To facilitate the currently expanding suite of driver assistance and safety systems, modern vehicles are equipped with a multitude of sensors to perceive their environment. To utilize the obtained raw sensor data, complex models are used. For multiple applications, a model of the position of vehicles surrounding an observer, as well as their predicted trajectories, is needed. Floating Car Observers (FCOs) are vehicles that are capable of identifying and mapping vehicles in their vicinity. In this paper, we introduce a framework to simulate FCO capabilities on top of an existing microscopic traffic simulation. Our framework is capable of simulating FCO embedded as part of ordinary traffic streams and calculating which vehicles an FCO would detect, taking into account, among other things occluded vehicles, sensor range, or stochastic detection errors. To demonstrate the capabilities of our framework we simulate traffic along a 1300 m long stretch of urban street including two intersections. We show how many FCOs are needed to detect certain percentages of all vehicles.