- The paper presents the MBot as a low-cost, modular platform that overcomes cost and flexibility challenges in hands-on robotics education.
- It details a versatile software architecture including sensor drivers, mapping, localization, and the MBot Bridge API for projects ranging from basic sensing to advanced autonomous navigation.
- The paper demonstrates MBot's successful adoption across diverse institutions, showcasing its adaptability for both introductory courses and advanced robotics labs.
Introduction
The emerging field of robotics has created a demand for educational platforms that can effectively train the next generation of roboticists. Real robot platforms are integral to hands-on robotics education, allowing students to experience the unique challenges of programming physical systems. However, limitations such as high cost and lack of flexibility often hinder the integration of such robots into academic curricula. The Michigan Robotics MBot, developed to address these challenges, is a low-cost, modular mobile robot that facilitates scalable and flexible instruction in robotics.
Development of the MBot Ecosystem
The MBot was initially developed to enable undergraduate instruction at the University of Michigan. Over time, the platform has evolved to support a diverse range of courses, including advanced topics like robot localization and control. It has grown into an ecosystem adaptable for both introductory and graduate-level courses. With the capability to be assembled from off-the-shelf components, the MBot platform features a versatile backbone: the MBot Robotics Control Board. This board's modularity allows educators to adjust the robot's complexity based on the course's requirements, from a budget version for basic sensing to advanced configurations featuring 2D Lidar and RGB sensor suites.
The Architecture of the MBot
At the core of the MBot's architecture is a user-friendly software stack that offers a wide range of functionalities, such as sensor drivers, mapping, localization, and path planning. Students can directly engage with the stack for advanced applications or use the synchronous MBot Bridge API for simpler projects. The incorporated web app makes the robot accessible through a browser, enabling visualization and teleoperation without additional software installations. The MBot platform's programming flexibility and its array of tools make it suitable for a broad scope of educational objectives and student experience levels.
Teaching Possibilities and Accessibility
Since its inception, the MBot has been influential in courses at the University of Michigan and other institutions, including Berea College, Howard University, and Morehouse College. The platform supports a wide variety of learning opportunities, from basic introductory courses to advanced robotics labs. Additionally, with an emphasis on user-friendliness and remote learning capabilities, the MBot makes robotics education more accessible to students, regardless of their backgrounds or physical locations. Whether students are learning basic control or diving into complex autonomous navigation, the MBot provides an ideal environment for cultivating robotics expertise.