Simulation of urban service operations for locational decisions

Abstract

Locational issues arise in the provision of urban services; for in stance, finding an optimal home location for a service facility in an urban setting is difficult when the interarrival times of ser vice calls and times are governed by nondeterministic processes. In the steady state, there exists an analytical solution to the prob lem of finding a location minimizing the expected response time.² This solution is obtained under the assumptions that (1) the service calls are generated by a homogeneous Poisson process; (2) the server always returns to the facility; (3) service requests are handled on a first-in-first-out basis; and (4) the travel speed is constant. Our experiments with a simulation model where the above assumptions have been relaxed indicate that the results are significantly affected by these assumptions and that the steady state may take too long to reach.