Near-infrared (NIR) sorting systems in waste sorting plants operate under multiple settings, creating distinct input-output relationships that challenge predictive modeling. Conventional neural networks, such as multilayer perceptron (MLP), often suffer from catastrophic forgetting under continual training, limiting reliability across settings. This study evaluates Kolmogorov-Arnold Networks (KAN) for continual regression modeling of multi-setting NIR systems. KAN assign nonlinear transformations to network edges using localized spline grids, enabling structural isolation between input regions. We introduce controlled input-space manipulations (shifting successive settings to adjacent or non-overlapping grid regions) and compare KAN performance with MLPs of comparable parameter count. We also examine single-input versus multi-input configurations to assess dimensionality effects. Results show that KANs with sufficient input-space separation maintain previously learned knowledge with pe... [more]

The pharmaceutical industry is placing growing emphasis on sophisticated process modeling to enhance the efficiency of drug design and production pipelines. Optimal control over these models can significantly improve manufacturing performance by lowering costs, boosting productivity, and ensuring rigorous quality compliance. However, the intricate nature and heavy computational load of these models often require the adoption of more practical or simplified alternative strategies for optimisation such as simulation-based approaches. In this work, we introduce a simulation-based framework including a "top-level" gradient-based mathematical programming optimisation model coupled with a "low-level" simulation scheme, to optimise multi-scale drug substance manufacturing flowsheets. The proposed framework optimises critical quality attributes, such as yield and purity, including green metrics such as process mass intensity, aligning with digital platforms (e.g. gPROMS) used in the pharmaceut... [more]

The global transition toward sustainable energy has intensified research into biofuels, with bioprocess optimization playing a central role in achieving decarbonization goals. Biobutanol, in particular, is a high-value molecule for sustainable fuel applications due to its superior energy density and compatibility with existing infrastructure. However, model-based optimization of its production is hindered by traditional semi-structured kinetic models that often suffer from limited predictive robustness. To address this challenge, within this study we developed a hybrid modeling framework for Clostridium saccharoperbutylacetonicum that integrates mechanistic mass-balance equations with Gaussian Processes (GPs) aiming to describe the biobutanol formation rate. Here, we investigate the effect of data normalization techniques on hybrid model's prediction capabilities comparing min-max normalization, z-score normalization, and no transformation. For each data treatment strategy, 8, 000 hybr... [more]

In pharmaceutical process engineering, accurate prediction of tablet properties is crucial for ensuring product quality, optimizing manufacturing efficiency, and advancing sustainable production practices. This study presents a physics-informed neural network (PINN) framework for predicting the physical properties of pharmaceutical tablets from near-infrared (NIR) spectra. The PINN framework integrates revised Kubelka-Munk theory and physical constraints to ensure physically consistent predictions while requiring less training data than conventional artificial neural networks. Tablets were manufactured using acetaminophen and microcrystalline cellulose formulations with varying compositions and compression settings. The PINN framework successfully predicts critical quality attributes, including tensile strength, porosity, and density. It offers a data-efficient, interpretable solution for pharmaceutical tablet quality control.

Mathematical models are increasingly used in pharmaceutical process development within quality-by-design (QbD) frameworks to reduce experimental effort and enable rational process design. However, model validation is still often based on deterministic performance indicators, which do not explicitly account for experimental variability, measurement noise, and model uncertainty. This work proposes an uncertainty-aware framework for model validation in pharmaceutical processes that quantifies predictive reliability in probabilistic terms, consistent with regulatory concepts. The framework explicitly integrates uncertainty in operating conditions, measurements, and model parameters, and evaluates model performance based on the probability that prediction satisfy predefined acceptance criteria rather than on single-point accuracy indicators. An in-silico case study of crystallization was performed to demonstrate the approach, where synthetic experimental data with controlled uncertainty wer... [more]

Maintenance is critical for industrial plants to ensure operational reliability and worker safety. In process industries, fouling, the accumulation of solid residues in equipment, poses a significant challenge, causing inefficiencies and productivity losses. Effective modeling of fouling evolution over time is essential for maintenance planning to prevent equipment from operating under suboptimal conditions. Traditional approaches to fouling prediction include equation-based models, which offer high precision but may struggle with continuously changing process bound-aries, and machine learning techniques, which are more adaptable but less effective at capturing rapidly evolving trends driven by complex underlying physics. This study introduces an innova-tive hybrid machine learning approach for predictive maintenance, combining the strengths of both methods. Pressure differential is modeled using an equation-based approach that links pressure data with fouling thickness, while the foul... [more]

