Active Travel Podcast
Active Travel Podcast - data in active travel, part one
Big data is a big issue right now - and we are perhaps about to realise just how much information Google and Apple have on us. Data is hugely important in understanding how we travel, but while we've been very good at measuring car traffic, how we measure cycling and walking is far more primitive.
David McArthur, at Glasgow University's Urban Big Data Centre, is trying to change that. Using Strava Metro data, and 'spare' CCTV camera capacity he was busy trying to work out who cycles and walks where - until the COVID crisis hit. Now his work is being turned to measure some of the changes we are facing around how we move around, and the new importance active travel is playing in the new normal.
Most methods of measuring active travel only give us part of a picture, however - and while the granular data on our lives is held by tech companies like Apple and Google, we might be glad that data isn't more widely available.
No one method of can capture everyone, though. Is there a way of making sure we are all visible in the right ways, in this new big data world? Is a national data centre for active travel the answer? And where on earth does government cycling data come from?
You can find out more about the Urban Big Data Centre, and David McArthur's work, here: https://www.ubdc.ac.uk/
David McArthur interview FINAL MP3.mp3
Laura Laker [00:00:00] Hi and welcome to the active travel podcast. A brand new podcast brought to you by the Active Travel Academy, which is an academic think tank on all things cycling, walking and micro mobility. It's part of the University of Westminster in London and works in collaboration with folks from inside and outside the university. That's people like me. I'm Laura Laker an active travel journalist working with the Active Travel Academy on this podcast. Amongst other projects, and this is the first of a two part on data in active travel.
Laura Laker [00:00:28] The Active Travel Podcast is joined by David McArthur, who is a senior lecturer in urban studies at the University of Glasgow. David is with us today to talk about two pieces of research. The first is using crowdsourced data from Strava Metro to establish cycling patterns. And the second is using spare CCTV capacity to identify pedestrian volumes and movement, which is not as 'Big Brother' as it seems. David assures me so. Welcome, David. Nice to have you with us.
David McArthur [00:00:56] Thank you very much. Pleasure to be here.
Laura Laker [00:00:58] So you specialise in big data around transport and urban analytics. Can you just tell us a bit about how that works?
David McArthur [00:01:05] I'm based at the Urban Big Data Centre. So this was a center funded a few years ago with the idea that the UK wasn't making the most of the big data revolution. Our job was to try to establish ways in which new forms of data could be used to address substantial social science questions. So my stream of work was in transport. We've tried to look at what datasets are out there, what can they tell us about our transport network and how to improve our cities and what make the limitations of this sort of data. People were quite it was the hype curve where people were very excited it was going to change the world. We were trying to be a bit critical to those ideas.
Laura Laker [00:01:44] And people get very excited about new tech developments as shiny new toys kind of. But it's not always as wonderful as you might think. So can you tell us where transport data is right now and where it's going and presumably focusing on active travel?
David McArthur [00:01:59] It's quite interesting. There is amazing data out there. It's not always accessible, though. One thing we tried to do in the center was to price it out of the hands of data owners. But that's not always so successful. Sometimes there's legal regulatory licensing issues with the data. So if some local authority has used the commercial product or ordinance survey data, they can't necessarily share that data with a third party afterwards. There's also issues of perhaps it's commercially sensitive. So with a deregulated bus network, for instance, the data may be helped by the operator of a bus service. So it might not be available easily to outside researchers, which is a shame because it would be nice to have better data on who takes the bus and where do they go, but it's commercially sensitive information. So there's lots of great data, but the governance issues tend to pose far more challenges than the technical issues of analysing it.
Laura Laker [00:02:55] Obviously, there's going to be privacy issues around people's data, and especially if it contains demographic data or even personal data. So you've got to be very careful about who gets that, haven't you?
David McArthur [00:03:05] Absolutely. We would definitely want the data owners to protect the data subjects. And it's a legal requirement after GDPR especially. Well we always had data protection legislation but I think GDPR sharpened people's focus on this idea. But some of the data, I don't think needs to be shielded quite as much. So cycle counter data of how many people go past that particular point in time, I'm not sure it's so sensitive, but certain people are not happy to share it or they're worried that something might be done with it that they don't like.
Laura Laker [00:03:38] Really? cycle counter data - numbers?
David McArthur [00:03:42] Yes, I've had some arguments with local authorities because they don't want to release it, even though it's six people past this point in an hour. So I think it's as far removed from personal data as you might be able to get.
