International Journal of Scientific Research and Engineering Development

As industrial data has expanded, monitoring systems have become crucial for management decisionmaking. Contemporary technologies, like sensors derived from the IoT, might aid in monitoring the production process. This paper suggests a system for ongoing monitoring that makes use of IoT sensors, big data processing techniques, and a hybrid predictive model. A sensor for the Internet of Things was created to gauge temperature, humidity, accelerometer data, and gyroscope information. The sensor data generated by IoT in factories is real-time, massive in volume, and unstructured. Apache Kafka is used for alerts, Apache Storm for real-time analysis, and MongoDB for storing factory sensor data on the large-scale data analysis platform. Secondly, the hybrid prediction model combined outlier identification based on DBSCAN with categorization using Random Forest to eliminate outlier sensor data and identify manufacturing defects. The vehicle was tested on a Korean car production line. The findings demonstrated that an efficient monitoring of production can be achieved through the use of IoT-based sensors in conjunction with the proposed big data processing framework. When it comes to fault prediction, the hybrid model that makes use of sensor data provides greater accuracy than alternative models. The proposed system is anticipated to improve managerial decision-making and reduce losses linked to manufacturing errors.

Vennila P, Maniraj V, "Optimization of Processes Through Machine Learning and Data Acquisition Based on IOT" International Journal of Scientific Research and Engineering Development, V9(2): Page(92-96) Mar-Apr 2026. ISSN: 2581-7175. www.ijsred.com. Published by Scientific and Academic Research Publishing.