Tableting is a dry granulation process for compacting powder blends into tablets. In this process, a blend of active pharmaceutical ingredients (APIs) and excipients are fed into the hopper of a rotary tablet press via feeders. Inside the tablet press, rotating feed frame paddle wheels fill powder into dies, with tablet mass adjusted by the lower punch position during the die filling process. Pre-compression rolls press air out of the die, while main compression rolls apply the force necessary for compacting the powder into tablets. In this paper, process variables such as feeder screw speeds, feed frame impeller speed, lower punch position during die filling, and punch distance during main compression have been systematically varied. Corresponding responses, including pre-compression force, ejection force, and tablet porosity have been evaluated to optimize the tableting process. After implementing an open platform communications unified architecture (OPC UA) interface, process variab... [more]

Understanding the complex dynamics of continuous processes in pharmaceutical manufacturing is essential to ensure product quality across the production line. This paper presents a data-driven modeling approach using Sparse Identification of Nonlinear Dynamics with Control (SINDYc) to capture the dynamics of a continuous direct compression (CDC) tableting line. By incorporating delayed control inputs into the candidate function library, the model effectively captures deviations from steady state in response to dynamic changes. The proposed model was developed by finding a balance between accuracy and sparsity, with focus on the ability to generalize to a wide range of operating conditions.

The paradigm shift into an era of Industry 4.0, also referred to as the fourth Industrial Revolution, has emphasized the need for intelligent networking between process equipment and industrial processes themselves. This has brought on an age of research and framework development for smart manufacturing in the name of Industry 4.0 [1]. While the physical and digital advancements towards smart manufacturing integration are substantial the inclusion of engineers themselves amongst this shift is often less considered [2]. There are educational efforts in Europe to create and implement smart manufacturing curriculum for non-traditional or adult learners already integrated in the workforce, but attention is also needed on a next generation smart manufacturing curriculum for pre-career students [3]. We, the teaching team of CHE 554: Smart Manufacturing at Purdue University, developed and implemented a curriculum geared towards the training of undergraduate, graduate, and non-traditional stud... [more]

Process control can be one of the most exciting and engaging chemical engineering undergraduate courses! This paper describes our experience transforming Chemical Process Control into Data Analytics, Optimization, and Control at the University of Notre Dame (second semester required course in the junior year). Our modern course is built around six hands-on experiments in which students practice data-centric modeling and analysis using the Arduino-based Temperature Control Lab (TCLab) hardware. We argue that state-space dynamic modeling and optimization are more critical for educating modern chemical engineers than topics such as frequency domain analysis and controller synthesis emphasized in many classical undergraduate control courses. All the course material is available online at https://ndcbe.github.io/controls.

Prompt engineering – improving the command given to a large language model (LLM) – is becoming increasingly useful in order to maximize the performance of the model and therefore the quality of the output. However, in certain instances, the user is not able to enrich the prompt with additional and personalized details, such as the preferred tone and length of generated response. Therefore, it is useful to create models that learn these preferences and implement them directly in the prompt. Current state-of-the-art inductive logic programming (ILP) systems can play an important role in the development and advancement of digitalization strategies. For example, they can be used to learn personal preferences of users without sacrificing human interpretability of the learned outcomes. These systems have recently witnessed the development of data efficient, robust, and human interpretable algorithms and systems for learning predictive models from data and background knowledge. In this paper,... [more]

Current progress in digitalization has led to a wide interest in learning more from available data. Advanced data analytics can be achieved through commercially available software; however, learning to program allows for more flexibility and, ultimately, more freedom in the potentially tailor-suited investigation. Among other programming languages, Python is one of the most requested, in industry and research alike. To intensify the earlier efforts and create both a pedagogical framework to teach programming to (bio)chemical engineers, and provide students with the opportunity to ask questions, we explore the integration of sPyCE and FermentAI into BioVL, a virtual laboratory for teaching (bio)processes, previously implemented by the authors. sPyCE is an open-source series of Python courses tailored to (bio)chemical engineers, FermentAI is a chatbot trained to answer questions about fermentation processes. The main goal of this work is to enable students to (i) learn (bio)processes and... [more]

Digital twins and digital shadows are frequently used terms by industry and academia to describe data-centric models that accurately depict a physical system intended for process monitoring and control. Processes restricted by a low degree of automation rely greatly on operator competencies in key decision-making; a digital shadow can here assist as a guidance tool [1-4]. This work presents a practical implementation of a digital shadow to support operators running a pilot scale-packed batch distillation column at the Technical University of Denmark (DTU) primarily used in education and teaching activities [5]. This operation is selected due to inherent unsteady process dynamics that are controlled by a set of manual valves, which the student operator must continuously balance to meet purity constraints without disrupting the operation. This realisation employ a modular software architecture, separated into four distinct modules compiled into Docker images and independently deployed. T... [more]

