Title Tagline Description (should be approximately 500 words); Explanation regarding which TM Forum APIs have been used Explanation and links regarding other platform capabilities used, example: IBM Bluemix, Ericsson Connected Car API, Nice Open Data etc Feedback regarding use of the APIs and platform from the event
Everyday that we go to work we see a lot of people running on one of the most polluted streets of Barcelona. So we thought it would be great if we could help those athletes to make a healthier sport. And so we created EcoRun
What it does
EcoRun is an application that allows athletes to take routes of the length they want or to the point the want that avoids polluted streets. Then, athletes that complete these healthy routes earn points that can be then exchanged for bus tickets, bikes services, and more services provided by the City Governments. The healthier the route, the more the points they earn.
How we built it
We've used a really cool stack of technologies in order to build our product. On the one side, our main server is powered by Go language and TMF Forum APIs. This server is responsible for doing all the communication, and serving everything needed in the store to buy and use products. We are using the TMF Catalog API to store our products, the TMF Customer API to manage all of our customers, their points and other information, and the TMF Inventory API to store what our clients have bought, with their QR code to use it later. This Go server is also connected with FI-WARE to retrieve information from mobile sensors in the city of Santander. We used this city for the demo because of the great quantity of Open Data they provide.
On the other side, we built a Java server responsible for calculating routes that avoid pollution. To do this, we used an OpenSource library called GraphHopper. This library basically converts an OpenStreetMaps file into a graph. We then modified this graph and added weight to its edges depending on the quantity of pollution in that point. Finally, we run a path finding algorithm in this graph to find the best route.
Challenges we ran into
We ran into several challenges while developing and deploying the project. Communication between Go, Java, TMF Forum APIs, real-time FI-WARE sensors wasn't easy. Another problem we had is that recalculating the weights for the graph was taking too long, so apart from optimising the process we moved this task to the background while still serving new route requests without disturbing the service.
Accomplishments that we're proud of
Most of all we are proud of having made a product that could already be used in a real city, in this case Santander. We are also very proud of the fact that we used Open APIs, so installing this application in another city just takes a breeze (if Open Data is available, of course).
What's next for EcoRun
We want to make this app bigger and available in more cities. We will improve the routing system to suggest even better routes. We will analyze where citizens walk and recommend them healthier routes, even if they aren't doing sport and they are just going to work. We will track running and biking stats such as kcal, distance travelled, etc. And we will sure have many more ideas in the future!
Everything in this project has been open-sourced and written during the hackathon. Have a look! https://github.com/novikk/openHack-routing-server https://github.com/novikk/openHack-api-server https://github.com/novikk/openHack-app