3D Sensor

TIAGo 3D sensor configuration in AMAS

Introduction

To setup TIAGo 3D sensor in AMAS user needs to:

• add SenseKit

• select calibration type with pose data (coming from TIAGo)

• fine-tune the calibration

• optionally enable 2D-3D fusion

Adding SenseKit

Follow SenseKit Addition and Configuration with:

• your TIAGo robot IP address

• default ROS port 9090

Selecting Calibration Type

Follow SenseKit to Robot Calibration with

• End Effector with Pose mount type

You don't need calibration board and don't need to take calibration samples. TIAGo is tightly integrated robot and calibrated pose data is available from the robot.

End Effector with Pose mount type should be selected even though 3D sensor is mounted on TIAGo's head.

Immediately after selecting End Effector with Pose mount type save the calibration without taking any samples.

Fine-Tuning the Calibration

For TIAGo with Astra S camera calibration data sent from the robot is just a bit off.

Follow Manual Sensekit Calibrationwith

• -0.02 for Right

• 0.02 for Up

• 0.02 for Forward

This should be enough to get correct 3D data overlap but system may be further fine-tuned if needed.

Enable 2D-3D fusion (optional)

For TIAGo with Astra-S camera the range of quality 3D data is limited.

You may optionally enable 2D-3D fusion to increase field of view and see 3D data emerging from 2D at cutoff distance.

Follow 2D - 3D fusion setup, typically with:

• 2 m Cutoff Distance and Cutoff Replacement

Fine tune cutoff distance/replacement as needed.

Below is an example of the 2D 3D view from the Tiago First person view:

• The chair and items on the chair are within the cut off distance and so shown in 3D

• The flooring and people in the background are outside of the cut off distance and so shown in 2D

Example of 2D 3D view