Create a basic simulation with bodies, geometries and shapes. 

Shows how to:

• Create a basic simulation with bodies, geometries and shapes.
• Can be started with the native python or agxViewer application. However, no graphics is generated in this script, so when you start it with agxViewer, you need to press 'g' to see the geometry.

• Start tutorial0_basic_simulation.agxPy in agxViewer
• View source code

Create bodies/geometries and attach rendering objects to each of them.

Shows how to: 

• Create a simulation including graphics which is loadable with agxViewer.
• Load a graphical object from a file, and connect it to a geometry so that it moves with the geometry.

This script can only be loaded with agxViewer, NOT python.

• Start tutorial1_graphics.agxPy in agxViewer
• View source code

Create bodies/geometries and attach rendering objects to each of them.

Shows how to:  

• Demonstrates how to create a Python script which can be loaded both with the native python.exe and agxViewer.
• Use arguments when python is used.
• Start tutorial2_python3_agxViewer.agxPy in agxViewer
• View source code

This tutorial show how to listen for simulation event and read forces from contacts and constraints.

Shows how to:

• Create a StepEventListener to listen for events in the simulation loop
• Access contact data/forces and write them to the screen
• Get all constraints in a system
• Access constraint data and read forces
• Draw simple text onto the screen

• Start tutorial3_events_and_reading_forces.agxPy in agxViewer
• View source code

This script randomly creates objects during runtime.

Shows how to:

• Use AutoSleep
• Create a StepEventListener to listen for events in the simulation loop
• Create and remove objects during runtime.

• Start tutorial4_creating_objects_during_runtime.agxPy in agxViewer
• View source code

• Read simulation data from .aagx file
• Group and move objects using agxSDK::Assembly
• Get objects from an Assembly
• Access data using names

• Start tutorial5_read_scene_files.agxPy in agxViewer
• View source code

• Read a mesh file from disk and create a TriangleMesh shape for collision detection
• Read a mesh file and use for rendering

• Start tutorial5_read_scene_files.agxPy in agxViewer
• View source code

• agxSDK.GuiEventListener for listening to keyboard and mouse events
• Change a shape in size (radius of a sphere), including the graphics

• Start tutorial6_read_keyboard_events_change_shapes.agxPy in agxViewer
• View source code

Shows how to:

Build a "lumped" object by attaching bodies/constraint to each other

• Set compliance to make lumped objects stiffer (all the way to infinite)

• Start tutorial9_building_deformable_objects.agxPy in agxViewer
• View source code

Shows how to:

Use the iterative solver specified on a per-material basis. The iterative solver does not give the same precision as the direct, but is a lot faster for large contact/friction system.

Also by using the ParallelPgs solver good performance can be gained with more than 1 thread.

• Start tutorial_iterativeContactFriction.agxPy in agxViewer
• View source code

This tutorial demonstrates all the constraints available in AGX.

• Start tutorial_constraints.agxPy in agxViewer
• View source code

This is a tutorial demonstrating how to use the TireModel class for simulating the wheels of a vehicle.

Control the car with up/down/left/right. If you have a gamepad connected it can be used to control the car.

• Start tutorial_tireModel.agxPy in agxViewer
• View source code

This script demonstrates how to build a hydraulic driven winch with wires. During simulation data will be plotted: rpm, pressure, flow etc.

Included in the tutorial is:

• RPM controller (written in Python)
• Engine
• Hydraulic pump
• Shafts
• Relief valve
• GearBox
• Clutch
  
• WireWinchActuator
  

• Start tutorial_hydraulicWinch.agxPy in agxViewer
• View source code

This script contains several scenes where each illustrate how to use the agxCable api.

Press key 1..7 to choose scene. Shows how to:

• Create and route a cable.
• Attach cable to objects.
• Set cable properties.
• Inspect dynamic state.
• Control initialization.
• Detect cable collisions.
• Attach objects on initialized cable.

This script demonstrates how to setup and use the CableDamage class available in AGX Dynamics.

Cable damage can measure the accumulated strain in a cable tue to bend/twist/stretch

This script demonstrates how to setup and use the CableTunnelingGuard class available in AGX Dynamics and shows off a few examples of parameterizing the class to avoid certain types of tunneling.

Tunneling caused by low simulation frequencies can be reduced by using the CableTunnelingGuard class. It offers default values to suit many situations but also exposes parameters to more aggressively prevent tunneling, if needed.

This tutorial shows how to use the FractureGenerator class for breaking and fracturing geometries in the simulation using 3D Voronoi tessellation.

Press key 1..4 to choose scene.

This scene demonstrates how to create a surface velocity conveyor belt # in order to set up a basic conveyor belt simulation transporting goods.

This scene demonstrates how to best make use of the various contact friction models available in AGX Dynamics

This script shows you how to set up a scene with accurate frictional contact simulation using the DIRECT solver

• Start tutorial_grasping.agxPy in agxViewer

View source code:

• tutorial_grasping.agxPy
• tutorial_grasping_sceneDescription.py
• tutorial_grasping_events.py

This file demonstrates the functionality of the virtual inertia class that adds virtual inertia elements to a constraint that will add extra resistance to translational and rotational movements in the general constraint axis. This can be used to emulate mechanisms modeled with constraints that have internal inertia

This demonstrates how to use MergedBody and MergeSplitHandler to automatically simplify a scene to achieve speedup.

This scene demonstrates how AMOR can speed up simulation in a scene where many bodies are created in a dynamic scene with a spinning "table".

This tutorial will show you how to use the agxModel::Track class to model a tracked vehicle/crawler.

