A Multi-Modal Agent Based Mobile Route Advisory System for Public Transport Network

In a metropolis such as Hong Kong, the transport network is massive, dynamic, and complicated, and therefore route finding is not an easy task, especially when routes comprising several modes of transport vehicles (such as bus, train, and ferry). This problem is even more important for e-tourism tourists and moving workforces, who may need to visit an unfamiliar part of the metropolis. To find a route that is the most cost-effective is even harder and time-consuming. In this paper, we present a conceptual model for a multi-modal public transport network and a multi-agent information system (MAIS) implementation architecture. We propose several intelligent approaches in the search agents to enhance the performance of route finding, in particular a knowledge-basket approach. Backend information agents gather updates periodically from transportation companies. We also consider hired transportation services, such as taxis and vans, which might offer similar costs for group commuters. We further support an agent-based auction sub-system for reservation of these mobile vehicles. We have built a route advisory system prototype with such features. Our system supports a flexible multiple user interface views on different mobile platforms. Our paper also includes some benchmark results to demonstrate the effectiveness our approach.