OpenCanine V1: Robotic Canine Assistance for the Visually Impaired

This paper introduces OpenCanine V1, a versatile and low-cost robotic canine that is fully 3D-printable and open source. The robot is made from PLA+ material which offers durability and slight flexibility and is equipped with 12 servos controlled by an Arduino Uno to ensure precise movements. OpenCanine V1 invites community engagement, allowing anyone to adapt and modify it for a range of purposes—whether for education, research, or even new robotics projects. While the open-source nature of OpenCanine V1 allows for broad adaptation, we have specifically developed it to assist visually impaired individuals. We have implemented a Sequential Detection and Classification System—a multi-stage AI approach designed to address various real-world scenarios. Although training a model to cover every possible situation is highly challenging, this paper highlights our focus on guiding the visually impaired through zebra crossings, identifying potential trip hazards, and recognizing different road conditions. This multi-stage approach to object detection can be adapted for a variety of other applications by the community, ensuring that the same methodology can be utilized for different purposes, from military to healthcare and beyond.