Advanced Deep Learning Autonomous Robot For Fire Fighting and Surveillance

Abstract

Advancement in deep learning resulted in tremendous research in the field of robotics over the last decades. Deep Learning utilizes artificial neural networks and is an intelligence function that behaves like the human brain while making a decision. This paper presents an autonomous, economical, and efficient service robotic system and its application in firefighting and surveillance in the real world. The system comprises a raspberry pi camera and fire-detecting algorithm based on the HSV color range to detect the desired fire at the very far distance. Moreover, the system can automatically detect and track the fired object and follow it and extinguish it. The result shows the system is capable of providing robust surveillance and reliability as it can track living and non-living objects distinctly in prohibited areas without human intervention.