Last modified: 19 June 2017
Back in 2015, me and a quickly assembled team were one of three first-round winners of The Open Data Challenge Series . The competition, run by Nesta and The Open Data Institute, challenged teams to use open data to improve access to jobs. We pitched to use the £5000 prize to create a system for measuring the quality of bus routes, with the aim of improving them in the future so that they would connect more people with more jobs.
The idea was pretty simple. We combined open data on where everyone in England & Wales lives with data on where they work. We then combined that with data on where all the buses run.
A simple idea, but actually doing it was difficult – the UK standard for public transport timetable exchange is complicated and there was no good open source code to convert it into something useful. The prize money was what let us write that converter.
At the end of the process we completed lots of forms, created a hugely hopeful business model, and failed to win the £50k prize to continue development of our product. The competition organisers combined the even more hopeful business models that other winners wrote, got PWC to stick it in a blender, and wrote a press release about how much value the series delivered.
The last sentence may sound sarcastic, and it is a little bit, but I see the need to estimate return on investment and I can’t suggest a better way of doing it. So really, I think it’s fine.
But here I want to explain how the prize that we won over two years ago actually helped to create value, and how that value has been much greater than anything we could have made up in a business model.
First, we tidied up our TransXChange to GTFS converter, and released it on Github under an open license. There are now four converters , learning from each other and improving. The UK’s public transport is now available in a more useful file format reasonably easily. You may benefit from that on your smartphone app, or online journey planner.
Second, creating our tool let us develop a number of new techniques that have formed parts of much larger projects. In the past two years these projects have been used over 100,000 times, and generated revenue of over £50k. That revenue supports a full-time employee and although it is impossible to assign a fraction of this revenue back to the prize series, it was necessary for some of our work.
Third, we’ve continued to improve and update our final tool, imactivate transport. We have used it to push West Yorkshire Metro to release more bus usage data, and to help convince Transport for the North to hold their data strategy consultations in the open with ODILeeds. That happens tomorrow at TravelHack.
Fourth, we are now ready to do some really exciting stuff. We know where buses go, when they go, and who they could provide journeys for. And we’ve hugely accelerated the code that calculates how good a particular bus route is. We think that we can now explore and compare every possible bus route in a city, and automatically design a system that connects the most people to the most jobs for the least money.
Without Metro Mayors in place two years ago, and with the buses bill still a distant dream, there was almost no chance we would win the main prize two years. But since then a lot has changed and English cities are now free to use the tool that we've created. For people like me who believe that getting people to work is a key part of creating more and better jobs, this a an exciting time.