Modeling in the Era of AI-Driven Industrial Automation and Optimization

Abstract

The growing influence of artificial intelligence (AI) is reshaping process systems engineering (PSE) and industrial modeling and optimization. While data-driven methods excel in predictive maintenance, anomaly detection, and pattern recognition, they still face challenges in safety-critical, data-scarce, and extrapolation-prone environments. This paper argues that first-principles models (FPMs) (rooted in fundamental physics, chemistry, and engineering) remain essential for mission- and business-critical workflows in industrial automation, both in process design and operations. We highlight the enduring strengths of first-principles and examine hybrid paradigms that combine mechanistic rigor with data-driven machine learning (ML) and large language models (LLMs) to enhance adaptability and efficiency. Case studies across process design, advanced process control (APC), real-time optimization (RTO), production planning, production scheduling, and supply chain management illustrate the value of retaining first principles as a backbone for modeling and optimization. We conclude that the future lies in AI-enabled systems grounded in first principles and mathematical optimization, where hybrid intelligence integrates mechanistic rigor with data-driven insights to deliver smarter design, safer operations, and more sustainable processes.