Abstract-In this paper, we propose a human mobility sensing system simulator called HumanS. HumanS is a multi-agent simulator that works with geographic information system (GIS). It models realistic movement of pedestrians and behavior of sensors that capture the mobility. Developers can evaluate performance of such sensing systems that detect location of people for pedestrian navigation, purchasing behavior analysis and so on under realistic model of environment. The simulator is designed based on GIS databases that store and manage the scanning data by appropriately tagging their time and location. The capability of the HumanS simulator has been validated by a scenario where accuracy of a pedestrian flow sensing system is tested with different settings of sensor types and location in Osaka downtown.