Issuu

International Research Journal of Engineering and Technology (IRJET)

e-ISSN: 2395-0056

Volume: 12 Issue: 12 | Dec 2025

p-ISSN: 2395-0072

www.irjet.net

EchoPaw: A Mini Robotic Dog P Yashwanth1, Y Deepika2, Shivakumara S3, Dr. Mohamed Saleem4 123UG Student, Dept. of Robotics and Artificial Intelligence, Bangalore Technological Institute, Karnataka, India

4Professor and Head of department, Dept. of Robotics and Artificial Intelligence, Bangalore Technological Institute,

Karnataka, India ---------------------------------------------------------------------***---------------------------------------------------------------------

Abstract – EchoPaw is a design of quadruped robotic

educating students on the concept of embedded systems, robotics as well as interactions with a robot.

system developed to connect human-robot interaction using offline voice commands, servo driven locomotion and AI based communication. The design is integrated with esp32 microcontroller, PCA9685 servo driver, sg90 continuous rotation servos, ultrasonic sensor and DFRobot offline voice recognition module to perform actions. A Bluetooth speaker is integrated in design where it can communicate with humans using LLM’s live voice chat. it performs basic actions like sitting, walking, standing and waving hand. The design is developed in low cost, modular hardware and smooth locomotion. The developed design results reliable voice recognition, smooth locomotion and obstacles detection. This work highlights the potential of EchoPaw as an educational and research-oriented mini quadruped robot.

1.2 Feasibility study

Key Words: Quadruped Robot, Voice Recognition, ESP32, Obstacle Detection, PCA9685, Servo Control, Human Robot Interaction.

Technical feasibility: its uses commonly available components ESP32, SG90 servos, ultrasonic sensors and servo driver. All these modules are easy to program and integrate in the system we opt.

Economic feasibility: all the components used in the design is low cost and affordable for student projects.

Operational feasibility: the design is used in indoor only.it gives a stable performance due to offline recognition module and safely operated due to obstacle detection

2. LITERATURE SURVEY 1. Paper Title: Speech Interaction with an Emotional Robotic Dog

1. INTRODUCTION Robots are also being used to serve as assistance, learning and interactions. Among the various robots the most commonly used robots are quadruped robots where the design we developed similarly looks like dog. Most of the robotic dogs available now requires controls via the internet or voice assistance. The EchoPaw includes the functionality to respond to verbal commands, carry out the dog like action and communicate with the user by give human like voice. The design developed in low cost components responds to human voice as response.

Author: Christian M. Jones, Andrew Deeming

1.1 Relevance Of The Study

2. Paper Title: On the Robustness of Speech Emotion Recognition for Human-Robot Interaction with Deep Neural Networks

Abstract and relevance: In the future, consumer electronics will be aware of how people feel in order to create experiences that will be more natural, engaging, entertaining, and productive. This paper describes a emotionally responsive robotic dog. The dog can tell how its user feels by listening to sounds and then acts in response. It explains how the acoustic emotion recognition works and how it is used inside the robotic dog.

This project is relevant because it is centered on developing an interactive robot that is low cost with offline control. In recent years, there has been a rising need for robotic companions and intelligent assistants. The problem with most modern robots is that they are costly to implement. The use of EchoPaw robot is a sensible alternative because it employs components that are readily available, such as ESP32, servos and sensors.

Author: Egor Lakomkin, M.A. Zamani, Cornelius Weber, Sven Magg, Stefan Wermter Abstract and relevance: The area of Speech Emotion Recognition (SER) is critical in human-robot collaboration. Furthermore, SER has attracted a lot of research interest. This paper investigates how noise from different real-world environments affects advanced neural networks used for SER, specifically through use of the iCub Robot. The study concluded that both robot ego noise and room acoustics had a negative effect on the accuracy of SER models. In addition, the study showed that the application of data augmentation

The implementation of offline voice recognition means that the robot is reliable even when there is no internet connection. This is particularly useful in a classroom setting, a laboratory or even indoors. The project is useful for

Impact Factor value: 8.315

ISO 9001:2008 Certified Journal

Page 521