FastSim: A Modular and Plug-and-Play Simulator for Aerial Robots
Can Cui,Xiaobin Zhou,Miao Wang,Fei Gao,Chao Xu
DOI: https://doi.org/10.1109/lra.2024.3391059
IF: 5.2
2024-01-01
IEEE Robotics and Automation Letters
Abstract:State-of-the-art robotics simulators are equipped with well-established mapping, planning, and control systems. However, they lack modularity and the convenience of plug-and-play functionality. In this work, we present FastSim, a high-fidelity and user-friendly simulation framework based on the Unity engine. FastSim enables users to build robot simulation scenarios efficiently, by decoupling various simulation tasks with customizable modules, which contains simulated sensors, integrated utilities, visualization tools, and template robots. Besides high-performance robot dynamics simulation and high-quality image rendering, hardware-in-the-loop and mixed-reality applications are also available in this framework. The distinguished merits of FastSim include: (1) Robot Operating System (ROS) compatible control interfaces and abundant visualization tools for researchers who prefer ROS-based toolchains and (2) its integration of state-of-the-art planning algorithms, which enables users, even beginners, to quickly master the deployment of highly autonomous robots in simulations. Finally, we demonstrate the flexibility of FastSim by several experiments and performance evaluations with open source examples in repository: https://github.com/ZJU-FAST-Lab/FastSim .