“Little Booker” AI Study Companion Robot

Project Introduction

With the development of computer technologies such as machine learning, service-oriented robots are gradually entering people's lives. Considering the current stage of educational robots - still in their early stages of development, with relatively limited functionality and types but high demand. We proposed a reading companion robot, named "Little Booker" which integrates computer vision and voice technologies.

This project primarily targets challenges faced by adolescent, such as heavy academic burdens, concentration difficulties and poor learning postures. The robot uses computer vision tech to monitor student's attention states, fatigue level and learning posture and employs voice interaction technology to provide reminders and corrections. Additionally, addressing the issue of busy parents lacking time for supervision, the robot will deploy communication channel to enable remote voice reminders and video monitoring.

Through these features, the "Little Booker" aims to help teenagers cultivate healthy study habits, improve learning efficiency, and safeguard their physical and mental well-being.

System Overview

Hardware Design

The robot adopts a customized mechanical design inspired by the Minions animated characters.

The core processing unit is built on the Raspberry Pi 3B+ platform, with peripheral devices including:CHI camera module for visual data acquisition, 3.5mm audio speaker for voice output, USB microphone array for speech input and AC adapter for power source.

The hardware architecture is illustrated in the following diagram.

Software Architecture

The system mainly comprises four modules.

• Computer Vision Module (OpenCV-based)
  Monitors children's learning status such as fatigue, focus, and posture by analyzing facial expressions (e.g., yawning), eye features (e.g., blink frequency), and head movements (e.g., head tilt angle).
• Voice Interaction Module (Baidu AI Platform API)
  Integrates TTS (Text-to-Speech), ASR (Automatic Speech Recognition), and NLP (Natural Language Processing) functionalities via Baidu's developer API to enable conversational interactions and voice reminders.
• Mobile App Interface Module
  Handles remote app interactions, including Real-time video streaming uploads and voice command transmission.
• Central Logic Controller
  Generates reminder messages based on vision detection results and manages the workflow of the above modules.

The modular framework is structured as follows.

Demonstration

During the mid-term project review, we independently completed the basic development and testing of the vision and voice modules. The results are shown in the figure below.

Achievements

During the mid-term review of the College Students Innovation and Entrepreneurship Training Program in Spring 2018 at the School of Aerospace, our project received a score of 93.3, ranking 2nd out of 45 projects , and was recommended as the 8th best project out of 37 projects university-wide for national-level project applications.

Regrettably, despite these early accomplishments, the project was eventually discontinued due to academic commitments that prevented team members from sustaining development efforts..

Acknowledgement

Special thanks to WANG Yanbo (Project Director) and LI Jialun (Computer Vision Lead) for their dedicated hard work throughout the project. This project was also a journey of collective growth for everyone involved.