An Agent-Based Self-Organizing Traffic Environment for Urban Evacuations

Traffic congestion has many negative effects on our daily lives, the economy and the environment. These effects get worse in critical situations (e.g., emergency urban evacuations) when human lives are at risk. As smart cities are becoming a reality, technologies known as Intelligent Transportation Systems (ITS) have been considered as viable solutions for the many negative effects of congestion. Several ITS strategies for urban evacuation have been investigated. Among these strategies, researchers and practitioners have proposed the use of road reversal operations as a mean to efficiently utilize the traffic network capacity in critical times [4, 14, 11, 13, 12, 6, 7, 8]. Although road reversal operations have shown an improvement in traffic flow management, the proposed approaches are based on the execution of mathematical models to identify upfront, optimal road-reversal settings for the entire evacuation process. Even though these approaches have shown an improvement in traffic flow, to the best of our knowledge and according to a study conducted by Wang et al. [9], none of these strategies considers the time needed to implement a safe road reversal operation. In addition to road reversal strategies, several studies have investigated the use of zoning strategies to manage evacuation operations [5, 10]. A zoning strategy is based on dividing the traffic environment into small areas called zones. Traffic within each zone is directed towards safe destinations based on priorities or a phased evacuation plan. To this end, the emergency type is used to determine the optimal number of zones needed for the entire evacuation process as well as the size of each zone. However, these fixed parameters do not reflect the highly dynamic feature of traffic which might require a redefinition of the number of zones and their sizes.