A Map of Time

Several months ago, I read about a project at Media Lab Europe that showed a lot of promise (I was shocked as anyone else) – Amble Time. Amble Time basically factors time into geographical maps, telling you where you could walk in a certain amount of time.

By using a GPS system and your average walking speed, it creates a bubble that indicates everywhere you could walk in an hour. Alternatively, given a final destination, it can show where you could roam along the way and still arrive on time. In the second situation, as your position changes and time ticks by, the bubble slowly shrinks and morphs until eventually it highlights the shortest path to your destination.

This is all very cool, but the question is, how did they create their temporal data? Are they just using the walking distance along streets, or are they taking into account the fact that some streets might be busier than others? And surely it’s a shame that Amble Time doesn’t work for driving?

Anyway, I thought up a possible solution this morning that ties into a similar project called Amsterdam Real Time, where a group of people carried GPS units around Amsterdam for two months and created ‘trails’ around the city which were eventually transformed into pretty maps.

The solution would involve giving lots of people GPS units that track their positions. By matching up their lat/long with actual streets and roads and measuring how long it took them to walk/drive/cycle along them, you could create a fantastically detailed idea of how long it would take to get from any point to any other point at any time of the day. So instead of pretending that all roads are equally traversible and distance is the only thing that matters, you could take into account narrow roads, wide roads, rush hour, congested spots – everything.

I don’t really know how you would go about coding this kind of system – it’d be non-trivial, but then it wouldn’t be impossible either. If you wanted to be really ambitious, each users’ map could ‘learn’ their average walking speed at different times of the day and week, and tweak the raw data accordingly. Even more ambitious would be to track people in real time to create up to the minute maps of how long it would take to get from A to B (or where you could walk or drive in x minutes).