Friday 21st of October we participated in the Dynacity Data Challenge in Gent. In short our challenge was: It there potential for a bike sharing system in Gent, similar to those in Antwerp and Brussels? We think so and we will try to prove it using the origin-destination data from Gent’s traffic model.
To calculate this potential we used OpenStreetMap, as always, as a base for our routing network. On top of this we positioned some bicycle stations similar to how they are positioned in Antwerp and Brussels. Using an algorithm that combines pedestrian routing with a ‘borrow and return’ scheme for the bicycles we then calculated, for all origin-destination pairs the best route, in our case, routes that are faster by using the bike sharing system compared to using a car. For each resulting route we assign the origin-destination count to the route’s segments.
The result is then put on a map where each segment varies in width depending on the number of bikes predicted. The previous and next segments are also stored enabling us visualise where the bikes on this segment come from or go to, also called Selective Link Analysis. The resulting demo website can be found here:
We estimate about 3600 bike rides per day using this sytem (120 rentals per station per day), and that’s just counting rides that are currently done by car. We think we at least showed that it’s not unrealistic to have this system in a city the size of Gent.
Our challenge wasn’t the only one, make sure you check the Dynacity website. It was a great event and we at least hope the Dynacity experiment leads to something more.