How open data helped TravelTime go from idea to SME
Charlie Davies, Co-founder of location-based search software provider iGeolise, explains how he used open data to build a business and measure distance in minutes, not miles
As with most businesses, the location-based search software provider iGeolise began because a problem was noticed that needed a solution.
I’d realised that when searching for local amenities like hotels, restaurants and businesses, the search results listed weren’t always relevant to me because they seemed impossible to get to. The lightbulb moment was seeing that these searching sites use miles radius to find the closest results, but when was the last time anyone said “my hotel’s only 2 miles radius from the beach” or “I’m running late, be there in 3 miles radius”?
We discuss distance in minutes because we can’t go as the crow flies. Search engines and apps can calculate time from point A to B, but what if you don’t know where you’re going?
My idea for the TravelTime platform was an API that could return ranked and sorted location results, based on minutes rather than miles. Searchers can specify their location, time of day and mode of transport to get more reliable results that are ‘within X minutes’ rather than ‘within X miles radius’.
To bring the idea to life, my business partner Peter Lilley and I needed developers to make sense of a wide array of location and transport data. The excellent thing about open data is that we didn’t have to invest money upfront for this. This was essential at our stage of development because we still needed investment to get going. If we’d had to pay for the data, the barriers to create our innovation would’ve been too high.
With this considerable reduction in risk, the focus could be on making sense of a wide range of open data. We used TNDS Traveline National Data Set data for bus and rail schedule times in Great Britain, NaPTAN data to pinpoint the location for every bus stop, railway station, ferry terminal and other transport stops, NOS (National Operator Codes) data to identify the operator of each public transport route, Ordnance Survey data for detailed street level maps and more.
Our biggest challenge with the data was making sure that all human input error was ironed out. This took time, and sheer persistence. You get the most out of open data when you acknowledge that there’s going to be rogue entries in all types of data and stamp them out. As an example, we realised that some routes started at 11am and finished at 11pm when the journey was only a mile! To resolve this we made sure we manually re-entered data points that were incorrect, as it would skew our software’s results.
Without open data an API like ours wouldn’t have been possible – or only possible for businesses with big budgets. I’d recommend using open data to anyone who wants to fuel new projects at a low cost. My words of caution would be to recognise that the data will probably need many man-hours to ensure accuracy.
I’m a big believer that open data can be used for almost any industry. So whether you’re dedicated to preventing crime, or convinced you can predict the weather better than the BBC, there’s probably some data you can use along the way.
Because open data made the TravelTime platform possible we’re now supplying household names with the API including Zoopla, OpenTable and VisitBritain. To find out more about the TravelTime platform and searching by minutes click here.