1964–1975

State-Space Process Control

Contemporary themes

The period from 1964 to 1975 marked a decisive shift from empirical tuning to theory-guided design, with state-space representations, optimal control, and modal analysis becoming central to chemical, electrical, and industrial processes. Real-time implementation and early computer-aided control connected theoretical developments with practice, enabling model-based control and rapid deployment across plant operations. Decomposition techniques and modal analysis supported management of large-plant and inventory contexts, while process safety and design-oriented control methods gained prominence in chemical and process industries, including heating and cooling applications, and electrical drives provided tangible demonstrations of robust and optimal control.

• Emergence of rigorous mathematical control frameworks—state-space, optimal control, modal analysis—being applied across chemical, electrical, and industrial processes, signaling a shift from ad hoc tuning to theory-guided design[7][8][13][16][17].
• Early integration of computational methods and real-time implementation, with model-based process control, real-time programming languages, and remote/industrial process software bridging theory and practice[1][9][19][20].
• Modal analysis and decomposition-driven strategies to manage complex plants and production systems, including large-plant control and inventory contexts[3][4][13].
• Process control in chemical and industrial settings emphasizing theory-to-practice linkage, process safety, and design-oriented control methodologies across chemical processes and heating/cooling applications[2][6][10][11][12].
• Electrical machines and drives as concrete control exemplars, illustrating robust and optimal control for synchronous machines and their outputs[5][16][17].

Influential works

The emergence of structured frameworks for linking dynamic process models with feedback design established the foundation for modern process control. Early adaptive control concepts and the integration of state feedback and observers shaped education and industry practice for decades. The period solidified the role of model-based design as a standard approach, paving the way for nonlinear analysis, robust control, and the broader adoption of state-space and model-based methodologies in subsequent eras.

• Process Systems Analysis and Control (1964) introduced a structured framework linking dynamic process models with feedback control design, enabling stability analysis and model-based optimization across unit operations; it catalyzed modern process control methods and education[7].
• Adaptive Control Processes (1964) presented early adaptive control concepts with parameter adjustment in real time, paving self-tuning controls and robust design; influenced later model-reference and PID adaptation strategies in industry[9].
• Modern Control Engineering (1971) codified state-space and modern design for process systems, integrating time-domain and frequency-domain tools, bringing state feedback and observers to routine practice and shaping curricula for decades[1].
• Oscillations in chemical systems. IV. Limit cycle behavior in a model of a real chemical reaction (1974) exposed nonlinear dynamics and limit cycles in reactors, prompting robust control strategies and nonlinear analysis that informed later control theory[8].

1976–1982

Integrated Model-Based Process Control

Contemporary themes

The period witnessed an emergence of integrated production-control ecosystems linking decision-support, architectural design, and management practices, moving from isolated control loops to systemic production governance. Formal modeling, specification, and optimization underpinned rigorous design of processes and control architectures, guiding regulator design and adaptive control. Real-time, microcomputer-enabled control rose to prominence in industrial settings, enabling digital control of fermentors, dissolved-oxygen systems, wastewater treatment, and autonomous microprocessor-based controllers, while process and energy-system optimization shaped performance metrics.

• Emergence of integrated production-control ecosystems combining decision-support, architectural design, and management practices, reflecting a shift from isolated control loops to systemic production governance[1][10][12][14][15].
• Formal modeling, specification, and optimization underpin a shift to rigorous design of processes and control architectures, using system specification languages and production-function formalisms to guide regulator design and adaptive control[5][6][13][20].
• Growing emphasis on real-time and microcomputer-enabled control for industrial processes, including digital control of fermentors, dissolved-oxygen systems, wastewater treatment, and autonomous microprocessor-based controllers[7][11][16][18][19].
• Process and energy-system optimization within chemical and energy domains, combining process safety, extrusion control, and solar power plant dynamics to shape control strategies and performance metrics[2][3][8][17].

Influential works

These developments established foundational methodologies for model-based design, multivariable control, and identifiability in closed-loop settings, catalyzing the shift toward predictive and adaptive control across industries.

