This thesis investigates how surveillance fragments spatial perception and memory through agent-based computational modeling. Using women’s navigation under surveillance in Tehran University’s Faculty of Fine Arts as a case study, this research examines how systematic threats reshape the cognitive maps of individuals in controlled environments. The methodology integrates spatial cognition theory (Tolman, Mallot, Ingold) with agent-based simulation to model navigation under stress and uncertainty. The simulation operationalizes three interacting systems: constrained perception (heading-dependent intake zones, occlusion), dynamic memory (dual-layer storage distinguishing spatial detail from learned threat, with differential decay rates), and stress-driven learning through which agents psychologically erase physically accessible paths.