Teaching Robotic System in ROS2

Abstract

The primary objective of this ALIGN grant proposal is to develop an educational platform using Robotic Operating System 2 (ROS2) for teaching concepts of a typical robotic system. The platform can be integrated seamlessly in the Computer Engineering curriculum.  Unique to his proposal is that the educational materials will offer a scaffolded learning process in laboratory sessions across two modules of the CEG program.

Robotic Operating System 2 (ROS2) is an open-source framework designed to develop and control robotic systems and has become a popular platform for industrial applications. Its unique feature is its ability to integrate devices, often from different manufacturers, into one coherent system. The growing demand for automation in various industry sectors requiring advanced manufacturing systems such as Automotive, Aerospace, Electronics, and Semiconductor industries, has led to significant advancements in robotic technology. The adoption of ROS2 for collaborative robotic platforms has become a key driver for industrial transformations in those sectors. Nearly 55% of total commercial robots shipped in 2024, will have at least one ROS package installed, increasing the demand for qualified engineers in this area.

Through this ALIGN grant, we intend to develop educational materials for SIT students, specifically in the Computer Engineering program, with the objective of teaching concepts and theories in robotics through a practice-oriented and applied approach. We will leverage the ROS2 collection of software development tools, drivers, applications, and simulations to design and develop multiple diverse robotic applications on computer vision and mapping using Simultaneous Localization and Mapping (SLAM) toolbox and cartography and autonomous navigation that can be used in lab exercises for two modules in the Computer Engineering program. The educational materials will offer a scaffolded learning process in laboratory modules. They will cover python programming to build modular robotic software systems, sensor fusion for integration of data, system architecture, which is modular and scalable, teleoperation control, SLAM, autonomous navigation, computer vision for object recognition and obstacle detection. Overall, a “learning by doing” approach will be applied covering development steps from simulation to real-world examples.

Gazebo is a ROS2 simulation tool. It provides a seamless bridge between virtual and real robots. Working with virtual robots with simulated sensors in the virtual environment provides students with more insights in designing and developing robotic applications. This approach is closely linked to SIT’s applied learning pedagogy by providing students with practical experience in training and testing a robot in a controlled and customizable virtual space while reducing the costs and potential hazards posed by working with hardware. Students will learn specialized concepts and use hands-on exercises for different specific applications. They will be able to identify the limitations and learn from them by incrementally improving their designs. The implementation of computer vision in robotics offers the opportunity to introduce Machine Learning (ML) techniques into the design process. For example, once a trained model is created in the simulated environment, it will be programmed into a physical robot that can perform tasks in the real world. It is crucial for computer engineering students to have a realistic visually enhanced experience of training an autonomous robot, experience with working on a real robot and finally, complementing it with a live demo of a full size real-word example, such as the SIT autonomous car.

In summary, this grant aims at developing a platform and framework to coach students by giving them a solid understanding of the basic working principles of ROS2, ML and understanding the communication concepts that can leverage on the multifaceted transformation into different applications within the robotic ecosystem.