Multi Agent Model Based Traffic Flow Control system-Section of Nairobi, Kenya

Printer-friendly version

Urban traffic flow control and management is a key concern for any country.  In many governments, the police departments are usually tasked with ensuring that there is not only smooth flow of automobile traffic for efficient and optimal economic and social activity, but also ensure the security of the countries’ people as well as minimisation of transportation costs.  The case of traffic management and the deployment of police to be responsible for the same finds presence in the Central Business District of Nairobi; the capital city of Kenya.  In the recently (2006) public website of the Kenya Police, they state that one of the obligation is to ensure free flow of traffic.

The development of multi agent concepts in Artificial Intelligence in recent years provides an opportunity to explore the possibility of deployment of such concepts to further enhance the functional operation of traffic lights in the problem of traffic flow control and management. A review of available literature deduces that much of the successful works so far have been to model in theory the urban traffic lights infrastructure but successful software based works and physical realisations are yet seen or published.

This research project sought to model three roundabouts traffic flow management problem in Nairobi Kenya’s Central Business District around multi agency concept. A case scenario of three traffic lights interconnection nodes; namely Railways roundabout, Harambee Plaza round about and Nyayo house round were selected. The criterion for the selection of the three nodes was the premise that the three junctions were the only ones which are directly interconnected with one another and a appeared to have the traditional traffic lights systems work. The second premise was that under the normal human “eye”, the three junctions appeared to be the ones that handle the bulk of vehicle volumes in all directly adjacent interconnecting segments.

It is observed and concluded that the use of multi-agent theory in the control of traffic flow in Nairobi is feasible and deployable.

UoN Website | UoN Repository | ICTC Website

Copyright © 2018. ICT WebTeam, University of Nairobi