Explainable Agency by Revealing Suboptimalityin Child-Robot Learning Scenarios

Revealing the internal workings of a robot can help a human better understand the robot's behaviors. How to reveal such workings, e.g., via explanation generation, remains a significant challenge. This gets even more complex when these explanations are targeted towards children. Therefore, we propose a search-based approach to generate contrastive explanations using optimal and sub-optimal plans and implement it in a scenario for children. In the application scenario, the child and the robot learn together how to play a zero-sum game that requires logical and mathematical thinking. We report results around our explanation generation system that was successfully deployed among seven-year-old children. Our results show trends that the generated explanations were able to positively affect the children's perceived difficulty in learning the zero-sum game.

For additional details about the research work you can check out our paper: Explainable Agency by Revealing Suboptimalityin Child-Robot Learning Scenarios

Test

To run the code:

• open a terminal and launch: $ roscore
• open a second terminal into your repository folder and launch minmax.launch: $ roslaunch minmax.launch

Config

• Install ROS melodic and catkin

• Clone the repository

  • $ git clone https://github.com/Silviatulli/suboptimax.git

• Make the python files executable Run the following command for each script:

• $ chmod +x filename.py

• Build a catkin workspace and source the setup file

  • $ cd ~/catkin_ws
  • $ catkin_make

• Add the workspace to the ROS environment

  • $. ~/catkin_ws/devel/setup.bash

• Make sure that the CMakeLists.txt file is configured properly

  • All the services and the dependencies should be as follows:
    • find_package (catkin REQUIRED COMPONENTS roscpp rospy std_msgs message_generation message_runtime )
    • add_service_files (FILES Decision.srv GameState.srv Plan.srv RobotExplanation.srv RobotTalk.srv )

• Make it work with the Robot

  • If you work with the NAO Robot uncomment the line 7, from 41 to 65 and 85 (self.robot_communication.say(self.explanation_text)) in the file robot_manager.py.
  • Create a folder sdk that contains the pynaoqi sdk required and modify your bashrc ($gedit ~/.bashrc) to indicate the python and library paths as follows:
    • export PYTHONPATH=$PYTHONPATH:~/sdk/pynaoqi-python2.7-2.1.2.17-linux64
    • export LD_LIBRARY_PATH=${LD_LIBRARY_PATH}:/home/<your_pc_name>/sdk/pynaoqi-python2.7-2.1.2.17-linux64
    • you can download the pynaoqi sdk following this guide