do an illegal move, what should happen and so on, I think for such a tool to go on to say this does not have, but not prevent to do it. 21:49.000 --> 21:54.000 You know what you need to know, but who will change something to start. 21:54.000 --> 22:12.000 So so this was a question and the answer all in one. So this was back about illegal moves and should it will warn or should it prevent and the audience member said it should just warn it shouldn't prevent and I'm kind of on the same page. 22:12.000 --> 22:19.000 I like to just move things out of the way and then do the other stuff and then I can put it back in. 22:19.000 --> 22:25.000 So I kind of lean towards warning right now as well. So I think you had a question. 22:25.000 --> 22:32.000 Yeah, I already looked at the page before and I also really love it. 22:32.000 --> 22:37.000 But maybe also I'm, because you said you didn't like that. 22:37.000 --> 22:41.000 But then I do that all the time and I kind of love it. 22:41.000 --> 22:54.000 But I still would like to have a tool like this. The thing which I really really hate about it is the one you already said that this cross order going. 22:54.000 --> 23:08.000 Then the data just drops. So I often have like five tabs open with this railway company, this railway company and this railway company to come on. 23:08.000 --> 23:15.000 So this again was a call to European data producers. Please produce better data. 23:15.000 --> 23:20.000 The last thing I saw that you had a booking point. 23:20.000 --> 23:22.000 Oh, yeah, there's nothing there yet. 23:22.000 --> 23:35.000 Yeah, sure, but I think that's also an important thing because because the the pricing is because often it's like a bit of magic. 23:35.000 --> 23:42.000 Oh, yeah, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I totally get it, hoping for the European Commission to make that booking thing happen. 23:42.000 --> 23:46.000 You know, you're, you're saying the room with a mark of sort of restarting. 23:46.000 --> 23:49.000 Um, I'm there's going to be hopefully, I'll certainly. 23:49.000 --> 23:53.000 I'm sharing feature of which just you, you're Ellen Codes all in necessary information. 23:53.000 --> 23:58.000 I think for more for it. If one of your trains doesn't turn up to pick you up and great plan. 23:58.000 --> 24:06.720 I'll take that with me, but it's not on the immediate agenda. 24:06.720 --> 24:15.960 The person far back, it looks like this time it's running late, which has to be mentioned 24:15.960 --> 24:18.760 at the beginning of the lecture. 24:18.760 --> 24:24.960 Okay, so this was a question about incorporating real-time data to deal with delays, which 24:25.040 --> 24:28.040 is very similar to the question beforehand. 24:28.040 --> 24:29.840 Yeah, would be cool. 24:29.840 --> 24:36.480 I think there's still way too much other stuff first to get done. 24:36.480 --> 24:38.720 But yeah, it should exist, totally. 24:38.720 --> 24:39.720 It should exist. 24:39.720 --> 24:40.720 Yeah, yeah. 24:40.720 --> 24:41.720 Totally. 24:41.720 --> 24:42.720 Yeah. 24:42.720 --> 24:43.720 Yeah. 24:43.720 --> 24:44.720 Yeah. 24:44.720 --> 24:45.720 Totally. 24:45.720 --> 24:46.720 Yeah. 24:46.720 --> 24:47.720 Yeah. 24:47.720 --> 24:48.720 Yeah. 24:48.720 --> 24:58.400 I think the pricing is crazy hard, but one like small thing you might be able to introduce 24:58.400 --> 25:04.480 is like a interrail ticket pricing, the prices are relatively minimal, and you could 25:04.480 --> 25:12.720 estimate those to like give some price, because most of the time other methods don't even exist 25:12.720 --> 25:14.720 by it. 25:14.720 --> 25:15.720 Yeah. 25:15.720 --> 25:21.120 What about the proposal about pricing and maybe interrail would be a good thing to 25:21.120 --> 25:22.520 target first? 25:22.520 --> 25:29.920 And actually, I'm thinking also of you, why not have delay prediction directly there in 25:29.920 --> 25:30.920 the calendar? 25:30.920 --> 25:36.760 Okay, so we still have more time than back there. 25:36.760 --> 25:59.800 Yeah, so this is a question about availability, and if that's also a context information 25:59.800 --> 26:05.560 that should be displayed, and yes, like, I mean if the data is there, it's easy to put it there, 26:05.560 --> 26:08.560 but is the data there? 26:08.560 --> 26:09.560 Yeah. 26:09.560 --> 26:14.560 I was just wondering if you knew about the money in 161? 26:14.560 --> 26:15.560 Yes. 26:15.560 --> 26:16.560 Okay. 26:16.560 --> 26:18.560 Maybe it could be a good idea. 26:18.560 --> 26:19.560 Yes. 26:19.560 --> 26:20.560 Yeah. 26:20.560 --> 26:29.560 So this was a question about men in, I know that one, about men in 161 and maybe getting 26:29.560 --> 26:36.560 in touch, and perhaps sharing, having people share during this experiences and so on on the 26:36.560 --> 26:37.560 website. 26:37.560 --> 26:38.560 Okay. 26:38.560 --> 26:39.560 But I think that's it. 26:39.560 --> 26:40.560 One question. 26:40.560 --> 26:41.560 Okay. 26:41.560 --> 26:43.560 That's one question. 26:43.560 --> 26:44.560 That's right. 26:44.560 --> 26:48.560 How do you determine the things of why you think private enterprise has been paid and 26:48.560 --> 26:51.560 say family building something like this? 26:51.560 --> 26:54.560 I mean, yeah. 26:54.560 --> 26:58.560 So this question is why doesn't, why doesn't this exist already in private? 26:59.560 --> 27:02.560 Commercial tools? 27:02.560 --> 27:06.560 I guess they don't actually try to plan metrics.