Simulation environments for
drones
UAV autonomy training needs cluttered, realistic environments, buildings, trees, wires, delivery targets, dynamic obstacles. Rigyd generates physics-enabled OpenUSD assets compatible with Isaac Sim, AirSim, and Gazebo.
The problem
Why existing workflows fall short.
Aerial environments are visually complex
Drones operate in dense, varied scenes: urban canyons, warehouse interiors, forests, mountainsides. Each environment type needs thousands of unique, varied 3D assets.
Lightweight physics still needs accuracy
UAV dynamics are sensitive to aerodynamic interactions with obstacles. Mass distribution and collision-mesh detail affect propeller wash modeling and rotor-wake-aware path planning.
Perception stacks need semantic variety
Vision-based drone autonomy trains on diverse object classes. Lack of semantic and geometric diversity caps policy robustness the moment a drone enters a new environment.
How Rigyd helps
AI-native infrastructure that automates the hard parts.
Dense environment catalogs
Convert entire 3D building, infrastructure, and landscape catalogs into SimReady assets, useful for urban delivery, infrastructure inspection, and search-and-rescue training.
Wire- and edge-aware collision geometry
Convex decomposition captures wing-catching edges, antenna-snagging protrusions, and wire-like thin geometry, critical for realistic collision-avoidance policy training.
Semantic labels for perception
Each asset ships with semantic labels (building, vegetation, vehicle, person), enabling perception-in-the-loop UAV policy training without a separate annotation pass.
cost reduction vs hand-modeled drone environments
better real-world performance with accurate physics
per asset, from upload to simulation-ready
Build drone training environments at scale
Upload 3D obstacle and environment models and get SimReady assets for UAV simulation.
Starts at $29/month. 30 credits included.
Frequently asked questions
Is Rigyd compatible with Isaac Sim, AirSim, and Gazebo?
Yes. OpenUSD output imports into all three. Physics properties (mass, collision geometry, friction) are preserved, so the same asset runs in every major UAV simulation stack without re-export.
Can Rigyd capture wire-thin obstacles for collision-avoidance training?
Yes. Convex decomposition captures edges and thin protrusions that simple bounding boxes miss, critical for realistic wire, antenna, and branch collision scenarios in UAV policy training.
Are assets suitable for dense urban environment simulation?
Yes. Bulk conversion makes urban, warehouse, and landscape catalogs practical. Each asset ships with semantic labels for vision-based drone autonomy and perception-in-the-loop training.
Does Rigyd support AirSim and Microsoft Project AirSim?
Yes. AirSim consumes Unreal Engine assets directly, and Microsoft Project AirSim is USD-native. Rigyd outputs UE-compatible FBX with embedded physics metadata for legacy AirSim workflows, and direct OpenUSD for Project AirSim. Semantic labels, collision meshes, and physics properties survive both pipelines, so drone policies can be trained on the same asset library across simulator versions.
Can Rigyd capture thin geometry like power lines, antennas, and chain-link fences?
Yes, these are critical for drone collision avoidance training. Convex decomposition handles cylindrical thin geometry (wires, antennas) with elongated hulls that preserve wing-catching edges. For chain-link or grid structures, Rigyd offers a "thin obstacle" mode that uses capsule chains instead of decomposition, ensuring the simulator can detect contact along the entire wire rather than just at bounding-box boundaries.
Related reading
In-depth guides on robotics simulation, OpenUSD, and SimReady assets.
21 MAY 2026
Synthetic data generation for computer vision in robotics
Real-world labeled robotics data is expensive and slow to capture. Synthetic data is fast and unlimited, but only useful if the underlying simulation has correct physics, semantic labels, and domain randomization. Here's the complete pipeline.
18 MAY 2026
Digital twin creation pipeline for manufacturing
A factory digital twin needs every object to behave physically, not just render. This is the end-to-end pipeline: CAD intake, BIM merge, physics layer, semantic labeling, simulation runtime, at the asset volumes (10K+ unique SKUs) real factories actually contain.
13 MAY 2026
How to set up mass, friction, and joint properties for robot training
The three pillars of robot physics setup, mass, friction, joints, determine whether your trained policy transfers to real hardware. Here's the calibration target for each, the schemas, and the common mistakes that quietly break training.
Explore more
SimReady for Isaac Sim
OpenUSD assets validated for NVIDIA Isaac Sim
Models for MuJoCo
Physically accurate assets for robot learning
Assets for Gazebo
Simulation-ready models for ROS 2
Warehouse Simulation
50,000+ SKUs need physics properties
Humanoid Robots
Whole-body interaction training assets
Sim-to-Real Transfer
Close the gap with accurate physics data