An Aviation Manned-Unmanned Teaming Simulation in Urban Environments to Compare Autonomous Flight Formations

Abstract

In response to the growing importance of urban drone operations and the strategic potential of aviation manned-unmanned teaming (MUM-T), we propose a high-fidelity simulation platform for MUM-T operations in urban environments. The platform integrates Unreal Engine 5, Cesium for geospatial data, Colosseum (an AirSim fork), and PX4 SITL to provide realistic flight dynamics, sensor modeling, and immersive cityscapes. It supports the testing of MUM-T strategies in dense, obstacle-rich urban settings with high physical and visual fidelity. To demonstrate its capabilities, we simulated eight three-quadrotor formation patterns: Column, Line Abreast, Echelon, V-shape, Inverted-V, Triangle, Circular, and Arrow. Each formation consisted of one human-operated leader and two autonomous followers. The results revealed trade-offs between coverage, collision risk, and formation stability. Wider formations improved coverage but caused more collisions, while compact ones enhanced safety with reduced coverage. These findings highlight the platform's utility for evaluating urban MUM-T strategies under realistic conditions.