Skip to content

RflySim Full-Stack Unmanned System Toolchain

A professional closed-loop development toolchain for research and education. Natively supports multiple heterogeneous vehicle systems, reshaping cutting-edge Sim-to-Real experimental scenarios through an industrial-grade graphics rendering engine.
Algorithm Prototyping ➔ Software-in-the-Loop Simulation ➔ Automatic Code Generation ➔ Hardware-in-the-Loop Simulation ➔ Real-World Deployment and Testing
🚀 Quick Start ▶️ Featured Cases ⭐ Download Installer

Important Notice

  • 📚 The official RflySim companion textbook, Intelligent Unmanned System Control Practice, has been officially released! Instructor teaching resources and experiments are included with the book. For details, see: 🔗🔗🔗
  • 📢 RflySim v5.01 official version has been released! Check the Changelog for the latest features.
  • 💼 Our team is continuously recruiting graduate students (Master's/PhD) and engineers/interns in unmanned systems (Beijing/Changsha). Send your resume to: service@rflysim.com

✨ Why Choose RflySim?

🎯 1. Consistency: Integrated Framework Design
Builds a standardized closed-loop development framework for “modeling–development–testing–evaluation,” comprehensively supporting multiple vehicle types, including multicopters, fixed-wing aircraft, VTOLs, unmanned ground vehicles, and surface vessels. The platform starts with Simulink as its control core and, together with core automatically generated C/C++ modules, seamlessly connects a cross-platform integrated design workflow.
Multiple Vehicle Types
⚡ 2. High Efficiency: Fully Automated Validation and Rapid Iteration
Built to accelerate prototype deployment: after completing logic design in the integrated environment, generate code with one click and automatically deploy it for software/hardware-in-the-loop simulation. Once verified, not only can real sensor parameter sources be switched in instantly, but the system modules “as is” can also run directly and seamlessly on actual vehicles and onboard processors, greatly reducing validation costs.
High-Efficiency Framework
🌌 3. High Fidelity: For Intelligent Perception and Scene Reconstruction
Based on the continuously evolving UE4 / UE5 engine, the platform delivers unparalleled physical-engine immersion and material-rendering lighting capabilities. It not only supports high-precision simulation of sensors such as RGB cameras, depth maps, point clouds, and infrared cameras, but also enables complex adversarial testing under lighting changes, weather transitions, and environmental occlusion.
High-Fidelity Rendering
🔄 4. Transferability: Fully Connecting Sim-to-Real
Bridging the boundary between virtuality and reality. In the virtual environment, rendered images can be directly transmitted to replace real-vehicle camera inputs, while computed actions are delivered through the network to the underlying PX4 flight controller for closed-loop simulation. Vision modules refined within this framework require no cumbersome migration procedures; with one-click switching to real camera data, they can be directly deployed for real-world flight.
Sim2Real Migration
🌐 5. Large-Scale Swarms: Native Collaboration and Distributed Support
Its inherently highly flexible and loosely coupled network topology supports distributing a single simulation task among multiple multi-core hosts, numerous PX4 physical-level modules, and independent distributed rendering processes for parallel computation. Its unique pure-software and virtual-real hybrid simulation modes make adversarial game research involving hundreds or thousands of swarm nodes especially easy and low-latency.
Distributed Topology


  • 💡 6. Ease of Use: A Lightweight Experience for Research and Education

    Supports efficient and stable operation on Windows with an embedded WSL / Linux toolchain. One-click installation and uninstallation avoids the complexities of ROS dependency assembly. The platform includes hundreds of ready-to-use open-source examples and provides deep underlying support for domestic operating systems such as Kylin, making it the preferred choice for all types of R&D and courses.

  • 🧩 7. Functional Completeness: Full-Stack Unmanned System Coverage

    Comprehensively integrates all component environments “from bottom to top,” covering both fundamental model evaluation and a complex stack including “advanced perception, self-inspection and diagnosis, route planning, execution decision-making, and even networking and swarming.” It can also be flexibly extended to implement heterogeneous compatibility solutions and dynamic system health injection, practically filling every missing link across the intelligent system chain.

🧭 Quick Navigation Guide

To help developers from different backgrounds quickly find the resources they need, we provide a clear knowledge architecture in the left sidebar. We recommend exploring along the following paths:

Your Role Recommended Learning Path Core Goal
🎓 Students / Beginners Quick Start → Core Courses Follow the 10-chapter supporting curriculum system, from theoretical foundations to hands-on implementation
💻 Algorithm Developers Python SDK / Simulink Directly consult APIs and interface specifications for secondary development independent of the underlying implementation
⚙️ System Integrators Software Manual → Hardware Ecosystem Understand the complete chain from networking and sensor configuration to hardware selection
🚀 Frontier Explorers Featured Cases → Advanced Topics Starting from published top-tier journal paper cases, explore cross-disciplinary applications such as LLMs

📚 Companion Tutorials

The RflySim platform is accompanied by the Design and Control of Multirotor Unmanned Aerial Vehicles textbook series, authored by Prof. Quan Quan's team. It systematically explains the comprehensive knowledge framework for UAVs, from theoretical foundations to practical applications. For more details and resources, please refer to Companion Textbooks Overview.

Textbook Title Description Target Audience
Design and Control of Multirotor Unmanned Aerial Vehicles A systematic foundational textbook on multirotor design, modeling, perception, control, and decision-making Senior undergraduates, graduate students
Practice of Design and Control of Multirotor Unmanned Aerial Vehicles Eight progressive experiments based on the RflySim platform and an MBD development workflow Students taking laboratory courses
From Principles to Practice of Multirotor Unmanned Aerial Vehicles An introductory textbook and complete operational guide from basic concepts to assembly and tuning Beginners, hobbyists
Remote Control Practice of Multirotor Unmanned Aerial Vehicles A comprehensive remote-control practice textbook for vocational students and lower-level undergraduates Vocational or lower-level undergraduate students
Flight Control Practice of Mini Fixed-Wing Unmanned Aerial Vehicles Full-stack fixed-wing UAV development, covering design through vision algorithms Fixed-wing enthusiasts, researchers
Intelligent Unmanned System Control Practice Official companion textbook for RflySim RflySim beginners, developers

📊 Toolchain Version Tiers

The RflySim toolchain not only provides a free basic version for general developers, but also offers advanced solutions for more complex, high-quality enterprise or defense-industry testing requirements. It is currently divided into three major version tiers:

Version Name Main Features & Included Content Applicable Scenarios & Target Users Acquisition Method
Free Edition Includes single-machine SITL/HITL simulation core, Python/ROS development interfaces, supporting tutorials and courses, and other foundational simulation and validation capabilities. University teaching, individual learning, and entry-level algorithm development and validation Download the ISO installation package directly from the official website.
Full Edition Includes multi-machine distributed and cluster simulation, UE5 high-fidelity rendering support, complex environment adaptation, the latest PX4 firmware, and hardware integration with cluster boxes. Research institutions and validation of collaborative and swarm unmanned-system projects Released selectively; please contact the official after-sales team.
Custom Edition Interface customization for specialized projects, third-party autopilot/avionics black-box integration, high-level safety testing (such as FPGA simulation), and specialized training services. Enterprise R&D, defense applications, and high-level safety-standard testing and certification Sign a requirements contract for in-depth custom development.

Version Comparison Details

For more detailed information about feature coverage and technical support differences, please directly visit 🔎Version Comparison Details🔎 . For the Full Edition and above, please contact the Feishi Lab team directly via email (service@rflysim.com) to obtain them.

Core Paper: RflySim Principles and Architecture

This paper is the core publication introducing the fundamental principles, design philosophy, and overall architecture of the RflySim toolchain. It systematically explains model-based design methods and how modeling and simulation, control, and swarm decision-making work together to support the development workflow from simulation validation to real-world deployment. Readers wishing to gain a deeper understanding of how RflySim works may access the full paper through the links below.

Links: 🔗 10.23919/JSEE.2025.000079 📥 View PDF 📚 BibTeX Citation

About This Documentation

This site is designed to provide you with the clearest and most intuitive RflySim user guide and developer reference. The left navigation bar serves as the knowledge graph—feel free to browse and explore at your convenience.
If you encounter any issues during use or have suggestions for improving this documentation, please submit feedback to us via GitHub Issues.

Copyright © Feishi Lab · Beijing Zhuoyi Intelligent Technology Co., Ltd.