METHODOLOGY

The study proposed a model to determine the optimal station and coverage for each response unit within the Traffic Operation Center-7’s coverage area to minimize the average travel time of the response units and the estimated total delay induced by incidents during morning rush hours (7:00 – 9:30) on weekdays. The optimization model was formulated based on various assumptions, including (i) response units will be posted at their designated stations and be dispatched after an incident is detected, (ii) they will return to their stations when the incident has been cleared, (iii) each response unit will manage incidents within its designated coverage area, and (iv) response units can travel on shoulders during periods of intense traffic congestion. The proposed model’s performance was compared with historical data from CHART incident database and those obtained by the “p-median” model, a traditional model that assigned optimal positions for available incident response units to minimize the average response time. The average travel time of the response units and the estimated total delay induced by incidents were compared to evaluate the performance of the strategies. 

 FINDINGS

• For a fleet size of three, the total delay was 4,659,967 vehicle-hours and 4,729,356 vehicle-hours for the proposed model and the traditional p-median model, respectively. This represented a 17 percent and 15.7 percent reduction in the total delay compared with what was practiced by the current CHART program.  
•  The average response time for a vehicle fleet size of three of the proposed model and the traditional p-median model was 7.51 minutes and 6.90 minutes, respectively, which is 3.6 percent and 11.4 percent shorter than that from the CHART program (7.79 minutes).