V3I2P16

Development of iCAT: A Smart Identifier Pet Robot  Integrating Computer Vision and Affective Computing

Josh Nathaniel M. Lanto¹, Jim Patrick G. Martinez²,   Ramuel D.J. Panganiban³, Eliza B. Ayo⁴*

Abstract

This study presents the design, development, and evaluation of iCAT, an interactive robotic companion that integrates machine learning-based object detection with affective feedback mechanisms. The prototype employs a dual-processor architecture comprising a Raspberry Pi 4 Model B for high-level computational tasks and an Arduino Uno for real-time sensor management and actuator control. The system utilizes OpenCV for computer vision-based object classification and the pyttsx3 library for speech synthesis, enabling real-time auditory feedback. Affective responses are simulated through servo-controlled tail movements and LED-based eye animations, designed to enhance human-robot interaction (HRI) quality. A descriptive developmental research design guided the iterative hardware-software integration process. System evaluation employed comprehensive software quality assurance (SQA) criteria, assessing functionality, usability, reliability, maintainability, and detection accuracy. Controlled testing under varied environmental conditions revealed object detection accuracies ranging from 80% to 100%, contingent upon object placement, background contrast, and ambient lighting. Mean system response time measured 0.75 seconds (SD = 0.25s) from object detection to vocal output. The device sustained continuous operation for approximately 2.5 hours in active mode and 4 hours in idle mode per charge cycle. User evaluation (N = 30) demonstrated high satisfaction ratings (M = 4.6/5.0, SD = 0.5]) across usability dimensions. The findings demonstrate iCAT’s viability as a modular platform for educational technology, assistive robotics, and emotionally responsive human-robot interfaces, particularly within smart environment applications requiring intuitive multimodal interaction.

Keywords:

Object detection, interactive robotics, speech synthesis, human-robot interaction, affective computing, open-source hardware, computer vision