An AI-Powered System for Encouraging Cleanliness and Managing Screen Time

Abstract

This article introduces Clean Room Monitor, an AI-powered solution created to automate the supervision and management of children’s room tidiness. The system uses an IP Webcam to capture images and employs a Convolutional Neural Network (CNN) to categorize spaces as neat or cluttered following uniform preprocessing. When a messy room is identified, automatic responses, like SMS/Email notifications and router-imposed internet blocks are activated to maintain impartial parental control. A structured logging module records all predictions for analytics and periodic model retraining. Experimental results demonstrate high classification accuracy and reliable real-time performance, indicating that the system effectively reduces parental supervision effort while promoting responsible behaviour. The proposed solution is low-cost, scalable, and suitable for integration into smart home and digital parenting applications.

Introduction

Maintaining tidy rooms is important for discipline and healthy habits, but traditional parental oversight is subjective, inconsistent, and labor-intensive. Existing parental control apps focus mainly on screen time and device usage, neglecting real-world behavioral aspects like room cleanliness.

The proposed Clean Room Monitor leverages AI and IoT to objectively assess room tidiness and enforce rules automatically. Images from an IP Webcam are preprocessed and classified as clean or messy using a CNN. The system then triggers automated actions—SMS/Email alerts and router-based internet restrictions—based on cleanliness detection. All events are logged, and the model is periodically retrained to adapt to changes in lighting, room configuration, or behavior.

The methodology integrates image preprocessing, CNN-based classification, IoT decision automation, and continuous learning, providing a scalable, unbiased, and real-time monitoring solution. The backend, built with Python and Flask, coordinates prediction, notifications, and router control, while a dashboard enables parents to monitor status, view images, and manage settings.

Results show that the system reliably detects clean versus messy rooms under varied lighting and clutter conditions, effectively automates parental interventions, and provides consistent, objective monitoring, enhancing accountability and reducing manual effort in digital parenting.

Conclusion

The Clean Room Monitor system introduced in this study offers an automated and smart solution for evaluating room tidiness through AI-powered image classification and IoT-enabled action management. By combining a CNN model with a dependable preprocessing work-flow the system delivers consistent and impartial identification of tidy and untidy settings removing the bias and variability linked to manual monitoring. The automated enforcement feature which involves SMS/Email alerts and router-level internet restrictions enhances control by guaranteeing prompt and fair reactions, to the state of the room. Ongoing logging and scheduled model retraining additionally improve the system’s flexibility and sustained effectiveness. Test outcomes verify that the Clean Room Monitor functions across various real-life scenarios while preserving excellent accuracy and quick response. The user-friendly dashboard and automated processes lessen effort and encourage improved child discipline via data-informed reinforcement. Moving forward the platform can be enhanced with object- detection, cloud analytics and compatibility, with voice assistants to boost accuracy and ease of use. Overall, the proposed solution demonstrates strong potential as a scalable, cost-effective, and intelligent tool for modern digital parenting and smart home management.

References

[1] S. Rattanaprathum, P. Kaena, and S. Hunta, “Object detection in smart home system for visually impaired person,” Journal of Spatial Innovation Development, vol. 12, no. 2, pp. 45–59, 2023. [2] D. Mereddy et al., “Vision-Based Automatic Groceries Tracking System — Smart Homes,” Vision Systems Research, vol. 15, no. 4, pp. 110–124, 2023. [3] Anonymous, “A framework of context-aware object recognition for smart home,” in Proc. 5th Int. Conf. Smart Homes & Health Telematics, 2010, pp. 112–125. [4] C. Wu and H. Aghajan, “Recognizing objects in smart homes based on human interaction,” in Advanced Concepts for Intelligent Vision Systems (ACIVS 2010), vol. 6475, Springer, 2010, pp. 131–142. [5] R. Bagade and K. Killiketar, “Object detection for indoor localization system,” International Journal of Innovative Science and Research Technology, vol. 9, no. 9, pp. 200–212, 2024. [6] J. Smith, A. Patel, and T. Lee, “Scene and object recognition fused framework for indoor smart home automation,” Journal of Ambient Intelligence and Humanized Computing, vol. 15, no. 7, pp. 3450–3465, 2024. [7] M. Akhila, E. K. Adhil, S. Amay, U. Rithil, and A. Prasad, “Autonomous Room Sanitizing Machine,” International Journal of Engineering Research & Technology, vol. 10, no. 4, pp. 150–157, 2022. [8] S. Hizal, Ü. Çavu?o?lu, and D. Akgün, “IoT-Based Smart Home Security System with Machine Learning Models,” Academic Platform Journal of Engineering and Smart Systems, vol. 1, no. 1, pp. 28–36, 2024. [9] A. R. Javariya and H. Lohiya, “Deep Learning Model-Based Intelligent Smart Home Control and Security System,” International Journal of Computer Engineering and Technology, vol. 13, no. 1, pp. 53–64, 2022. [10] “Enhancing Smart Home Security Using Deep Convolutional Neural Networks and Multiple Cameras,” Wireless Personal Communications, vol. 132, no. 5, pp. 4057–4075, 2024.

Copyright

Copyright © 2025 Abhishek N, Tejas B N. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.