Real-time decision support systems in chemical and process engineering

Abstract

Artificial intelligence (AI)-driven decision support systems (DSS) are transforming Industry 4.0 by integrating internet of things (IoT), big data analytics, and cyber-physical systems into manufacturing. This research explores data acquisition and integration within DSS for chemical engineering, emphasizing advanced sensor technologies, data management, and pre-processing methods to refine raw data for meaningful analysis. The study examines various decision models and optimization strategies, including real-time optimization, machine learning (ML) models, and simulation-based decision support. Additionally, it underscores the significance of user interface design and visualization tools, ensuring seamless integration with process control and automation. Industrial sensors and IoT devices collect extensive process data, including temperature, pressure, flow rates, and chemical composition. Near-sensor and in-sensor computing enhance efficiency by minimizing data transfer. Data pre-processing ensures accuracy and consistency. Technologies like MapReduce and stream processing frameworks (Apache Kafka, Flink, and Spark Streaming) facilitate real-time analytics. ML techniques, including artificial neural networks, reinforcement learning, support vector machines, linear regression, and decision tree, are widely applied in predictive maintenance. Manufacturing simulations utilize discrete event modeling and Monte Carlo methods for optimization. Process control systems, distributed control systems, and supervisory control and data acquisition play vital roles in automation and monitoring. However, challenges like network delays, system load, and algorithm complexity can impact real-time DSS efficiency. Developing and maintaining these systems demand advanced hardware and continuous updates. © 2026 Elsevier Inc. All rights reserved.