Rmagine

A robot wants to simulate too

Code   •   Documentation   •   Videos   •   Issues   •   Examples

Description

Library for fast and accurate simulation of range sensors in large 3D environments using ray tracing. The simulations can even be computed on embedded devices installed on a robot. Rmagine has been specifically developed to

• perform multiple sensor simulations simultaneously and in realtime
• distribute computations to various devices: CPU, GPU, RTX..
• hold data at device of computation
• reduce graphical overhead (offscreen-rendering)
• solve runtime critical tasks

Spherical, Pinhole or fully customizable models.	Query several attributes at intersection.

Installation / Usage

See Wiki page for further instructions.

Example

This examples shows how to simulate ranges of 100 Velodyne VLP-16 sensor using Embree backbone. First, the following headers need to be included:

// Map
#include <rmagine/map/EmbreeMap.hpp>
// Sensor Models
#include <rmagine/types/sensor_models.h>
// Simulators
#include <rmagine/simulation/SphereSimulatorEmbree.hpp>

namespace rm = rmagine;

The following code loads a map "my_mesh.ply" and simulates 100 Velodyne VLP-16 scans from certain predefined poses. Hits and Ranges are chosen as return attributes.

// loading a map from disk
std::string path_to_mesh = "my_mesh.ply";
rm::EmbreeMapPtr map = rm::load_embree_map(path_to_mesh);

// defining a model
rm::SphericalModel velo_model = rm::vlp16_900();

// construct a simulator
rm::SphereSimulatorEmbree sim;
sim.setMap(map);
sim.setModel(velo_model);

// 100 Transformations between base and map. e.g. poses of the robot
rm::Memory<rm::Transform, rm::RAM> Tbm(100);

for(size_t i=0; i < Tbm.size(); i++)
{
    rm::Transform T = rm::Transform::Identity();
    T.t = {2.0, 0.0, 0.0}; // position (2,0,0)
    rm::EulerAngles e = {0.0, 0.0, 1.0}; // orientation (0,0,1) radian - as euler angles
    T.R.set(e); // euler internally converted to quaternion
    Tbm[i] = T; // Write Transform/Pose to Memory
}

// add your desired attributes at intersection here
// - optimizes the code at compile time
// Possible Attributes (rmagine/simulation/SimulationResults.hpp):
// - Hits
// - Ranges
// - Points
// - Normals
// - FaceIds
// - GeomIds
// - ObjectIds
using ResultT = rm::Bundle<
    rm::Hits<rm::RAM>, 
    rm::Ranges<rm::RAM>
>;
// querying every attribute with 'rm::IntAttrAll' instead of 'ResultT'

ResultT result = sim.simulate<ResultT>(poses);
// result.hits, result.ranges contain the resulting attribute buffers
std::cout << "printing the first ray's range: " << result.ranges[0] << std::endl;

// or slice the results for the scan of pose 5
auto ranges5 = result.ranges(5 * model.size(), 6 * model.size());
std::cout << "printing the first ray's range of the fifth scan: " << ranges5[0] << std::endl;

More detailed examples explaining each step and how to customize it to your needs are explained in the Wiki. More examples can be found here: https://github.com/aock/rmagine_examples.

Citation

Please reference the following papers when using the Rmagine library in your scientific work.

@inproceedings{mock2023rmagine,
  title={Rmagine: 3D Range Sensor Simulation in Polygonal Maps via Ray Tracing for Embedded Hardware on Mobile Robots}, 
  author={Mock, Alexander and Wiemann, Thomas and Hertzberg, Joachim},
  booktitle={IEEE International Conference on Robotics and Automation (ICRA)}, 
  year={2023}
}

Rmagine-accelerated Applications

• rmagine_gazebo_plugins
• RMCL
• radarays_ros

Contributions

We are open to all contributions. These can be issues, pull requests or feedback that help us to improve this OpenSource project.

Build Status

	Ubuntu 20	Ubuntu 22
rmagine::core
rmagine::embree
rmagine::cuda	...	...
rmagine::optix	...	...

News

Dec 5th 2023

New version 2.2.2 is available now and brings convenience updates for ROS-users. Just place Rmagine into your ROS-workspace and it will compile. Via find_package(rmagine COMPONENTS [...]) you can still find Rmagine's components as if you would install it globally on your system. We tested it with

• ROS1 - noetic
• ROS2 - humble

Normally you would set OptiX_INCLUDE_DIR via cmake flags. Now we provide an additional option: Set the environment variable OPTIX_HEADER_DIR for example in your .bashrc-file:

export OPTIX_INCLUDE_DIR=~/software/optix/NVIDIA-OptiX-SDK-7.4.0-linux64-x86_64/include

Especially if you place Rmagine into your ROS-workspace this option becomes very handy.

Older news ...