While in the past, city planners designing for civilian protection guesstimated the movement of people by looking at the distribution of places (churches, shops, schools, parks) in the maps, today we can use mobility statistics, 3D models and road networks to quantify the diurnal movement of people through the urban landscape and the accessibility of shelters. With the help of computer simulations and agent-based modeling (ABM) our project aims to assess the evacuation potential of the Aarhus shelters in order to evaluate how effective, even, and complete their coverage was with regard to 1970-80s population and building distribution. For example, we will design an agent-based model that will investigate the efficacy of the shelter system under different conditions, such as varied shelter readiness or the level of awareness among the general public about the shelter location. Agents will be asked to move to the nearest shelter they are aware of, enabling us to reconstruct what proportion of the population successfully reached protection and whether they were evenly distributed between locations. This way we can assess the utility of the shelter system and take into account the differences in population numbers and activity in the course of the day and different levels of awareness of evacuation procedures. Computer algorithms coupled with historical data will reveal how well the local implementation of civil defense matched the democratic and equitable ideal of the national strategy.