JINKOBOT

Therapeutic Robot for Children with Autism Spectrum Disorder (ASD)

Bachelor's Thesis Project (Honors Distinction) by Diana Hernandez Soler
Bachelor's Degree in Interactive Technologies
Universidad Politécnica de Valencia
Gandía Campus · September 2022

📋 About the Project

Jinkobot is a ROS-based therapeutic robot designed to assist therapists working with children with Autism Spectrum Disorder (ASD). The robot guides children through therapeutic exercises and routines that have been previously prepared by psychologists or therapists through a web application. The system collects objective data during therapy sessions—such as time spent on each task, number of correct answers, and unanswered questions—allowing therapists to track progress over time through detailed analytics and graphs in the web application. The goal of this project is to help both children and therapists during therapy sessions by collecting and processing objective data that would otherwise be much more complex to monitor manually.

• Keywords: Robotics, ASD, Autism, ROS, Python, Firebase, AWS, Real-time data processing

🛠 Technologies Used

• ROS (Robot Operating System): Robot control and navigation
• Python: State machine, image recognition, real-time processing
• TurtleBot Platform: Physical robot hardware
• Firebase: Real-time database and web application backend
• AWS Polly: Text-to-speech for robot interaction
• Computer Vision: Real-time image recognition for therapeutic exercises
• Autonomous Navigation: ROS-based navigation system

🎯 Key Features

• Autonomous Navigation: The robot navigates independently to guide children through different therapy stations
• Real-time Image Recognition: Uses computer vision to verify children's responses in exercises
• Web Application: Therapists can prepare exercises and monitor progress through an intuitive interface
• Data Analytics: Automatic collection and visualization of therapy session data
• Interactive Communication: Text-to-speech integration for natural robot-child interaction

🔗 Project Repositories

This project consists of multiple interconnected repositories:

• State Machine (Python) - Core robot behavior and exercise coordination
• Emotion Exercise: Real-time Image Recognition - Computer vision module for emotion recognition exercises
• Orientation Exercise: Autonomous Navigation with ROS - ROS-based navigation for spatial orientation tasks
• Web Application - Firebase-based therapist interface

🎥 Demo

Video demonstration (Spanish)

🏗️ System Architecture

The system integrates several technologies:

1. ROS Layer: Controls the TurtleBot, manages navigation, and coordinates sensor data
2. Python Layer: Implements the state machine, processes real-time data, and handles exercise logic
3. Web Application: Firebase-based interface for therapists to configure exercises and view analytics
4. Cloud Services: AWS Polly for speech synthesis
5. Hardware: TurtleBot platform with sensors (camera, IMU, proximity sensors)

📚 Academic Context

This project was developed as a Bachelor's thesis for the Interactive Technologies program at Universitat Politècnica de València (UPV), earning an Honors Distinction for its technical complexity and real-world application in assistive therapy.

👩‍💻 Author

Diana Hernández Soler

• GitHub: @dianait
• LinkedIn: linkedin.com/in/dianait
• Blog: dianait.blog

📄 License

This project was developed for academic purposes as part of a Bachelor's thesis at UPV.