Major release of AGX Dynamics with new features!

A new release, AGX Dynamics, is now available for download!

This time we have been working hard on some new interesting features. As Machine Learning is becoming more central to simulations we now make it easier for you to get up and running with such applications. In addition to stability and performance improvements, this major release also contains a number of new features:

  • Support for URDF models now available.
    The feature is accessible from C++, C# and Python.
  • A new WheelJoint for modelling of four wheeled vehicles including Ackermann steering and suspension.
  • A brand new implementation of a TorqueConverter.
    Note: Restoring an older archive including a torque converter will restore the deprecated model.
  • A gradual merge algorithm introduced to the agxTerrain::Terrain model that will improve the visual quality when particles are merged with the terrain. Introduces the agxModel::GranularPostProcessor that computes GranularFields.
    The available fields includes: Mass field, Momentum field, Velocity field and Custom field (weighted sum from custom buffer).
  • Adding two Machine Learning (ML) tutorial python scripts that show how to integrate AGX Dynamics with OpenAI Gym.
    See the directory data/python/RL and Appendix 6 in the documentation.
  • RL with AGX Dynamics Mesh reader now support a large range of file formats including: STL, Collada, etc.
  • Example Python script demonstrating Lidar and Camera (color/depth) sensors.
  • Also, make sure you checkout our new plugin for Unity, AGX Dynamics for Unity: