OTN Unleashes Student Innovation at Birmingham Hackathon
On 24-25 October 2015 OTN participated as Silver Sponsor in BrumHack, a 24 hour student hackathon that was attended by more than 150 students from across the UK. The event was run as a hack in the Major League Hacking UK Fall season, and was held for the third time by the organisers, three Birmingham University Students.
Students had 24 hours to build innovative and catchy apps using sponsors’ data and APIs. OTN captured the imagination of two groups. One group from Aston University (pictured) decided to use Birmingham real time traffic data to build PrimeMover. Named after a spacecraft from an Ian Banks Culture novel, PrimeMover is a hack tackling the problem of reaching a destination in an optimal way while taking into account the requirement for parking. The team used Open Transport Network data and Google Direction services to plan routes to nearby parking garages and plotted the routes on a map. Access to the OTN data was made possible via an API created over the hackathon using the Slim Framework for PHP. It provided an easy way to access the information from Open Data via a web app and allowed students to translate and fill in the missing data required for the demonstration. An example of automatically finding missing data was the API using the NIA Parking garage names to get the geo-location points. These were set to zero in the dataset. The API translated the BNG values to longitudes and latitudes to ease the web app plotting on the map.
Aman Soni and Hassan Qulqass, two students that created an app, plotted the available routes first to nearby parking garages that were currently open and with available parking spaces, and then from the parking garages to the desired destination taking into account a change to map walking routes after the car has been parked. Students opted for an optimal route using an Ant Colony Optimisation algorithm in which the routes returned from the Direction Services were edges for the graph. This was done in order to keep the algorithm limited to the road network and so using valid travel paths. Additional local data, including available parking spaces, traffic flow and reported incidents were used to weight the individual edges and an optimal route found by taking both sets of the route navigation into consideration.
Aston team also attempted to create an interesting visualisation to demonstrate the workings of the algorithm for the hack, however due to the time constraints and the effort required for such visualisation there was simply insufficient time to investigate whether it would have met the requirements.
The second team, which included just one first year computer science student from the Birmingham University, produced an app called ResourceFrogger. Written entirely in Java, RF is a frogger type game in which the player attempts to cross a series of roads while avoiding cars. The creator used OTN flow data to populate roads which the player has to cross. This involved downloading the historical data XML files for the flow rate. Each of these files contain the name of the location, the date and time the data was recorded, and the data itself.
Michael Peran Truscott, the student that developed an app, assigned each road on screen with it's own XML file. From this, he extracted the name of the road and displayed it next to the road on screen. He then created two ArrayLists, one which contains the date, and one which contains the flow data associated with the date.
When the game starts the earliest date in the set is loaded, and each road has the number of cars on it as there actually were at that point in time. Every ten seconds, RF refreshes the raids with the data from the next five minutes and shows the new number of cars. If the data is missing for that time slot, the road shows the data from the last valid time.
All three students that worked on these amazing apps received a prize, a RPi 2 kit, and an invitation to collaborate closely on similar projects in the future.