Cloud-based Digital Twin for Cognitive Robotics

TL;DR

The paper addresses the barrier of hardware-intensive cognitive robotics education by proposing a cloud-based digital twin platform. It integrates containerization, Kubernetes, ROS, Jupyter, RVizWeb, and XPRA to deliver web-based, per-user learning environments. Key contributions include semantic digital twins, cloud deployment via BinderHub, and visualization-enabled teaching of KRR, knowledge acquisition, and task executives with CRAM. The work demonstrates educational effectiveness and potential to democratize access to cognitive robotics.

Abstract

The paper presents a novel cloud-based digital twin learning platform for teaching and training concepts of cognitive robotics. Instead of forcing interested learners or students to install a new operating system and bulky, fragile software onto their personal laptops just to solve tutorials or coding assignments of a single lecture on robotics, it would be beneficial to avoid technical setups and directly dive into the content of cognitive robotics. To achieve this, the authors utilize containerization technologies and Kubernetes to deploy and operate containerized applications, including robotics simulation environments and software collections based on the Robot operating System (ROS). The web-based Integrated Development Environment JupyterLab is integrated with RvizWeb and XPRA to provide real-time visualization of sensor data and robot behavior in a user-friendly environment for interacting with robotics software. The paper also discusses the application of the platform in teaching Knowledge Representation, Reasoning, Acquisition and Retrieval, and Task-Executives. The authors conclude that the proposed platform is a valuable tool for education and research in cognitive robotics, and that it has the potential to democratize access to these fields. The platform has already been successfully employed in various academic courses, demonstrating its effectiveness in fostering knowledge and skill development.

Figures (4)

• Figure 1: Assignment on Pick&Place robots on the left, physics simulation on the right, showing the robots believed world state.
• Figure 2: The cloud service framework built on BinderHub ragan2018binder initiates a new docker container for each connecting student. The Figure \ref{['fig:arch-2']} showcasing the robot software framework within the container.
• Figure 3: To control a real-world robot, ROS gives commands to it. ROS is running in the Docker and is loaded with AI Applications, e.g. the Cognitive Robot Abstract Machine (CRAM). The Digital Twin Simulation, a 3D game engine, is visualized in Jupyter via RVizWeb and/or a lightweight display for visual applications. Jupyter Notebooks run code to control the AI Applications, while a Terminal allows for direct access to the ROS processes.
• Figure 4: ROSBoard JupyterROS widget in Jupyter to visualize laser scan (top left), the front camera feed (top mid) and the recorded map (top right). Gazebo Physics Simulator (bottom left) and RViz visualization of the environment map (bottom right) on the XPRA VNC virtual desktop, navigating the Tortugabot through the digital twin on the green path.