| dc.description.abstract | With development taking the centre stage, life has become fast-paced journey that time is one of the most important factor. Transport sector has not been left behind on this sphere of development.
Transport technologies have been of great help to humanity, but have come with their own challenges. Traffic jam, congestion, accidents and pollution are the economic losses due to them. Looking at the challenge of road congestions, the management of the intersections traffic, contribute greatly to this phenomena. The intersections in Nairobi are operated through pre-timed electronic controlled traffic lights. These pre-timed traffic lights operating from one intersection, execute their duties at that particular intersection; however, their coordination to the corresponding traffic lights in an adjacent intersection is very inefficient, thereby leading to the deployment of traffic police officers to control the traffic at the intersections. This leads to wastage of manpower energy as well as treating a dynamic situation as static. To solve this unpredictable traffic congestion situation in Nairobi, this project details the use of multi-agent based framework to model the situation as dynamic in nature. We design and implement the framework, which serves to monitor the changing traffic condition and manages the traffic in accordance to density and wait time. This framework seeks to compare traffic density of lanes, compare wait times and release each lane accordingly. We use JADE development framework and Prometheus methodology in modeling the agent based solution as it offers an elaborate and stage-by-stage requirements for the system design and code guidelines to ensure system specifications are fully met. This framework presents aspects of observe-think-and-act model, thereby ensuring coordination of traffic during the whole process of movement. This system if fully implemented, seeks to benefit the road users that includes me and you, thereby increasing the uptake of agent technology in addressing complex and dynamic issues | en_US |
