Abstract
The novel paradigm of Internet of Things (IoT) is gaining recognition in the numerous scenarios promoting the pervasive presence of smart things around us through its application in various areas of society, which includes transportation, healthcare, industries, and agriculture. One more such application is in the smart office to monitor the health of devices via machine learning (ML) that makes the equipment more efficient by allowing real-time monitoring of their health. It guarantees indoor comfort as per the user's satisfaction as it emphasizes on fault prediction in real-life devices. Early identification of various types of faults in IoT devices is the key requirement in smart offices. IoT devices are becoming ubiquitous and provide an assistant to supervise an office that is regulated by ML and data received from sensors is stored in cloud. A recommender system facilitates the selection of an appropriate solution for faults in IoT-enabled devices to mitigate faults. The architecture proposed in this paper is used to monitor each and every office appliance connected via IoT technology using ML technique, and recommender system is used to recommend solutions for fault patterns without much human intervention. The ultrasonic motion sensor is used to fetch the information of employee availability in cubicles and data is sent to the cloud through the WiFi module. ATmega8 is used to control electrical appliances in the office environment. The significance of this work is to forecast the faults in IoT appliances which will have an impact on life and reliability of IoT appliances. The main objective is to design a prototype of a smart office using IoT that can control and automate workplace devices and forecast whether the device needs repairing or replacing, thus reducing the overall burden on the employee and helping out in increasing physical as well as mental health of the person.
Highlights
In addition to the previous, Mundada et al [40] emphasize on the software fault prediction technique which is based on an artificial neural network with a back-propagation learning algorithm
E proposed architecture fetches data from office appliances to enhance the performance of each appliance. e data being collected by IoT is sent to the machine learning (ML) algorithm for future fault prediction. is data stored in database is connected to a cloud server. e prediction and classification process of faults is done via an ML algorithm. en, this information is shared with the end-user as well as the solutions will be recommended using a mobile application. e mobile application with the end-user comprises software that keeps track of all devices and alerts the user on deviation or abnormal functioning of any appliance
Machine learning is a growing technology that plays a vital role in the advancement of the IT sector. e proposed architecture presents a prototype of a smart office that uses IoT, cloud, and ML technology for office people so that their efficiency can be increased
Summary
A comprehensive literature study has been carried out for fault prediction based on different ML techniques in IoT applications. In addition to the previous, Mundada et al [40] emphasize on the software fault prediction technique which is based on an artificial neural network with a back-propagation learning algorithm. Arun et al [45] presented a smart office system based upon the IoT application. In 2019, Xenakis et al [48] presented an IoT and cloud-based framework for fault prediction and machine condition monitoring for Industrial IoT. Ese technique improves the various performance parameters of the system like the reduction of noise, recovered lost data, increase prediction rate (using device location and state) to achieve better and faster results. Each keyword has a minimum frequency of 15 and the anomaly detection, sensors, deep learning are the latest keywords
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
