TEMPERATURE SENSOR DATA QUALITY ASSESSMENT IN MANUFACTURING ENVIRONMENT USING HAMPEL FILTER AND QSD

Abstract

In the Industry 4.0 era, integrated temperature sensors in system production become source main data for taking decisions. However, the quality of the data produced often influenced by noise, missing values, and disturbing anomalies accuracy of analytical processes. Research This proposes a monitoring pipeline designed data quality For environment manufacturing based on the Internet of Things (IoT), with focus on usage Hampel Filter and Quality Score Delta (QSD) methods. Hampel Filter is used for detecting and handling outliers in temperature data in a way adaptive, while QSD is used for measure dynamics change data quality from time to time. Architecture system built with using Apache Kafka for data ingestion, InfluxDB For time-series storage, and Grafana for real-time visualization. Case study performed on temperature sensor data from the conveyor motor, and the results show that method. This capable detect degradation data quality in general proactive. Findings show potential big in increase reliability industrial monitoring system as well as support maintenance predictive data- based. Research This give contribution significant in developing modular and adaptive approach for management data quality in the manufacturing sector.