Laura Laker [00:03:55] That's interesting. I remember writing an article last year, I think it was, collecting cycle counter data from around the UK. And I got maybe a handful, and those are just the visible ones with the totem poles. But it was quite hard to get hold of, which was quite a surprise. And I think I was working on it for a few months, actually, partly because there were a number of issues. Some of the cycle counters broke down and some of the London ones have broken down. So I was kind of waiting on them. But also, like you say, it's quite hard to get information from people, and that's just the ones with the totem poles and the numbers on that are visible. And I guess there must be a lot more embedded in pavements that you just never see.
David McArthur [00:04:33] Yes, there are, there’s some hidden. So the council will have data on them, but maybe you get it, maybe not. But it's a shame not to have that data available for people to use.
Laura Laker [00:04:44] So you're working on both these projects, the pedestrian project and the cycling project, and that was pre-lock down, and obviously life changed for everyone. Since then, people stopped moving around as much. And I'm just wondering obviously the scope of the project is changing as the transport environment changes. And you wrote a couple of blogs about this, didn't you? The phenomenon of COVID and the changes that are happening. And I'm just wondering how much you've changed what you're doing since then.
David McArthur [00:05:10] It's been a really interesting time for transport data because we've often had this fragmented ownership of the datasets, trouble having access to them. Suddenly, though, everyone needed data on who was where and who was moving where and what modes of transport they were using. It's been interesting to see that the tech giants, Apple and Google have been the ones stepping in to provide consistent data across the UK. But a bit of a black box in terms of how does it go from raw data into these aggregates that they're publishing. But this has been used to formulate policy now, so we might be a bit concerned that if we had our own data and we had a national data service for transport data and it had all been there [LNE1] in a consistent way, we could easily have pulled up the information that we needed. But at the moment, as you said, it's a big job to try and gather all of it and that other people have stepped in to provide other versions of it. So it's interesting.
Laura Laker [00:06:07] And where is this Apple and Google data coming from?
David McArthur [00:06:10] I believe Google's using their location service, which sense for people are through combination of G.P.S. and Wi-Fi, looking at what Wi-Fi networks are nearby. I believe Apple is using where people are searching for directions about. So from that, they can infer something about the purpose of the travel was and where it is. And then they've published these mobility reports that you may have seen getting some media coverage, about how activity at different locations has changed over time. So it's very valuable information at the moment, but it's unfortunate we don't necessarily know all the details about how robust is it and is it excluding certain types of people from the analysis.
Laura Laker [00:06:52] People without mobile phones?
David McArthur [00:06:54] Yes. It's one of the key challenges for big data. So it could be people without mobile phones or the privacy conscious people who've opted out of sharing this sort of information. Apple data, it's a particular subset of people that use Apple products. So if you formulate policy based on a subset of people using the technology who are you excluding and who's not been seen?
Laura Laker [00:07:19] Yes, transport poverty is a big issue and we know a lot of inequalities are being exacerbated by the crisis. And Apple products are extremely expensive, not everyone has a smartphone even so, it's fascinating. Can you tell us about how your crowdsourced cycling project works?
David McArthur [00:07:38] One of the first datasets we acquired at the Urban Big Data Centre was from Strava. So you may be familiar with going on, you have some physical activity, usually running or cycling and you log it and then it gives you some information about how fast you are, and did you beat people? Strava takes this raw G.P.S. data matches onto the route network and then they provide an aggregate data product where you can't identify individuals, but you get information about how many people are on each road at different times of day, and what are the origins and destinations that people are moving between. So we've been working with this for several years, what we've been trying to do is to say what can you get out of the data, what are the limits of the data, and what are the biases in the data. So, again, this is another example, where probably not all cyclists are logging all of their trips on Strava. So whatever you see in the Strava data, you have to think this is for a certain type of person. So men dominate the use of Strava, men are overrepresented in cycling anyway, but they're even more overrepresented in Strava. So you have to be careful with your conclusions that you don't end up designing things for men who use Strava, try and design for all kinds of people. And then there's the issue of including people who don't cycle yet, but we might like to encourage the cycle, but they're not in the data. So a couple of things we tried to do, one was to compare it to cycle counter data, to see how the flows that pass a particular point on Strava match up with the flows that count everyone. The evidence there is not bad, it gets the order of magnitude. You can certainly pick out the busy versus the quiet locations. It's not especially accurate in giving you precise numbers. We believe you can probably monitor trends with it the overtime, whether locations are becoming busier or are quieter. So we did some work doing that, many other academics have also done similar work trying to understand what the what the biases are and how well it represents it. We've then gone on to look at if you put in new infrastructure, what impact does it have on cycling flows. So we have a couple of papers which look at that. The interesting thing, maybe it's obvious, but it's nice to measure it is that it seems that it needs to be infrastructure of a certain quality. So if you have segregated infrastructure, you can get something like a twelve to 18 percent increase in the number of people using that route. If it's some sort of bus lane that you let cyclist cycle in and you call it bike infrastructure, it doesn't seem to be particularly popular and doesn't seem to have a measurable effect. So it's nice to be able to put magnitudes on what are the effects of doing these things and what it's worth doing, what isn't worth doing.
