Social Navigation App For Cyclists

Over the course of several months, we created a mobile application that includes:

• a map with cycling lanes

• infrastructure points such as bike parking, repair shops, bicycle shops and rental locations, as well as places of interests for cyclists, water spots, etc., along with the ability to filter them on the map

• cycling lanes navigation

• The ability for users to add new points on the map

• User feedback

For the map itself, we chose Mapbox, an open-source service for custom-designed maps. It offered the best set of features for our solution when compared to other providers. The data is gathered from open data sources, such as OpenStreetMap and NASA, as well as purchased proprietary data sources such as DigitalGlobe.

Mapbox uses anonymised data from telemetry pings, such as Strava and RunKeeper, to identify missing data in OpenStreetMap with automatic methods. Fixes are applied manually, and reports on the issue submitted to OSM contributors. Additional features of Mapbox that made it the clear choice include a Studio tool that allows developers to import their own data to make choropleths, scaled point maps, and more This gave our engineers full design control: like adding our own fonts, icons, and textures to make your map fit seamlessly with your brand. But the most important feature is that it has its own cycling navigation, based on OSM data and the ability to work offline.

For the Android application we used Android SDK, Kotlin, AndroidX, Clean Architecture MVVM, RX Java, Koin and Retrofit HTTP for Back-end services. The address search function on the map is reproduced by Google Places. The native Mapbox solution for a map search appeared to have some issues in the target location; it doesn't recognize numerical street addresses, so you cannot enter a specific destination in a search. Solving this problem was mission-critical so the team decided to implement the map search feature using Google Places with Mapbox geocoding services for data reproduction.

The native iOS Application is created with a traditional UI Kit with Swift. All application styles are created inside the Mapbox Studio tool, which is easily customisable. The map search is implemented by Apple Maps, which is more efficient for Apple users.

As the full map and data coordination is implemented inside the Mapbox tool; the back-end services include an admin page to work with user feedback. It is written on Flutter and is hosted on FireBase Hosting. Cloud Fire Store is used for feedback production (a brand new tool from FireBase that can gp for Real-Time Database), FireBase Authorization, FireBase Links (for location sharing), and Firebase Crashlytics.