• Process Dynamics and Control (1978) introduced a comprehensive synthesis of dynamic modeling and feedback design for chemical processes, standardizing transfer-function and state-space methods and catalyzing model-based control in industry[2].
• The application of fuzzy control systems to industrial processes (1977) advanced fuzzy logic to real processes, proposing rule-based control with linguistic variables to handle uncertainty and nonlinearities, spurring decades of fuzzy and adaptive controllers[5].
• Process Analysis (1981) offered a rigorous, interdisciplinary framework for analyzing process dynamics and control loops, promoting systematic stability, robustness concepts, and performance metrics that informed both textbooks and software tools[1].
• Advanced process control (1982) consolidated APC concepts, highlighting multivariable control, predictive strategies, and optimization-driven approaches, accelerating the shift from single-loop controllers to integrated plant-wide control[6].
• Identification of processes in closed loop—identifiability and accuracy aspects (1977) tackled identifiability in closed-loop identification, revealing how feedback affects parameter estimation and accuracy, shaping best practices for system identification in practice[11].

1983–1989

Robust Multivariable Process Control

Contemporary themes

During 1983–1989, process control research centered on robustness and multivariable, model-based strategies, increasingly capable of handling uncertainty through predictive control and constrained optimization. Real-time fault detection, disturbance analysis, and online parameter estimation emerged as essential tools for stability, safety, and performance in processing systems. Integrated planning and control for manufacturing linked process models with production scheduling and design flow, while adaptive and cognitive methods began to inform learning-based control.

• Model-based control and optimization matured to handle uncertainty, integrating multivariable modeling, constrained optimization, and predictive control to improve dynamic performance across chemical and industrial processes[12][13][14][15][18].
• Real-time fault detection, disturbance analysis, and online parameter estimation emerged as essential for stability, safety, and performance in processing systems[9][10][11][20].
• Integrated planning and control for manufacturing systems and industrial product development, linking process modelling with production scheduling and design flow[5][8][16].
• Adaptive and cognitive approaches shaped process control through reasoning about learning, adaptive control, and empirical insights from workshops and studies[17][19][20].

Influential works

These developments established a unified paradigm of robust multivariable control that guided both theory and practice, introducing foundational concepts in multivariable design, robustness under uncertainty, and integration with quality and planning systems, shaping subsequent research through the late 1980s and beyond.

• Robust process control (1988)[3] introduced a systematic focus on maintaining process performance under model and disturbance uncertainties, laying foundational ideas for robustness in process industries and shaping later robust-control frameworks.
• Principles and Practice of Automatic Process Control (1985)[5] offered a comprehensive bridge between control theory and chemical/process engineering practice, codifying design methodologies and benchmarking practices that informed both education and industrial implementations.
• Linear Multivariable Control (1985)[14] pioneered MIMO control concepts for processes with multiple inputs and outputs, establishing foundational design strategies and stability insights that influenced subsequent multi-variable control research.
• A New Measure of Process Capability: Cp (1988)[1] formalized Cp as a quantitative metric for process capability, integrating statistical quality control with manufacturing performance, catalyzing SPC adoption and quality improvement programs.

1990–2009

MPC-Driven Robust Control

Contemporary themes

Model Predictive Control and robust control frameworks dominated process design, linking predictive modeling with real-time optimization and resilient performance. Data-driven, online monitoring became foundational across batch and continuous processes, employing multivariate tools such as multiway PCA/PLS, ICA, and latent-variable models to reveal structure, variability, and variable-contribution patterns in real time. Plant-wide coordination and holistic optimization of production networks emerged as a key priority for large-scale operations.

