TIAGo PlayGround

easynav_playground_tiago simulates PAL Robotics’ TIAGo in the AWS RoboMaker small house: its mobile base (pmb2), lifting torso, pan-tilt head with an RGBD camera, 7-DoF arm with the PAL gripper and wrist force/torque sensor, base laser, sonars and IMU. The base drives with EasyNav, and the torso, head, arm and gripper move under ros2_control.

See PlayGrounds to install it.

Launching EasyNav

Each easynav_<config>.launch.yaml starts Gazebo, TIAGo, EasyNav with that configuration and RViz2:

ros2 launch easynav_playground_tiago easynav_costmap_rpp.launch.yaml

Once RViz2 is up, send a goal with the 2D Goal Pose tool.

Launch file	Configuration
easynav_costmap_rpp.launch.yaml	Costmap from home2.yaml, costmap AMCL with the base laser, A* and Regulated Pure Pursuit.
easynav_navmap_bonxai_amcl.launch.yaml	NavMap (home2_20cm.navmap) and Bonxai (house.pcd) maps, NavMap AMCL against the Bonxai map with the base laser, NavMap A* and Regulated Pure Pursuit. The head camera can be added as a second sensor in the parameters.

They accept params_file, rviz_config, gui (false runs Gazebo headless) and rviz (false skips RViz2).

Notes for your own configurations:

• TIAGo’s base laser is 0.095 m above the floor. EasyNav’s localizers and obstacle filters drop points below min_height (0.1 m by default) as floor hits, so these configurations set min_height: 0.05. In the NavMap obstacles filter, min_height is the height above the NavMap surface.

• The Bonxai map’s resolution must be the one house.pcd was built with (0.05 m): at the default 0.3 m, the laser shares a voxel layer with the floor and the NavMap AMCL sees every ray blocked.

• The base takes geometry_msgs/TwistStamped: EasyNav runs with use_cmd_vel_stamped: true, and the launch files remap cmd_vel_stamped to /mobile_base_controller/cmd_vel and odom to /mobile_base_controller/odom.

Simulation only

To run Gazebo and TIAGo without EasyNav or RViz2:

ros2 launch easynav_playground_tiago_worlds gazebo_sim.launch.yaml

The arm and torso start in PAL’s home posture. The torso, head, arm and gripper have joint_trajectory_controllers (torso_controller, head_controller, arm_controller, gripper_controller), so they take FollowJointTrajectory goals or trajectories on <controller>/joint_trajectory. For example, to look down:

ros2 topic pub --once /head_controller/joint_trajectory trajectory_msgs/msg/JointTrajectory \
  "{joint_names: [head_1_joint, head_2_joint], points: [{positions: [0.0, -0.6], time_from_start: {sec: 1}}]}"

Packages

Package	Contents
easynav_playground_tiago	EasyNav launch files, parameters and RViz2 configurations
easynav_playground_tiago_worlds	The small house world, its maps (costmap, NavMap and Bonxai) and the Gazebo launchers
easynav_playground_tiago_description	The TIAGo model: URDF, meshes and ros2_control controllers

The TIAGo model is PAL Robotics’ work (Copyright (c) 2022-2024 PAL Robotics S.L., Apache-2.0), expanded from the Gazebo Harmonic ports of its tiago_robot, pmb2_robot and pal_gripper packages. TIAGo is a robot by PAL Robotics.