Pegasus Simulator provides a framework built on NVIDIA Omniverse and Isaac Sim for simulating drone dynamics. It offers a simulation interface for PX4 and ArduPilot integration, along with a custom Python control layer. The project focuses on simplifying the simulation workflow and provides tools for research in unmanned aerial vehicles (UAVs).
Pegasus Simulator distinguishes itself through its straightforward integration with Isaac Sim, simplifying UAV dynamics simulation compared to building custom solutions. The project offers modular design, supports both PX4 and ArduPilot, and provides a flexible Python interface for control and experimentation. The documentation is comprehensive, supporting easy adoption for users with varying levels of experience.
- Multi-Platform Support: Supports PX4 and ArduPilot autopilot systems for realistic drone behavior.
- Python Control Interface: Offers a flexible Python API for custom control algorithms and experiments.
- Isaac Sim Integration: Seamlessly integrates with NVIDIA Isaac Sim for high-fidelity physics and sensor simulation.
- Extensible Architecture: Designed with a modular architecture for easy addition of new features and vehicle types.
- Simulation Environment: Provides a detailed and physically accurate simulation environment.
- Open Source: Available under the BSD-3 license, facilitating community contributions and customization.
- Research Focused: Built for research purposes with a focus on providing a robust and reliable simulation platform.
Pegasus Simulator is an active research project under development, with regular updates and bug fixes. Major releases are aligned with new versions of Isaac Sim, ensuring compatibility. The project has a helpful documentation and an active community contributing through GitHub discussions and issue tracking. The project's long-term maintenance is supported by the author’s Ph.D. work plan.
Pegasus Simulator benefits researchers and developers working on UAV control, navigation, and autonomous systems. It enables rapid prototyping, testing, and validation of algorithms without the need for physical hardware. The framework streamlines the simulation process, offering a valuable tool for advancing UAV research and applications.