Design and Development of the RelBot AI Box: A Modular Platform for Enhancing Sensoring, Odometry, and AI Capabilities in Existing Robotic Systems

MSc assignment

The RELbot is an educational robotics platform equipped with foundational hardware, including a Raspberry Pi (RPI), an ICO board FPGA, wheel encoders, a small RGB camera, and ToF sensors for real-time collision avoidance. It serves as an excellent tool for learning about data transfer across platforms, implementing control loops, and understanding real-time systems. However, its potential extends beyond its current configuration.

Robotics education includes diverse topics such as navigation, sensor fusion, image processing, and artificial intelligence (AI). By enhancing the RELbot with additional advanced sensors and computational capabilities, it could transform into a comprehensive platform that supports practical exercises in robotics education. Such a platform would empower students and educators with hands-on opportunities to explore robotics concepts in greater depth compared to concept projects and simulations. The RELbot’s versatility makes it an ideal candidate for this enhancement, enabling it to bridge the gap between theoretical knowledge and practical application in advanced robotics.

The goal of this project is to design and develop an AI-powered sensor hub add-on called the ASAI Addon (Algorithms, Software, and Artificial Intelligence) for the RELbot. This modular box will expand the platform’s capabilities, enabling it to perform advanced robotics tasks such as SLAM, odometry, and other advanced robotic techniques. The sensor hub will integrate seamlessly with the RELbot while
also being compatible with other platforms using Raspberry Pi (RPI) and the appropriate ROS (Robot Operating System) distribution.

The approach involves developing a plug-and-play solution that is easy to set up and use. The design will focus on simplicity, ensuring that teachers, technicians, and students can work with the add-on effortlessly. The sensor hub will act as an interface for advanced sensors and AI processing, providing the RELbot with enhanced situational awareness and improving its utility as a robotics education tool. By incorporating sensors such as LiDAR, depth cameras, and IMUs, and leveraging the computational power of an AI module, the ASAI Add-on will significantly expand the RELbot’s capabilities and potential for use in robotics courses.

Figure 1: Quick sketch of the RELbot in combination with the ASAI Add-on