• Data-driven, multivariate monitoring and diagnosis became foundational across batch and continuous processes, integrating multiway PCA/PLS, ICA, and latent-variable models to reveal structure, variability, and variable-contribution patterns in real time.[1][2][3][16][4][17][15][18]
• Model-based control and predictive strategies dominate process industry design, emphasizing model predictive control, robust control, self-optimizing systems, and plant-wide coordination to improve production efficiency and safety.[5][9][19][7][12][13][14][8]
• On-line and dynamic data modeling for batch processes focuses on real-time updates via dynamic PCA/PLS, multiway dynamic data, and online monitoring to support startups, grade transitions, and ongoing process supervision.[18][16][17][4][1]
• Process modeling, simulation, and formal representations provide the theoretical scaffolding for process design, planning, and analysis, enabling systematic reasoning about evolution, trends, and control needs.[8][13][11][12]
• Plant-wide control and industrial process coordination emphasize holistic optimization of production networks and system-wide control architectures for large-scale operations.[7][19][12][5]

Influential works

Foundational advancements introduced robust process control concepts, formalizing design approaches that withstand model uncertainty and disturbances. Linear matrix inequality techniques for H-infinity control enabled convex optimization to guarantee robust performance, catalyzing widespread adoption in practice. Model Predictive Control matured into a comprehensive framework for dynamic optimization, with retrospective analyses that solidified its status as a dominant paradigm for process control in industry and academia.

• Robust process control. (1990) introduced systematic robustness concepts for chemical process control, outlining design strategies to withstand model uncertainty and disturbances, laying the groundwork for later H∞ and robust control in process industries[7].
• A linear matrix inequality approach to H∞ control (1994) presents an LMI-based formulation for H∞ controller design, enabling convex optimization to guarantee robust performance; it catalyzed broad adoption of LMI techniques in process control research and practice[8].
• Model Predictive Control (MPC) (2007) articulates the core framework for dynamic optimization of process systems under constraints, linking predictive models with real-time optimization and control; it became a standard tool in industry and academia[6].
• Model predictive control: past, present and future (1999) provides a comprehensive retrospective of MPC developments, comparing formulations, constraints handling, and implementation challenges; it solidified MPC's status as a dominant paradigm in process control[12].

2010–2016

Data-Driven Model Predictive Control

Contemporary themes

During this period, data-driven analytics became the central engine for process control, unifying monitoring and diagnosis across batch and continuous operations through kernel principal component analysis, hidden Markov models, particle filters, Gaussian mixtures, and process mining, complemented by reviews on data-based monitoring and data-driven control. Model predictive control with scheduling, design, and uncertainty management emerged as the core paradigm for batch and multiproduct processes, supported by multi-parametric MPC, adaptive neural networks, and design-time simulation, along with large-scale system design strategies. Progress in process modeling, simulation, and model updating laid the groundwork for accurate control across large-scale chemical systems, connecting process dynamics, simulation, model repair, and distillation design/control simulation, while domain-specific advances in laser-based manufacturing demonstrated advanced monitoring and fault detection.

• Data-driven monitoring, fault diagnosis, and information fusion unify process analytics across batch and continuous operations using kernel principal component analysis[2], hidden Markov models[13], particle filters[14], Gaussian mixtures[12], and process mining[16], with reviews of data-based monitoring[11] and data-driven control[19].
• Model predictive control (MPC) with scheduling, design, and uncertainty management emerges as a core paradigm for batch and multiproduct processes, integrating multi-parametric MPC[20], adaptive neural networks[7], and design-time simulation[9], complemented by large-scale system design strategies[17].
• Process modeling, simulation, and model updating (repairing models, reflecting reality) provide foundational accuracy for control and design across large-scale systems and chemical processes, connecting process dynamics[5], simulation[6], model repair[4], and distillation design/control simulation[9].
• Advanced manufacturing monitoring and safety-oriented fault detection in laser-based processes demonstrates domain-specific monitoring and optical measurements, including laser melting failure detection[15], influential process monitoring in laser-assisted manufacturing[1], and fault diagnosis via dynamic monitoring methods[14].

Influential works

These developments established a shift toward data-centric, predictive control frameworks that informed both education and industry practice, with AI-augmented control becoming pervasive and laying the foundation for subsequent hybrid physics-data models. The period solidified Model Predictive Control as a mainstream engineering tool for constrained, multivariable processes and multiproduct operations.