Laura Laker [00:10:30] You mentioned the National Data Centre. I'm just thinking about the problem of who you're capturing with this data and the fact that there's different data out there that may have those people in to tech, they might be using it or they or people who want to ride fast, they might be using Strava. But then you've got people in families or older people who don't really have the time or inclination to be faffing around with apps for every element of their lives. But there's potential, perhaps and you mentioned National Data Centre perhaps suggesting that there could be a collection of some of this information and the best way to map out where routes might need to be, for example.
David McArthur [00:11:08] It would be good to have that. I think we're making progress in that direction. The Urban Big Data Centre's helped a bit. So we've gathered some of this data together. Also, organisations like Cycling Scotland, in Scotland at least trying to do some of this work to pull out this data together. But the other point you mentioned there is whether it's comparable between different areas. If you have different sensor technologies, if the maintenance regimes aren't consistent across places, so you have sensors breaking or giving faulty readings, it's not necessarily easy to have something you can compare between locations or over time, but at least having it all in one place so that researchers can go to it and see what's happening I think is useful, it's something we try to help bring about.
Laura Laker [00:11:54] And then if one thing breaks, physical counter breaks, or if you're not capturing every kind of person, then you have other elements of the data as a backup in a way.
David McArthur [00:12:02] Yes. Interestingly, maybe it's the typical story, there's much more progress made on this for counting cars and vehicles. That's taken a lot more seriously. And there is much better data available on collating all the route country data so we can see what the cars are doing. But it would be nice to know what the people are doing.
Laura Laker [00:12:24] And we measure what we care about. And historically, we've cared about car traffic. Prioritising that, reducing delays for drivers in this country, haven't we? And cycling and also walking, more so walking, have really been forgotten in this piece. And another thing you're doing with the Big Data Centre is this pedestrian CCTV projects. We are not spying on people! But you are you're looking at pedestrian volumes and pedestrian movements in Glasgow.
David McArthur [00:12:51] Yes, very, very keen to start with definitely not spying on people! The is work led by my colleague Mark Livingston, I've been working with them on it for a while now. We wanted to look at how can we measure what the pedestrians are doing and what's available. And there are a few spark sensors, which kind of pedestrians, and there wasn't much else. And it didn't cover the areas of Glasgow that we were interested in it at the time. So we said, well, machine vision is a lot better now, we can use machines to count people in objects and images. There's cameras all around Glasgow and the CCTV network. So we said, do you think if we go to the Council and discuss it with them, it might be possible to do something with that? After months of negotiation and safeguards and ethical approval at the university we did a pilot study where we put a machine that sits in the secure CCTV suite controlled by the Council. It goes to four cameras when we started, it moves them to a particular position and snaps an image. The images then run through this algorithm, which counts a number of people. Then the image gets discarded and out of it we get how many people were spotted at this location at this particular point in time.
Laura Laker [00:14:05] I just love the idea of all the CCTV cameras positioned around Glasgow. You've got about 30, haven't you? That's up from four at the beginning. I just love the idea of them sort of sitting there and then every 15 minutes, or half an hour, you're able to take control of them and then turn them around and take a photo. Then they just resume their normal life. There's something a bit James Bond about that.
David McArthur [00:14:27] Yes, certainly every time I walk past one of our cameras, I check my watch to see if it's due to move, but it never has been so far, and now we're stuck at home more. [LNE2] So I have been desperate to see to see it move around. This was another challenge we had to address in the project because the cameras are used for security. So we had to adjust our approach so that it won't swing around if someone's using it and then missing something that the operators are trying to watch. So we were really keen to try and use open software, free software, not interfere with what the camera's main purpose is, but to get something useful out of it. Originally, it was nothing to do with COVID, but now it's become even more useful to have this daily report that we get on with what are pedestrians doing, how many are in different locations, and how is that changing over time? Are there areas that need to be watched?
