Route recommendation system to support multiple destinations and multiple routes to minimise road congestion

Multiple destinations routing is important for car navigation when a user requests for a service instead of a specific destination. Existing navigation systems can suggest k shortest routes to a destination, where the cost is defined in terms of distance, travel time, or other parameter. In case of a special event like a festival or an unforeseen situation such as tsunami, a large number of users would require the same service, e.g., a parking lot or a safe shelter, not a specific destination. When service points are known, the proposed algorithm first computes multiple near-optimal routes to those destinations, and then optimally distributes the traffic along those routes based on respective costs. This reduces congestion both on road network as well as at service points. Our proposed algorithm has been simulated with real-life traffic data on real city road network, showing encouraging results compared to conventional routing.