Innovations in science and technology, particularly in the area of electronics, have had a significant impact on people's lives in today's rapidly evolving world. Robots are a well-known example of how technological advancements have enhanced and streamlined our way of life. Robots are designed to assist humans with their daily tasks in an effort to make work easier and more efficient. One such chore that people usually put off since it is tedious and exhausting is floor cleaning. The issue is that cleaning floors is a labor-intensive task because buildings have enormous floor spans. Brooms and mops are used in traditional cleaning methods, which have been demonstrated to be time-consuming. Furthermore, dust is unintentionally dispersed by these techniques. To overcome this problem a novel approach on cleaning robot Using Information of Things (IOT) based on Deep Reinforcement Learning algorithms (DRL) in introduced. The goal of the proposed work is to develop an inexpensive robot for floor cleaning. Deep Reinforcement Learning algorithm (DRL) was introduced for identifying and detecting the obstacles on the floor. The robot identifies the obstacles on the floor by using sensors. Various sensors are deployed for identifying obstacles on the floor like ultrasonic sensor, LIDAR sensor on floor cleaning robot. The sensor is used to identify, collect and deposit the trash and accelerometer. Deep reinforcement learning is combined with deep learning algorithm for training the floor cleaning robot. The proposed work DRL achieves about 2.04% and 3.06% than MCDM and SWCVAE algorithms in accuracy.