Laura Laker [00:15:20] And is that because of capacity issues on the pavements?
David McArthur [00:15:23] Yes, some of the locations we've been trying to think of this as a social distancing problem, but certainly in some of the locations, once the counts get over a certain amount, it suggests people are probably going to be too close together because the pavements are smaller or the area the cameras are looking at, isn't that large. And if there's dozens of people visible on the frame at one point, they may not be maintaining social distance, although you can't just tell that's necessarily a problem. It may be people in the same household, it might be people who are compliant. It's still interesting, though, to see and to spot the pattern. So when you get the nice weather or the warm days, the dry days, you see upticks in the number of people out and about. It's quite interesting to see how the patterns change.
Laura Laker [00:16:09] And I think it's useful in planning pavement space, potentially, if there are too many people, because it gets to the point where there's so many people on the pavement, it becomes impossible to socially distance.
David McArthur [00:16:19] Yes, one of my colleagues, Nick Vess, came up with some very clever method to extract from the map of the pavement widths for the whole of the UK. So we were looking at that and then trying to map where is there may maybe space for social distance and where isn't it possible, and perhaps we need to have an intervention. The work was done a few days after we did it as we released a similar product. So I'd like to see that we beat them, we beat them with our open source, transparent way of doing it.
Laura Laker [00:16:46] That was reported in the in a couple of newspapers, wasn't it?
David McArthur [00:16:50] It was. So we should have pushed ours better because we did it first and we were clear about how we did it and set out exactly how you go from the OS data, how you process it and what comes out and what it looks like. But it still is extremely valuable information to have. If you're telling people to stay two meters apart and it's a very narrow pavement, then you have to do something about it.
Laura Laker [00:17:13] Yeah, it makes quite a strong case for action on pavement space and things like pavement parking, perhaps, or closing rows to through traffic.
David McArthur [00:17:20] Absolutely. And I think it's helpful to have the data on pavement widths, to have information about where are the pedestrians, when are the pedestrians there, and then to use that to inform decisions about where might we need to abolish some street parking or close roads or take some other kind of action.
Laura Laker [00:17:37] There are actually very few ways of counting pedestrians, aren't there? There aren't counters in the same way that there are for cycling.
David McArthur [00:17:45] There are fewer of them around. Some of them are used more with a view to looking at how many shoppers are out and about. So maybe we can get some data on shopping streets. But we want data on other places as well. Some of the other counters are placed at very busy points to monitor what's going on. But if we want a picture of the whole city, then this was a sensor network which existed, the cameras were there. They weren't being monitored all the time, so being able to use them to extract this useful data we thought was quite a good idea. And it's something that can be done elsewhere. The software is open source, the methods are clear. It's something any local authority could potentially implement. We also did some validation work, I should say, checking that the counts machine was producing, actually represented how many people were in the image. And it gives a very good performance. Better than we expected because we thought, oh, well, when it rains, probably it's not going to work or if the lighting's poor, might not work, but we didn't really see that. It performed pretty well under all these different conditions.
Laura Laker [00:18:50] So it was originally intended to provide some before and after data, some baseline data ahead of some pedestrian realm improvements in Glasgow, right?
David McArthur [00:19:00] Yes. Glasgow's got a very large public realm improvement project called the Avenues Project, I think is a hundred and forty million pounds, they're spending on upgrading several of their key streets in Glasgow, particularly active travel focused. So they want more people walking, more people cycling, fewer cars on the road. Often, though, the only thing to measure afterwards what's happened, and it's hard to do an evaluation if you only see what happens afterwards. So we tried to get in there to start to say, well, let's start counting now, so that after we can see what does it look like, and hopefully demonstrate the value that you can get from making these improvements. They will certainly know about what happened to the cars, and we don't want to lead people's thinking about what's happening to cars and congestion. Lets present them with their an X percent increase in pedestrians, and a Y percent increase and cyclists, and that's beneficial.
Laura Laker [00:19:54] Yeah, well, it's cheaper than having people on the street counting.
