A Self-Balancing Robot With Redundant Actuators for Obstacle Avoidance

Planar robots, such as self-balancing two-wheeled systems, are designed to operate in two-dimensional space and maintain stability while in motion. While these systems can navigate simple terrains, they struggle with complex obstacles requiring traversal from above or below. This paper presents a novel self-balancing robot capable of autonomous morphological reconfiguration between two-wheeled and four-wheeled configurations, enabling it to overcome such challenges. To ensure robust stability throughout these transitions, an adaptive PID control strategy is implemented, facilitating real-time gain adjustment based on the robot's dynamic physical state. Experimental results demonstrate that the adaptive controller maintains balance during both transformation phases, in contrast, a fixed-gain PID controller fails to maintain balance during these phases. The system illustrates promising capabilities for mobile robotics applications such as search and rescue or navigation in unstructured environments, while demonstrating a practical method for analytically adjusting PID gains in response to measured tracking error under changing dynamic conditions.