• Model Predictive Control (MPC) (2012)[15] introduced the optimization-based framework for constrained, multivariable process control, predicting future trajectories to satisfy constraints while minimizing cost. This formalism rapidly spread to industry and became a teaching pillar, shaping subsequent research in constrained control.
• Process Dynamics and Control (2011)[10] unified dynamic modeling, control design, and stability analysis for chemical/process engineers. As a foundational textbook, it trained generations of practitioners and scholars, anchoring modern process control education and practice worldwide.
• A Combined Adaptive Neural Network and Nonlinear Model Predictive Control for Multirate Networked Industrial Process Control (2016)[11] couples neural network–assisted process modeling with nonlinear MPC to handle multirate, networked dynamics. It pioneered AI-augmented MPC concepts and influenced later networked-control research.
• Survey on data-driven industrial process monitoring and diagnosis (2012)[5] surveys data-driven fault detection and monitoring methods for process industries, highlighting pattern recognition, statistical process control, and data mining. It catalyzed a data-centric shift in monitoring and informed subsequent data-based process diagnosis work.

2017–2023

Digital Twin Process Control

Contemporary themes

Digital twin–driven, data-rich frameworks unify production management, process data analytics, and real-time control across factory floors, enabling closed-loop optimization in Industry 4.0 contexts. Predictive and prognosis-driven monitoring emphasizes early fault detection, trend-based forecasting, and explainability to support trust and actionable insights in process control. In-process sensing and control for additive manufacturing leverage characterizations of melt pool, plume, spatter, and nondestructive testing data to detect defects and adjust process parameters, enabling improved quality assurance in AM manufacturing. Hybrid control schemes combine predictive control with iterative learning and fault tolerance to handle batch process uncertainties, time-delays, and nonstationary dynamics within constrained control objectives. Nonstationary process monitoring leverages slow feature analysis and cointegration techniques to extract slowly varying patterns, enabling robust adaptive monitoring and diagnosis under changing conditions.

• Digital twin-inspired, data-driven frameworks unify production management, process data analytics, and real-time control across assembly and manufacturing floors, enabling closed-loop optimization and proactive decision-making in Industry 4.0 contexts[3][9][14][20].
• A shift toward predictive and prognosis-driven process monitoring emphasizes early fault detection, trend-based forecasting, and explainability to support trust and actionable insights in process control[3][10][14][19].
• In-process sensing and control for additive manufacturing leverage characterizations of melt pool, plume, spatter, and nondestructive testing data to detect defects and adjust process parameters, enabling improved quality assurance in AM manufacturing[6][7][8][12].
• Hybrid control schemes combine predictive control with iterative learning and fault tolerance to handle batch process uncertainties, time-delays, and nonstationary dynamics within constrained control objectives[15][16][17].
• Nonstationary process monitoring leverages slow feature analysis and cointegration techniques to extract slowly varying patterns, enabling robust adaptive monitoring and diagnosis under changing conditions[1][4][13][18].

Influential works

Foundational works formalized safety and constraint handling through barrier functions; big-data process monitoring frameworks for Industry 4.0, and digital twin-based production management and control for complex shop floors; additive manufacturing process monitoring and control by non-destructive testing techniques established in-process sensing as a basis for feedback control and quality assurance.

• Control Barrier Functions: Theory and Applications (2019)[7] formalizes barrier functions to enforce state constraints, enabling safety guarantees in nonlinear control, robotics, and CPS, and shaping subsequent safety-oriented control design.
• Industrial Process Monitoring in the Big Data/Industry 4.0 Era: from Detection, to Diagnosis, to Prognosis (2017)[5] provides a structured framework for real-time monitoring, fault diagnosis and predictive prognosis using big data in process industries, catalyzing data-driven maintenance.
• Digital twin-based smart production management and control framework for the complex product assembly shop-floor (2018)[4] introduces digital twins to unify design, simulation and control for complex shop floors, accelerating Industry 4.0 adoption.
• Additive manufacturing process monitoring and control by non-destructive testing techniques: challenges and in-process monitoring (2017)[14] surveys NDT-based in-process sensing to guide feedback control, elevating quality assurance in AM processes.