David McArthur [00:19:57] It's much cheaper. There is some of that that goes on. So there's some manual counts done in Glasgow. I think now, the they have someone watch a camera and count the number of people manually. But this is a much more scalable approach. I believe some local authorities also take videos of a point and then they ship the video off to some other country, and someone on a very low wage sits and manually watches hours and hours of video and notes down the number of people. So this is a much more scalable, cheaper solution.
Laura Laker [00:20:27] And how do you see this being used going forward?
David McArthur [00:20:30] It would be nice, ideally if local authorities installed and made the data usually available, then any researcher doing anything on transport or infrastructure or evaluation would have the data available. At the moment, we're still hoping that we can gather baseline data for our evaluation project. But the data certainly now seems more useful for a covert response and trying to understand what's happening. It will also be interesting. This is probably the first time that government has discouraged the use of public transport. So it's going to be very interesting to see what happens to all those trips that would have been made by public transport. Are we going to see them go on to cycling? Are we going to see them become pedestrians or are they going to work from home? So at least in Glasgow now we will have the car data, we will have maybe some bus data, we will have pedestrians and cyclists. So hopefully get a much clearer picture of what's going on.
Laura Laker [00:21:26] When you get things on car traffic levels, increasing or decreasing there's sometimes comparable data on cycling, walking, and presumably they come from fewer data sets?
David McArthur [00:21:37] We often try to find out where these come from and we never seem to get anywhere. So I have a colleague who spent some time trying to find out the source of some of this official data that was presented. But he wasn't successful, many emails and many "Yes, we'll get back to you", but we don't know where they come from, so I'm also a bit puzzled about where some of these statistics come from, what the data is. You might be right that yes, if you tried digging, it's not you don't always get to the bottom. So it may be that it's the cycle counter data that they sort of aggregate and come up with something. It's it's not clear, though.
Laura Laker [00:22:13] Fascinating. But it's easier to understand where the car data comes from, presumably. I'm imagining there's more counters out there and that they're more accurate.
David McArthur [00:22:22] Yes. Many more counters and it's better documented where they are. And it's much it's much clearer what's happening with that.
Laura Laker [00:22:30] So who documents the car figures?
David McArthur [00:22:32] I think Department for Transport has collected a bunch of the trunk road sensor network, even for Scotland, which is Transport Scotland doing it separately. And then local authorities also possess some for their Traffic Signal Control Systems. They may have separate sensors. We've also looked at some of that. Glasgow thankfully makes at least some of its detector data open. So we've done some analysis and looking at what's happening to road traffic in different places. But the national statistics are sometimes somewhat of a mystery. And again, the big tech companies may know better because Google can detect where you're going, and they also have the ability to do some more detection of what modes of transport you're using. So they may have a better idea of what the mode use says and different areas.
Laura Laker [00:23:18] Do you think there's any responsibility there for the tech companies to release this data anonymised?
David McArthur [00:23:24] They've released some of the aggregate data so you can get the overall trends and mobility. But we don't know what's happening underneath. And sometimes that's for good reason because the location data from smartphones can be very disclosive. So it's something that you really do have to be careful with. But it might be nice to have the detailed methodology on how is it all processed, from raw data up to all these indicators that they provide.
Laura Laker [00:23:50] The track and trace system or the proposed system has raised a lot of questions about privacy and phone companies giving people's data out to government and the ethical implications of that.
David McArthur [00:24:01] It's such a difficult one because the data can be so useful for all sorts of purposes. It can help us really, to transport new ways, if we can understand people's door to door journeys. At the moment, the data gets fragmented between, we have one data set for trains, which is patchy, we have some stuff on cycling, we have some stuff on walking, but we don't necessarily understand. Someone starts the day, they go into their different things, different modes, and how does that connect up? And how might we reconfigure things to get more sustainable choices? But then from this data, you can identify where people live, where they work, where their children go to school, all kinds of stuff that you wouldn't want being made available easily. So it's a very interesting trade-offs there, about what people are what people are willing to share and also who they share with. We share a lot with the tech companies, often unknowingly, but sometimes I think when government does it, it's done in a more explicit way and then people maybe react to it more to it. But they aren't necessarily aware of the amount of information they're already disclosing. But I don't know what the solution to all of that will be.
Laura Laker [00:25:13] It's a commodity, data, which is why we have free apps like Strava, though they've now introduced a subscription. But that information is valuable and presumably you pay for Strava Metro cities like Glasgow.
