Performance Analysis Internet of Things Based on Sensor and Data Analytics

Abstract

The Internet of Things (IoT) is a transformative development. Sensors are known in scientific science as a prospective field. IoT sensors are used successfully to develop an intelligent world in different IoT implementations. To make a familiar operating image, ubiquitous sensing abilities provide shared knowledge. IoT systems can fail if one of the sensors stops working due to any failure, which causes serious living problems to some families. Thus, the problem consists of predicting when the subsequent loss will occur. A two-phase approach is proposed. First, an exploration of the provided data will take place. The purpose of such analysis is threefold: (1) to get familiar with the data, (2) to assess the quality of data and decide which methodology to employ accordingly, (3) to determine the features of interest. This research will serve as the basis for further theoretical work in the same field in the prospect. Analyzing these sensor-derived data is an important task that can find valuable latent information in addition to the data itself. Since the Internet of Things includes some sensors, the measurement data obtained by these sensors are multi-type data, often including information from the time series. Depending on the configuration flow rate, vibration, impeller speed, and even temperature, the sensors with solid correlation can be closely correlated with some delay involved. Data ranges investigation was done before correlation check. Results indicate that our procedure is accurate for the machine's regular, damaged, and recovering state.