Developing ultra- or high-performance liquid chromatography (HPLC) methods for analysis or purification requires significant amounts of material and manpower, and typically involves time-consuming iterative lab-based workflows. This work demonstrates in two case studies that an autonomous HPLC platform coupled with a mechanistic model that self-corrects itself by performing parameter estimation can efficiently develop an optimized HPLC method with minimal experiments (i.e., reduced experimental costs and burden) and manual intervention (i.e., reduced manpower). At the same time, this HPLC platform, referred to as Smart HPLC Robot, can deliver a calibrated mechanistic model that provides valuable insights into method robustness.

This paper introduces a prototype for constructing an edge AI system utilizing the contemporary Machine Learning Operations (MLOps) concept. By employing microcontrollers such as the Raspberry Pi as hardware, our methodology includes data scrubbing and machine learning model deployment on edge devices. Crucially, the MLOps pipeline is fully developed within the ecoKI platform, a research platform for ML/AI applications. In this study, we thoroughly investigate the performance of our ecoKI platform by comparing it with the established Edge Impulse platform. We deployed the ML model with different weight quantization methods, such as FP32 and INT8, to compare accuracy variations and inference speed between these two platforms and quantization strategies on edge devices. In our experiments, we identified that the average accuracy performance of the ecoKI platform is 3.61% better than the edge impulse. Moreover, real-time AI processing on edge devices enables microcontrollers, even those w... [more]

Compartmental models are widely used to simplify the analysis of complex fluid dynamics systems, yet subjective compartment definitions and computational constraints often limit their applicability. The CompArt algorithm introduces an AI-driven framework that automates compartmentalization in Computational Fluid Dynamics (CFD) simulations, optimizing both accuracy and efficiency. By leveraging unsupervised clustering techniques such as Agglomerative Clustering, CompArt identifies coherent flow regions based on velocity and turbulent kinetic energy dissipation rate, ensuring a data-driven, physically consistent segmentation. The methodology integrates a connectivity-based clustering strategy, where compartments are dynamically optimized using the Silhouette score and adjacency matrix. This approach enables the reduction of high-resolution 3D CFD simulations into a network of interconnected sub-systems, significantly lowering computational costs while preserving system heterogeneity. The... [more]

As industries embrace the digitalization of Industry 4.0, the abundance of process data creates new opportunities to optimize industrial control systems. Traditional Proportional-Integral-Derivative (PID) controllers often require manual tuning to address changing conditions. This paper introduces an automated, adaptive PID tuning method using historical data and machine learning for a continuously evolving, data-driven approach. The method centers on training a surrogate model using historical process data to replicate real system behavior under various conditions. This enables safe exploration of control strategies without disrupting live operations. An RL (Reinforcement Learning) agent interacts with the surrogate model to learn optimal control policies, dynamically responding to the plant's state, defined by variables like operational conditions and measured disturbances. The agent adjusts PID parameters in real-time, optimizing metrics such as stability, response time, and energy... [more]

Maintenance is critical for industrial plants to ensure operational reliability and worker safety. In process industries, fouling, the accumulation of solid residues in equipment, poses a significant challenge, causing inefficiencies and productivity losses. Effective modeling of fouling evolution over time is essential for maintenance planning to prevent equipment from operating under suboptimal conditions. Traditional approaches to fouling prediction include equation-based models, which offer high precision but may struggle with continuously changing process boundaries, and machine learning techniques, which are more adaptable but less effective at capturing rapidly evolving trends driven by complex underlying physics. This study introduces an innovative hybrid machine learning approach for predictive maintenance, combining the strengths of both methods. Pressure differential is modeled using an equation-based approach that links pressure data with fouling thickness, while the foulin... [more]

The increasing complexity of industrial biotechnology demands advanced modeling techniques capable of capturing the intricate dynamics of bioreactors. Traditional regression-based and empirical methods often fall short when confronted with the highly nonlinear behavior and limited experimental data characteristic of bioprocesses. Addressing these challenges requires a more intelligent approach—one that leverages domain knowledge to model complex bioprocess dynamics effectively, even with sparse data, while maintaining interpretability and robustness. In this study, we introduce a process-informed, data-driven methodology for modeling the dynamics of industrial bioreactors, leveraging the capabilities of the rising field of Scientific Machine Learning (SciML). Our approach leverages Physics-Informed Neural Networks (PINNs) to seamlessly integrate domain knowledge encoded in physical laws with sparse experimental data and deep learning techniques, enabling precise simulation and modeling... [more]