David McArthur [00:25:26] We do. I think they're currently looking at how their model's going to work. They've traditionally been quite helpful, I think, because they've been founded by real cycling enthusiasts. They have had a bit of a social mission with their data, to try to make it available and try to improve cycling around the world. So they've taken quite a different approach, I think, from the other companies. They've seen it as less of a moneymaking tool and a bit more of a campaigning tool. So we do pay for it, but it's not as expensive as they might be able to make it.
Laura Laker [00:26:01] Another thing you've been looking at is hire bike data since lockdown.
David McArthur [00:26:08] Yes, that was in Edinburgh and Just Eat Bikes, they all had some open data. So I thought, I'll have a quick look at it and see what patterns that we can see in it.
Laura Laker [00:26:14] And that was fascinating, wasn't it? Because just prior to lockdown, when people were being told work from home, the number of people who were going on those bikes actually increased. But then there was good weather at the same time that was potentially a bit, 'Well, which one was it?' But then after lockdown, obviously, the number of trips decreased but the distance of the trips and the proportion of round trips increased, which suggests that they were leisure trips. Which is fascinating.
David McArthur [00:26:37] Yeah, I think this is, the weather has been very interesting because as soon as we had locked down, the weather got fantastic, so it made some of the analysis a bit more difficult. I've been working for some statistical modelling recently trying to strip out these effects so we can monitor the underlying trends a bit better. But the hire bike stuff has raised this question about how are these people, perhaps people who didn't cycle before. Are people looking at the empty roads and thinking, oh, maybe I could try cycling? And if they're doing that, will they keep doing it afterwards? Will they get a taste for it? There's been reports of bike sales going up. There's these suggestions that hire bikes have been used in new ways. So it will be very exciting to see if this is something that's going to cause a shift in the number of people that are cycling or is it going to die way afterwards, which in part will depend on the policy response and whether these nice new temporary cycling lanes that have gone up everywhere become something more permanent or whether it's going to go back to the cars afterwards.
Laura Laker [00:27:38] I guess some of the cycling data are amalgamating is going to be useful for that, perhaps pre- and post-.
David McArthur [00:27:44] Yes, unfortunately our Strava data comes in quarterly deliveries. So we have up until the end of March at the moment. So we just got the start of that, we're waiting on July when we get our next quarterly delivery, when we can perhaps start to track a bit more. Where are the journeys coming from? Where are we already going to? Which areas became busier and quieter, and was there more leisure cycling and what was going on? So it should be quite interesting.
Laura Laker [00:28:13] You really need something a bit more agile in these times because things are changing so quickly. I mean, since the end of March, the world's changed beyond recognition, really, hasn't it? The government has told councils they should be doing emergency cycle lanes, making road space for people on bikes and on foot, which we never had before. And that's such a guidance.
David McArthur [00:28:30] Yes, it's one of the benefits of big data supposed to be we get it quickly. But it doesn't always work out that way. So with our CCTV camera data, we get it daily, so we have a very up to date picture. Strava, they have the data, but they just haven't supplied it to their data users. They're currently reengineering the way they supply their data. So I think they're not wanting to get into a position of doing custom deliveries and custom cuts. So they've said, well, look, to stick to the schedule and wait until July and then you'll get your you'll get your new data. So we are like children counting down to Christmas. We turn on our use Strava data to play with. But the time seems to be flying through.
Laura Laker [00:29:12] So it's a whole concept of time has changed. Hasn't which days at what time is it, where am I?
David McArthur [00:29:18] It certainly has, but I'm very interested to see this idea of whether people do change their transport habits. And will people who used to drive have maybe tried to bike for leisure during Lockdown, will they be tempted to say, "Oh, well, I could take it to work" or "I've tried out the route to work at my leisure time, it was alright", maybe "the infrastructure was better than I thought, maybe I will keep doing that ".
Laura Laker [00:29:40] Yeah, that's right. Well, thanks, David, for talking. It's fascinating to hear from you and hear what's going on with big data in active travel. And you have to keep us posted what's happening in July with that Strava data?
David McArthur [00:29:51] Yes, that's the date we're counting down to. Thank you for having me. I was happy to discuss transport data.
Laura Laker [00:30:00] You've been listening to the Active Travel Podcast. You can find us online on our Web site at http://blog.westminster.ac.uk/ata/podcast/ and you can follow us on Twitter and Instagram at Active Travel Academy @Active_ATA. Let us know what you think. Drop us a tweet or an email at firstname.lastname@example.org. Thanks for listening. Until next time.