Development and control of a single-wheel robot: Practical Mechatronics approach

This article presents a Mechatronics approach to make the complex dynamical system satisfy the specification as desired. The Mechatronics approach has several phases: analytical design, system integration, sensing & control, and evaluation. The Mechatronics approach means that cycles of design, implementation, sensing, and control are repeated until the system satisfies the goal through evaluation. A single-wheel robot system called GYROBO after a robot that uses gyroscopic effect is developed and controlled by the Mechatronics approach. The goal of GYROBO is to navigate its terrain while maintaining the stable balance. However, successful balancing and navigation of a single-wheel robot are quite difficult and challenging since one point contact may fall down in lateral direction with ease. To have a successful balancing performance, many problems have to be solved a priori before applying any advanced control algorithms. Among several phases of analytical design, integration, sensing & control, and evaluation, the most important phase is the analytical design. However, the analytical design cannot guarantee successful performances due to the complexity of the system. Practical Mechatronic approach is to repeat cycles of system integration, sensing & control, and evaluation. After several modifications of mechanical assembly and relocation of components inside the wheel housing, simple linear controllers enable GYROBO to perform successful balancing and navigation. GYROBO is able to follow the specified trajectory given by a remote operator. Experimental studies of control of balancing, driving forward and backward, turning, and climbing over an obstacle of GYROBO are conducted to demonstrate and evaluate its functionality and support the concept of the Mechatronics approach to control complex systems. (C) 2013 Elsevier Ltd. All rights reserved.