Industrial process monitoring can benefit from utilizing historical data, providing insights for decision-making and operational efficiency. This study develops a soft-sensor-based approach leveraging multi-fidelity modeling to correct discrepancies between online sensors and laboratory analyses. A Gaussian process-based strategy is used to predict deviations between high-frequency low-fidelity sensor data and less frequent high-fidelity laboratory measurements. By exploring static and dynamic modeling frameworks, we assess their suitability for capturing process dynamics and addressing time-dependent variability. The multi-fidelity soft sensor noticeably improves predictive accuracy, outperforming high-fidelity and low-fidelity methods. This approach demonstrates applicability across various industrial settings where integrating diverse data sources enhances real-time process control and monitoring, reducing reliance on costly laboratory sampling.

In chemical process control, decision-making often involves balancing multiple conflicting objectives, such as maximizing production, minimizing energy consumption, and ensuring process safety. Traditional approaches for multi-objective optimization, such as linear programming and evolutionary algorithms, have proven effective but struggle to adapt in real-time to the dynamic and nonlinear nature of chemical processes. In this paper, we propose a framework that combines Reinforcement Learning (RL) with Multi-Objective Evolutionary Algorithms (MOEAs) to address these challenges. Specifically, we utilize MOEAs, such as NSGA-II, to optimize the parameter space of policy neural networks, resulting in a Pareto front of policies. This Pareto front provides a diverse set of policies that enable operators to dynamically switch control strategies based on real-time system conditions and prioritized objectives. Our proposed methodology is applied to a Controlled Stirred Tank Reactor (CSTR) case... [more]

This study focuses on optimizing production scheduling in multi-product plants with shared resources and costly changeover operations. Specifically, two main challenges are addressed, the unknown changeover behavior of new products and the need for rapid schedule generation after unforeseen events. An innovative framework integrating Machine Learning (ML) techniques with Mixed-Integer Linear Programming (MILP) is proposed for single-stage production processes. Initially, a regression model predicts unknown changeover times based on key product attributes. Then, a representation where distances correlate with changeover times is compiled through multidimensional scaling, allowing constrained clustering to group production orders according to available packing lines. Ultimately, the MILP model generates the production schedule within a constrained solution space, utilizing optimal product-to-line allocation from cluster segmentation. A case study inspired by a Greek construction material... [more]

Process Systems Engineering (PSE) provides the advanced conceptual framework and software tools to formulate and optimise well-considered integrated solutions that could accelerate the sustainability transition within the industrial sector. The landscape of advanced PSE is poised to undertake a considerable transformation with the rise in popularity of open-source and script-based software platforms with predictive modelling capabilities based on modern mathematical optimization techniques. This paper highlights three leading equation-based platforms—IDAES, Modelica, and GEKKO-that are increasingly utilised for the modelling, simulation, and optimisation of complex systems within the advanced PSE domain, alongside the strengths and limitations of each approach. Following this, we present a framework through which emerging techniques within the domain of Software Engineering could be leveraged to address these limitations, with a vision of improving the accessibility and flexibility of... [more]

Real-time carbon accounting is crucial for advancing policies that effectively meet sustainability objectives. This work introduces a carbon tracking tool specifically designed for the European electricity grid. The tool collects hourly data on electricity consumption and generation, cross-border power exchanges, and weather information to assess the real-time environmental effects of electricity use, employing locally-specific emission factors for the generation sources. It utilizes weather data from various stations across Europe to produce week-ahead forecasts of carbon intensity in the grid. Predictions are created using a random forest regressor, integrated within the optimal controller of an operational industrial batch process. This prediction-based optimizer seeks to reduce total emissions tied to the process schedule's electricity consumption by implementing a rolling horizon strategy. By leveraging enhanced energy flexibility, the controller provides significant opportunities... [more]

Process monitoring is essential in industrial settings to ensure system functionality, necessitating the identification and understanding of fault causes. While a substantial body of research focuses on fault detection, fault diagnosis has received significantly less attention. Typically, faults originate either from abnormal instrument behavior, indicating the need for calibration or replacement, or from process faults, signaling a malfunction within the system. A primary objective of this study is to apply the proposed fault diagnosis methodology to a benchmark that closely mirrors real-world conditions. Specifically, we introduce a fault diagnosis framework for a wastewater treatment plant (WWTP) that effectively addresses the challenges posed by imbalanced big data commonly encountered in large-scale systems. In our study, four distinct fault scenarios were investigated: fault-free conditions, process faults only, sensor faults only, and simultaneous sensor and process faults. To e... [more]