Open-source Gazebo world collection adds 42 simulation environments for robotics testing
A new Apache 2.0-licensed repository provides 42 Gazebo Harmonic simulation worlds generated with Python and Blender for navigation, SLAM, and exploration research.
Summary
Developer Mert Güler released world-collection, an open-source repository of 42 simulation environments for Gazebo Harmonic. The worlds include mazes, offices, warehouses, forests, and moving obstacle scenarios, all generated via Python and Blender and exported to .sdf format. The project targets robotics researchers and developers working on navigation, exploration, LiDAR, and SLAM workflows who need reproducible test environments.
What's in the collection
- 42 2D plan-based environments: mazes, offices, warehouses, forests, real-world place approximations, moving obstacle worlds, and challenge-oriented layouts
- Generation pipeline uses Python plus Blender, then converts to .sdf world files for Gazebo Harmonic compatibility
- Output format allows integration with other simulation stacks, math calculations, and programmatic customization
- Current limitation: simple untextured designs; 3D and textured worlds planned for future agentic workflow, perception, 3D mapping, and object detection use cases
- License: Apache 2.0
- Repository: mertgulerx/world-collection on GitHub
Why it matters
Gazebo world creation has long been a friction point for robotics developers; a programmatic, open-source alternative lowers the barrier to reproducible simulation testing for academic and hobbyist projects alike.
Original reference: discourse.openrobotics.org →