Robodog

Autonomous Legged Robots for Assisted Living

According to a recent report from the external page WHO, approximately 2.2 billion people worldwide have a near or distant vision impairment, including 217 million with moderate to severe impairment and 36 million who are blind. This represents a significant portion of the population that faces challenges in navigating and traveling independently.

Traditional aids like white canes and guide dogs offer limited assistance: canes cannot detect distant obstacles, and guide dogs are costly and scarce. Our project aims to enhance navigation aids by developing robotic guide dogs. These are built on relatively affordable, commercially available quadruped robots, which we equip with specialized sensors and computational units. Unlike wheeled robotic platforms, these robotic dogs can navigate stairs, tight spaces, and challenging terrains, overcoming the same obstacles a human would encounter.

This work leverages the expertise of the Center for Project-Based Learning, specializing in low-power embedded systems and novel sensor technologies. Utilizing this specialized knowledge, we equip the robotic guide dogs with sensor fusion and autonomous navigation algorithms. These algorithms enable the robots to guide visually impaired individuals from one location to another in unfamiliar environments while safely avoiding obstacles in real time. This integrated approach combines high-level navigation capabilities, commonly found in smartphone apps, with immediate, low-level obstacle detection traditionally provided by canes or guide dogs.