Geolocation management system
Insights from a real time tracking of tourist buses
This project represents a system for supervising and gaining insights from geospatial data specific to a fleet of smart devices. The devices were phisically mounted on-board of charter traveling buses and they allowed a constant connection between the passangers, the bus and the cloud systems. This geolocation project is part of a much larger system for passanger entertainment and it represents a major component of it by giving realtime visual information for bus managers and also realtime features to the passangers
- Build a system that offers realtime tracking of buses
One of the big challenge was to predict the upcoming WIFIsignal strength area and the distance till the bus would reach that area. The problem was that we did not know where the bus is going to move. The solution we found for that was to construct a network of geofences generated from all buses locations. And for the current bus we would compare his current position with the nearest geofence every few seconds.
We would use this comparison to conclude if the bus is closing or not on that geofence and if it was, we could know with a certain probability that that's the next geofence in which the bus will enter i.e. the next WIFI signal area.
The central piece of the geolocation system is in the data layer. For data persistence we used Elasticsearch along with Filebeat and Logstash as the pipeline for storing the data.
Nodejs servers were used for the processing part and Leaflet on the front end in order to display the heatmap. The geolocation system was also used in React Native passenger facing app in order to inform the user of an eminent signal change.