An IoT Platform with Monitoring Robot Applying CNN-Based Context-Aware Learning

Area of Science:

• Robotics and Automation
• Artificial Intelligence
• Internet of Things (IoT)

Background:

• Traditional closed-circuit television (CCTV) systems are limited by fixed installations.
• There is a growing need for dynamic and intelligent surveillance solutions.
• Internet of Things (IoT) technology offers new possibilities for connected and autonomous systems.

Purpose of the Study:

• To propose an IoT platform with an intelligent surveillance robot as an active CCTV alternative.
• To develop a machine learning model for enhanced image and context recognition in surveillance.
• To enable smart monitoring and remote control for improved security and convenience.

Main Methods:

• Development of an IoT platform integrating an intelligent surveillance robot with an onboard IoT server.
• Implementation of sensors for line tracing and contextual information acquisition.
• Application of Convolutional Neural Network (CNN)-based machine learning for precise image detection and recognition.
• Design and training of a CNN model using a dataset of images acquired from the robot.

Main Results:

• The intelligent surveillance robot can perform line tracing and detect abnormal environmental statuses.
• The system achieved a learning accuracy of over 0.98 in image context recognition using the designed CNN model.
• An alarm is sent to a manager's smartphone upon detection of an abnormal status.
• A client application allows for remote robot control, enhancing user convenience.

Conclusions:

• The proposed IoT platform with an intelligent surveillance robot effectively overcomes the limitations of fixed CCTV systems.
• The developed CNN model significantly enhances the image and context-aware learning capabilities of the IoT platform.
• The system demonstrates high potential for detecting abnormal situations in diverse industrial fields, including factories, smart farms, logistics warehouses, and public spaces.