Modeling Urban Warfare: Joint Semi-Automated Forces in Urban Resolve

摘要

The United States military is performing operations in urban environments with increasing frequency. Current Department of Defense doctrine is poorly suited to train and equip today's warriors with the tools and experience necessary to fight and win in modern sprawling cities. In order to "close the gap," the U.S. Joint Forces Command's Joint Experimentation Directorate led an effort to run a massively distributed simulation of a synthetic urban environment utilizing human-in-the-loop operators called URBAN RESOLVE. The synthetic environment simulated the city of Jakarta with over 1,000,000 buildings and structures and over 120,000 civilian entities. A Red force retreated into the city while a Blue force attempted to determine the enemy's Order of Battle. The exercise generated over 3.7 terabytes of data in seven distinct trials. This research evaluated the time required to identify targets after detection and the accumulation of identifications over time, and searched for trends between the seven design trials and between target groups. Two trends emerged from this research. First, there was a notable difference in the time required to identify a target once it has been detected based on its target group. Second, two design trials that are expected to demonstrate show counter-intuitive results.