Adding a Controller in Isaac Sim — NVIDIA Robot Tutorial

Title: Adding a Controller — Isaac Sim Documentation

URL Source: https://docs.isaacsim.omniverse.nvidia.com/latest/core_api_tutorials/tutorial_core_adding_controller.html

Published Time: Tue, 21 Oct 2025 19:24:36 GMT

Markdown Content: Adding a Controller#

Learning Objectives#

This tutorial describes how to create and use a custom controller to move a mobile robot. It then describes how to use the controllers available in NVIDIA Isaac Sim. After this tutorial, you should be more comfortable with adding and controlling robots in NVIDIA Isaac Sim.

10 Minute Tutorial

Getting Started#

Prerequisites

Review Hello Robot prior to beginning this tutorial.

Begin with the source code open from the Hello Robot tutorial, by clicking the Open Source Code button in the Hello World Example window.

Note

Pressing STOP, then PLAY in this workflow might not reset the world properly. Use the RESET button instead.

Creating a Custom Controller#

First, code an open-loop controller that uses the unicycle model for differential drive. Controllers in NVIDIA Isaac Sim inherit from the BaseController interface. You will need to implement a forward method, and it has to return an ArticulationAction type.

1from isaacsim.examples.interactive.base_sample import BaseSample 2from isaacsim.core.utils.nucleus import get_assets_root_path 3from isaacsim.robot.wheeled_robots.robots import WheeledRobot 4from isaacsim.core.utils.types import ArticulationAction 5 from isaacsim.core.api.controllers import BaseController 6import numpy as np 7 8 9 class CoolController(BaseController):10 def init (self):11 super(). init (name="my_cool_controller")12 # An open loop controller that uses a unicycle model13 self._wheel_radius = 0.0314 self._wheel_base = 0.112515 return1617 def forward(self, command):18 # command will have two elements, first element is the forward velocity19 # second element is the angular velocity (yaw only).20 joint_velocities = [0.0, 0.0]21 joint_velocities[0] = ((2 * command[0]) - (command[1] * self._wheel_base)) / (2 * self._wheel_radius)22 joint_velocities[1] = ((2 * command[0]) + (command[1] * self._wheel_base)) / (2 * self._wheel_radius)23 # A controller has to return an ArticulationAction24 return ArticulationAction(joint_velocities=joint_velocities)25 26 27class HelloWorld(BaseSample): 28 def init (self) -> None: 29 super(). init () 30 return 31 32 def setup_scene(self): 33 world = self.get_world() 34 world.scene.add_default_ground_plane() 35 assets_root_path = get_assets_root_path() 36 jetbot_asset_path = assets_root_path + "/Isaac/Robots/NVIDIA/Jetbot/jetbot.usd" 37 world.scene.add( 38 WheeledRobot( 39 prim_path="/World/Fancy_Robot", 40 name="fancy_robot", 41 wheel_dof_names=["left_wheel_joint", "right_wheel_joint"], 42 create_robot=True, 43 usd_path=jetbot_asset_path, 44 ) 45 ) 46 return 47 48 async def setup_post_load(self): 49 self._world = self.get_world() 50 self._jetbot = self._world.scene.get_object("fancy_robot") 51 self._world.add_physics_callback("sending_actions", callback_fn=self.send_robot_actions) 52 # Initialize our controller after load and the first reset53 self._my_controller = CoolController()54 return 55 56 def send_robot_actions(self, step_size): 57 #apply the actions calculated by the controller58 self._jetbot.apply_action(self._my_controller.forward(command=[0.20, np.pi / 4]))59 return

Using the Available Controllers#

NVIDIA Isaac Sim also provides different controllers under many robot extensions. Re-write the previous code using the DifferentialController class and add a WheelBasePoseController.

1from isaacsim.examples.interactive.base_sample import BaseSample 2from isaacsim.core.utils.nucleus import get_assets_root_path 3from isaacsim.robot.wheeled_robots.robots import WheeledRobot 4# This extension includes several generic controllers that could be used with multiple robots 5 from isaacsim.robot.wheeled_robots.controllers.wheel_base_pose_controller import WheelBasePoseController 6# Robot specific controller 7 from isaacsim.robot.wheeled_robots.controllers.differential_controller import DifferentialController 8import numpy as np 9 10 11class HelloWorld(BaseSample): 12 def init (self) -> None: 13 super(). init () 14 return 15 16 def setup_scene(self): 17 world = self.get_world() 18 world.scene.add_default_ground_plane() 19 assets_root_path = get_assets_root_path() 20 jetbot_asset_path = assets_root_path + "/Isaac/Robots/NVIDIA/Jetbot/jetbot.usd" 21 world.scene.add( 22 WheeledRobot( 23 prim_path="/World/Fancy_Robot", 24 name="fancy_robot", 25 wheel_dof_names=["left_wheel_joint", "right_wheel_joint"], 26 create_robot=True, 27 usd_path=jetbot_asset_path, 28 ) 29 ) 30 return 31 32 async def setup_post_load(self): 33 self._world = self.get_world() 34 self._jetbot = self._world.scene.get_object("fancy_robot") 35 self._world.add_physics_callback("sending_actions", callback_fn=self.send_robot_actions) 36 # Initialize our controller after load and the first reset37 self._my_controller = WheelBasePoseController(name="cool_controller",38 open_loop_wheel_controller=39 DifferentialController(name="simple_control",40 wheel_radius=0.03, wheel_base=0.1125),41 is_holonomic=False)42 return 43 44 def send_robot_actions(self, step_size): 45 position, orientation = self._jetbot.get_world_pose()46 self._jetbot.apply_action(self._my_controller.forward(start_position=position,47 start_orientation=orientation,48 goal_position=np.array([0.8, 0.8])))49 return

Press Ctrl+S to save and hot reload the example. Then press the LOAD button to reload the scene.

Summary#

This tutorial covered the following topics:

Creating a custom controller to move a mobile robot
Using Controller classes from NVIDIA Isaac Sim

Next Steps#

Continue to the next tutorial in the Essential Tutorials series, Adding a Manipulator Robot, to learn how to add a manipulator robot to the simulation.

Links/Buttons: