Real-Time Big Data Processing and Accurate Production Analysis in Smart Job Shops Using LSTM/GRU and RPA

Abstract

This paper explores the integration of Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) neural networks with Robotic Process Automation (RPA) for real-time big data processing in smart job shops.
Objectives: This include enhancing real-time data processing, automating production monitoring, optimizing production schedules, enabling predictive maintenance, and improving overall manufacturing efficiency.
Methods: This involve collecting real-time data from IoT devices, preprocessing it for LSTM/GRU models, and applying RPA to automate repetitive tasks. The integrated system predicts equipment performance, optimizes schedules, and reduces downtime.
Results: This demonstrates significant improvements, including an 8.2% reduction in downtime, a 0.837 increase in production efficiency, and enhanced predictive accuracy at 0.89.
Conclusion: This indicates that the proposed method effectively boosts decision-making processes, minimizes operational disruptions, and increases manufacturing productivity, making it a powerful tool for smart job shops in the Industry 4.0 era.