Using AI to evaluate non-thermal processing efficacy and assess the global impact

The food industry is increasingly prioritizing sustainable technologies to meet consumer demands and align with the global sustainable development goal of climate action. Unlike conventional thermal methods like pasteurization, non-thermal technologies such as high-pressure processing, pulsed electric fields, ultrasound, and non-thermal plasma offer innovative approaches to combat harmful microorganisms like Listeria, ensuring our food remains safe but also tasty and nutritious.

Advantages of non-thermal processing

Non-thermal processing technologies utilize various phenomena such as pressurization, electroporation, acoustic cavitation, and oxidation to provide effective microbial inactivation without the need for high temperatures. This approach not only preserves food’s sensory and nutritional quality but also aligns with sustainability goals by potentially reducing energy, water consumption, or food waste.

Current research and industry potential

Although promising, these technologies are not yet widely used in the food industry due to gaps in knowledge about their practical applications and legislative challenges. The current research in the domain aims to address these issues. Some researchers focus on understanding how microbes respond to these non-thermal processes, while others work on optimizing and up-scaling these technologies for industrial use. This project aims to merge the power of data science with existing research to enhance the impact of non-thermal processing, particularly for microbial inactivation.

A data-driven approach to get the most out of the technologies

To achieve this purpose, TRANSIT (https://transit-itn.com/) uses artificial intelligence to collect and analyze the data published in scientific literature worldwide. These findings help evaluate the efficacy of these technologies compared to thermal processing. The core idea is to create structural databases with relevant technical, biological, and food parameters and then use conventional and machine-learning techniques to analyze the results. In this way, TRANSIT evaluates not only the experimental data produced internally but also the data from every lab or pilot plant worldwide, providing a robust assessment of these technologies.

Identifying key paramers and evaluating non-thermal processing decontamination efficacy

One of the main goals is to identify the most important parameters affecting the decontamination efficacy to guide optimization procedures in industrial applications. By classifying the parameters involved in non-thermal processing, the professionals can prioritize these effects over others to explain most of “the problem” without adding unnecessary complexity. Although this ranking of the parameters is useful by itself, sometimes is needed to go one step further and provide quantitative answers. For this purpose, mathematical models are also being developed in the TRANSIT framework to predict the microbial population’s reduction under various processing conditions and in various food products. For instance, the time needed to reduce the microbial population by 90% can be used as a metric of the decontamination efficacy and how it can be affected by different processsing-, microbial-, and food-related properties. This becomes useful for professionals to assess if these technologies work and to what extent against different microorganisms and foods of interest.

Maximizing the impact: Guiding regulatory authorities and industry stakeholders

Furthermore, TRANSIT uses models to evaluate the impact of these technologies in cost-benefit analyses. For instance, it examines how the use of non-thermal processing methods could affect food waste in Europe, the incidence of foodborne diseases, and greenhouse gas emissions. This is done mainly through quantitative microbial risk assessment, where the different steps of the food supply chain are connected through predictive models, and in this way, the risk estimate of a foodborne pathogen can be assessed. This can be then translated to the estimated cases of foodborne illnesses in a specific area. For instance, the cases of foodborne illness after the consumption of fruit juices contaminated with Shiga-toxin producing E. coli in Europe can be estimated. Then, the effect of different processing tehcnologies (thermal or non-thermal) on this estimation can be examined to evaluate their impact. Similar approaches can be used for equivalent scenarios for spoilage microorganisms and food waste or greenhouse gas emissions. By providing these insights, TRANSIT aims to guide regulatory authorities and industry stakeholders in considering the adoption of non-thermal processing methods.

George Pampoukis

Verified

George Pampoukis combines his knowledge of food microbiology and data science to solve current problems in the food industry. His passion for understanding the science behind food led him to earn an Integrated MSc in Agriculture with a focus on Food Science and Technology from Aristotle University of Thessaloniki, Greece. He then moved to Athens to get another MSc in Food Science and Technology & Human Nutrition, specializing in Food Safety and Quality Management Systems, from the Agricultural University of Athens. During his studies, George focused on food microbiology and predictive modeling. From April to September 2021, he worked as a research fellow at the Laboratory of Food Microbiology and Biotechnology at the Agricultural University of Athens. Since September 2021, he has been an Early-Stage Researcher–PhD student at Wageningen University in the Netherlands, where he uses new data science techniques to explore sustainable processing technologies for the food industry.

Join our editorial team

Join our community and contribute to Wikifarmer, a platform based on the knowledge of a diverse community of farmers and people working in the agricultural industry from all over the world.

Article submission