Jorge Iñaki Gamero-Barraza1, Aylin Socorro
Saenz-Santillano2, Efren Delgado1, Cristian Patricia Cabrales-Arellano3, Damián
Reyes-Jáquez2*
1Food Science and
Technology, Department of Family and Consumer Sciences, New Mexico State
University, P.O. Box 30001, 88003-8001, Las Cruces, New Mexico, USA
2Department of
Chemical and Biochemical Engineering, National Technological Institute of
Mexico (TecNM) – Durango Institute of Technology (ITD), Blvd. Felipe Pescador
1830, Nueva Vizcaya, Durango, Durango, 34080, México
3Department of
Biology, Eastern New Mexico University, Eastern New Mexico University, 1500 S
Ave K Portales, NM 88130, New Mexico, USA
*For Correspondence
damian.reyes@itdurango.edu.mx
Extrusion is a versatile
thermomechanical technology with considerable potential for transforming
agro-industrial by-products into value-added food and feed products. However,
the complex interactions among raw-material composition, moisture, temperature,
screw speed, mechanical energy, and equipment configuration make extrusion
inherently nonlinear and difficult to optimize using conventional
trial-and-error approaches. This chapter examines sustainable food and feed
extrusion from an integrated perspective, linking the valorization of cereal,
fruit, legume, and oilseed-derived by-products with the physicochemical
transformations and functional properties generated during processing.
Particular attention is given to the effects of extrusion on starch, dietary
fiber, proteins, and lipids, as well as on hydration, expansion, bulk density,
hardness, and digestibility. Traditional optimization through response surface
methodology is discussed alongside emerging data-driven approaches, including
artificial neural networks, Random Forest, XGBoost, and Bayesian optimization.
These methods offer increased flexibility for capturing nonlinear interactions
and supporting multi-objective optimization involving product quality,
nutritional integrity, by-product inclusion, and resource efficiency. A
structured framework for data-driven extrusion is presented, encompassing
comprehensive ingredient characterization, robust experimental design, data
preprocessing, model validation, and multi-objective optimization. Finally, the
integration of predictive models, inline sensing, and intelligent control
systems is discussed as a pathway toward adaptive and more resource-efficient
extrusion. Combining extrusion technology with machine learning and rigorous
data science can accelerate the transition from empirical process development
toward predictive manufacturing while supporting waste valorization and the
principles of a circular bioeconomy.
Keywords
Extrusion Processing, Agro-industrial By-products, Circular Bioeconomy, Machine Learning, Response Surface Methodology, Bayesian Optimization, Artificial Neural Networks, Multi-objective Optimization, Process Digitalization, Waste Valorization
Download PDF