This tutorial illustrates how to use the agxDriveTrain.CombustionEngine class for simulating an engine.. Start tutorial_driveTrain_combustionEngine.agxPy in agxViewer
View source code

This scripts demonstrates how to use the WheelJoint constraint for modelling the ackerman steering system including suspension.

This script demonstrates the different steering models available for a WheelJoint:

• AckermannSteering
• BellCrankSteering
• ShortRackPinionSteering
• DavisSteering

Demonstrates how to initialize and set up a simulation using multiple Panda robots loaded from URDF. We compute and apply forward kinematics based on random joint configurations, ensuring valid poses by respecting joint limits.

• Start tutorial_set_robot_pose.agxPy in agxViewer
• View source code

Shows how to set up and perform gravity compensation with a Franka Emika Panda robot. We use inverse dynamics to calculate the torques needed at each joint to maintain the robot's position without acceleration, counteracting the effects of gravity. These torques are then applied to the robot's joints, ensuring it holds its pose.

• Start tutorial_gravity_compensation.agxPy in agxViewer
• View source code

This tutorial demonstrates how to generate straight-line trajectories in joint and task space and track them using feed-forward torque control. We compute the torques using inverse dynamics to actuate the Panda robot.

• Start tutorial_trajectory_tracking.agxPy in agxViewer
• View source code

Demonstrates how to track a moving target using position-based servoing. Postion-based servoing computes the tracking error between the end-effector's and object's pose. It then uses differential kinematics to compute the manipulator Jacobian and solves for the desired joint velocities. These joint velocities are input to the actuators of the Panda robot.

• Start tutorial_differential_kinematics.agxPy in agxViewer
• View source code

This script shows how to implement computed torque control, or the inverse dynamics controller, to track a trajectory. Along the trajectory, we unexpectedly let the robot collide with an object, causing it to diverge from the intended path. However, since computed torque control relies on feedback, it recovers gracefully.

• Start tutorial_computed_torque_control.agxPy in agxViewer
• View source code

This tutorial shows how to control a Panda robot to pick up boxes through frictional contacts and stack them. We use two controllers to manage the robot's motion. A high-level controller handles tasks related to the gripper and determines the robot's desired trajectory. A low-level controller tracks the target joint angles provided by the high-level controller and uses a force/torque sensor to estimate the load on the tool when moving different objects.

• Start tutorial_pick_and_place.agxPy in agxViewer
• View source code

This script demonstrates how to model electric motors for a controlled robot.

• Start tutorial_robot.agxPy in agxViewer
• View source code

This tutorial demonstrates how a robot can be modelled from individual parts. It also show how an ordinary PID controller can be used to control the movements of the robot to follow a predefined trajectory read from a file. PID controller chapter in the User manual goes through this example in more detail

This tutorial demonstrates how to model a custom suction gripper using the

agxModel::SuctionGripper

This tutorial illustrates how to model a gripping device. To be able to pick up an object using only friction, you will need to consider the material/solve model for the contacts between the gripper and the lifted object

This tutorial will demonstrate how to use the agxModel::Tree class to simulate a flexible tree structures with stem and branches. The trees can be broken into parts because of collisions or forces.

This script demonstrates how to render color and depth buffer to image If matplotlib is present it shows the buffers in a figure.

This script shows how to create a Lidar sensor and read back data (point cloud) from it. The Lidar uses GPU accelerated ray tracing.

Note: The Lidar simulation requires a graphics card that supports CUDA.

This script shows how to create an image based Lidar sensor together with depth and color cameras attached to a moving body. Note that this tutorial does not use the now recommended Lidar sensor, part of the agxSensor library which uses GPU accelerated ray tracing. See tutorial "Lidar" above for details of that Lidar sensor.

This tutorial demonstrates how to use the

agxUtil.reduceMesh()

This tutorial demonstrates how to use the

agxUtil.PrimitiveMeshGenerator.generateVoronoiPolyhedraMesh

This tutorial show how you can use the

agxCollide.computeShapeDistance

This tutorial show how you can use the

agxUtil.createOrientedBox, agxUtil.createOrientedCylinder

agxUtil.createOrientedCapsule

This tutorial shows how to use the agxModel::Beam class.

This tutorial demonstrates how to model a custom suction gripper using the

agxModel::SuctionGripper

This script demonstrates how to use predefined materials from the Material Library for various objects.

This script demonstrates how a per contact point friction model can be setup using ContactEventListener. A texture is used as a lookup table for getting non-homogeneous friction over the surface of a box. The lookup into the friction texture is based on projection down to the x/y plane.

This demonstrates basic friction modeling. The angle of the plane can be altered using left/right key. Friction coefficient can be altered using up/down key.

This script demonstrates the basics of how to set up and use the secondary constraints of a ball joint, specifically cone limit, twist range and friction controller constraints.

This script demonstrates how to use the Python API for accessing joystick/gamepad in a simulation.

This example demonstrates a tool that will recompute texture coordinates from a texture atlas.

• Start tutorial_recomputeTextureCoordinates.agxPy in agxViewer
• View source code

This script demonstrates the basics of enabling and accessing the position of center of buoyancy in a hydrodynamics scene.

ObserverFrame can be used to measure data for a rigid body including, position, velocity and acceleration.

agxPlot is an API (available in Python, C++, C#) for easy plotting of data. It uses a WebAPI for the actual plotting.

The reconfigure request is a utility class in agxUtil with functionality to compute new transforms for a collection of bodies given a set of values for constraint angles.

This tutorial shows how it can be used to change the pose of a wheel loader, an excavator as well as a robot.

The tutorial also shows extra handling which can be needed after updating many body transforms to maintain the simulation in a good state.