Getting a taxi in highly congested areas (e.g. airports, train stations) is both time consuming and expensive. While the emergence of novel Transportation Network Companies (e.g. Uber) has helped increase the supply of drivers during peak times, they have done little to reduce congestion in ‘hot spots’ such as airports, major stations, and stadiums. In these places, a steady stream of passengers arrive via some public transport mode, say train, and then depart to different destinations. The large volume and continuous flow of arriving passengers create a golden opportunity for ridesharing among departing passengers. In a ride sharing mechanism, the passenger submit a ride request shortly before departure specifying some parameters such as pickup time, passenger destination, willingness to walk to the destination, the maximum delay tolerated due to ride-sharing, maximum time the passenger is willing to walk and number of travelers in the party. In general, the goal is to allocate passengers with shortest and cost efficient ride.

This project evaluates ride-sharing algorithms on spatio-temporal data. The data in this project is an OSM data file of the road/highway of Mumbai city in India.