Development and Evaluation of an Intelligent Training Prototype to Improve Mobility and Reduce Anxiety in Visually Impaired Dogs

Abstract

Vision loss in dogs, with implications for anxiety and welfare, demands advanced technological solutions, as the market is currently limited to rudimentary approaches. This project aimed to develop an intelligent device to support mobility and reduce anxiety in visually impaired dogs. The iterative design methodology of the Costa Rica Institute of Technology was employed. The prototype integrates a proximity sensor (Sharp GP2Y0A21YKOF) and a Micro Arduino ATmega32U4 module, designed to activate multimodal alerts (buzzer and vibration motor) upon detecting an obstacle. A button allows the owner to select the most suitable alert mode for their pet’s individual needs. Usability tests, conducted through direct observation with five pairs of owners and senior dogs (age >8 years), validated the design hypothesis. A key and verifiable result was obtained: 80% of the participating dogs successfully assimilated the physical use of the device through brief positive reinforcement training, confirming the feasibility of learning in the senior population. Qualitative findings justified essential design improvements, such as the addition of a fastening strap to optimize stability and the refinement of the casing. Market analysis confirms the innovation of the proposal by not identifying intelligent solutions that actively address canine visual impairment. The main methodological limitation was the absence of quantitative validation for the alert parameters (decibels and intensity), a decision based on ethical criteria. Therefore, the project establishes a solid foundation for future work, emphasizing that the next critical steps involve the calibration and validation of the alert effectiveness with animal behavior experts in controlled environments, ensuring safety and welfare.