ROSUnitySim: Development and experimentation of a real-time simulator for multi-unmanned aerial vehicle local planning

Abstract

In this paper, we present a novel real-time three-dimensional simulation system, ROSUnitySim, for local planning by miniature unmanned aerial vehicles (UAVs) in cluttered environments. Unlike commonly used simulation systems in robotic research—e.g., USARSim, Gazebo, etc.—in this work our development is based on a robot operation system (ROS) and with a different game engine, Unity3D. Compared with Unreal Engine, which is used in USARSim, Unity3D is much easier for entry level developers and has more users in the industry. On the other hand, as we know, ROS can provide a clear software structure and simultaneous operation between hardware devices for actual UAVs. By developing a data transmitting interface, a communication module and detailed environment and sensor modeling techniques, we have successfully glued ROS and Unity3D together for real-time UAV simulations. Another key point of our work is that we propose an efficient multi-UAV simulation structure and successfully simulate multiple UAVs, which is a challenging task, running 40Hz LIDAR (Light detection and ranging) sensing and communications in complex environments. The simulator structure is almost the same as real flight tests. Hence, by using the developed simulation system, we can easily verify develop flight control and navigation algorithms and save substantial effort in flight tests.