ODI Leeds

#TravelHack part 2 - creating a data strategy for a new transport system

As a final push before lunch, Paul helped everyone join the working groups for the afternoon ahead. Having identified the core themes from the lightning talks, and added those that the audience agreed were priorities, these were put on to post-its on the wall. The attendees were then invited to put their name to the theme they wanted to explore in more depth, which ranged from legislation to customer choice and journey planning to new datasets.

With everyone happy with their working groups, it was time to take a break and refuel with "The ODI Leeds Summer Picnic" for the afternoon ahead.

Suitably fed and watered, the next few hours were devoted to exploring the themes, discussing data, and putting pen to paper. With diverse working groups, people could expect to have their ideas challenged and their scope broadened. In such a collaborative atmosphere, even organisations that have previously been at odds with each other can let their barriers down and work together on shared benefits. It is something we always try to do with innovation events here at ODI Leeds - by providing an outside space, removed from familiar faces and offices, there's an opportunity to concentrate on something without worrying about an overseer or furthering agendas.

You get together, you work together, you make awesome stuff.

At the end of the afternoon, it was time for the round up. Six groups had to give a 5-10 min brief summary of what they had been working on (with full details available in the collaborative document).

The first to speak was the 'data infrastructure' group, lead by Jamie Fawcett from ODI HQ and Carl Rodrigues from OpenDataSoft. They started with a very strong and valid comment - local authorities and other organisations in the transport industry would benefit from acquiring more data skills, right from basic abilities up to data science and developing tools. In terms of procurement opportunities, perhaps a reusable set of standards could be used to ensure suppliers all commit to publishing their data openly and to a format that is shareable and usable across the network. With this would come the idea of a 'minimum viable quality' to make sure the data is of a certain quality. Maybe use incentives to reward those who regularly meet these requirements? And finally, a comment from Data Mill North about a 'dataset of datasets' - one central catalogue of all of the different datasets, who publishes them, how often, etc, that could be managed by TfN.

Next to speak was Neil McClure, leading the group focusing on 'new data.' In particular, they were looking at the data generated by a potential contactless payment system. Working just like Oyster (as a system as a well as a card), the number of 'taps' would be consolidated, an appropriate fare calculated, and then the fare being charged to the customer. If this could span multiple modes of transport, that would be a lot of data! Some of it would have to be managed carefully, such as the personal and financial details of the customer. But other data could provide never-before-seen insights, such as how people are using/navigating urban transport. This would be very valuable to Local Authorities and planners, not to mention in modelling. Data about the volume of 'taps' at stations or bus stops also means that disruptions or events can be planned around these principal points.

Speaking of disruptions, that is precisely what the next group focused on. Guided by Paul Everson, they delved into issues such as the source of disruption data and how to get that data to the customer. They defined disruption as anything like delays, cancellations, or diversions, which as we all know can be caused by accidents, roadworks, events, etc. Paul commented that a lot of Local Authorities actually do produce human-readable documents about the impact of disruption but this never reaches the journey planners or other transport services. He called for tools to help bring all of that data together and perhaps publish a weekly digest of impact data to attach to the various systems across the transport network. He also echoed a sentiment we share here at ODI Leeds - sometimes it is better to just get the data out there and then let data users work out what's useful and what's not.

The next group to talk was the 'user choice' group, where the customer is put first (they also argued that much of what they had discussed could also apply to freight). They recognised that not every journey is entirely private or entirely public. For example, someone could drive to a Park&Ride and then use the bus for the rest of the journey. Users must be presented with a journey that best suits their needs - there's no point recommending a bus if the bus stop is too far to walk and still make it on time for the next arrival. There are also a whole host of other factors that might influence the user's choice of transport, factors that might not seem obvious to transport providers. Things like will there be space on the bus/train? What is the weather going to do? How busy are the routes? They also suggested that maybe there some physical changes (that would be less expensive that large infrastructure projects) that could help influence user behaviour - new Park&Rides, improved interchanges, better marketing of multi-mode transport, investment in data collection such as live car park spaces and train seats, etc. They believed a 'soft approach' could have huge benefits in the long term.

The fifth group to talk focused on how to get and keep customers engaged, including non-digital users. This meant taking a look at communication methods - for those people who don't own a smartphone, how can they interact with and use a smart travel system? Would this use more expressive forms of display at bus stops for instance? i.e. an interactive map with arrival times or moving icons And how would we get vital feedback from customers? With public input there is always the risk that you can never please everyone and that a project might stall due to the amount of back-and-forth. There is a case to be made for making something quickly, testing it, fixing it, testing it again, fixing it again, etc. Whilst the public are data-providers and consumers - they shouldn't necessarily have the final say. Perhaps data release could be gamified and points given which can be cashed in on support for a particular new development or proposal from TfN

The final group covered 'competition' and what that might look like in this new transport landscape. To enable a multi-modal system across all of the North, many operators and orgs would have to collaborate. If TfN has no power or plans to introduce fare changes or a blanket price across the whole system, then operators will have to compete on something else. This could be comfort or additional benefits, like free wifi. The group were keen to state several things to be cautious of in regards to competition - TfN would have to act to encourage best practice between players in the marketplace and improve standards across the board; and any changes would need to be well thought out and resourced before committed. For example, a new IT system that crashes could cause reputational damage to an operator. It was important that this group did have such a realistic and grounded focus (they kept saying 'they weren't all doom and gloom') as it can be all too easy to overlook this kind of thing. Better to find these potential bumps now rather than too late to fix them.

To conclude the day, it was back to Richard Mason of TfN, who started the day with a wow and was going to end with a wow. He was delighted with the turnout and the quality of the outputs. For his first 'hackathon' style event, he wasn't sure what to expect. Now he has pages of notes, sketches, photos, etc, from enthusiastic professionals and experts in transport or data that could all help TfN build a solid data strategy. He reiterated that TfN are keen to keep working with the open data community and they want to see this project through to completion. He doesn't want to lose the momentum - there will be more TfN events that engage developers, planners, customers, and more.

This project does not come to a stop just because the event is over. Behind the scenes, we and TfN will continue to push the right people towards this ambitious idea. We will encourage people to release their data openly so that we start to see the transport system as a whole. If you want to join us or help us in our endeavours, it's not too late to be involved.

  • Share your blog posts or tweets using the hashtag #TravelHack.
  • Sign up to our mailing list to receive news about any future events.
  • Talk to your transport operators, engage them and ask them to open their data.
  • If you have been creating your own data and would like to see it used, get it published and point us to it.
  • Let us know if someone is doing good things with transport data by tweeting us at @ODILeeds