Technology Internship in Advanced Digitalization of Food Processes Based on Data Integration, Modeling and Artificial Intelligence

The project will be carried out at AZTI's facilities in Derio, with specific infrastructure for its execution: advanced sensing, IoT systems and platforms, physico-chemical laboratories, platforms for data acquisition, management, and modeling, and software tools intended for the analysis and development of models. Additionally, the fellowship will be carried out in an environment with experience in the food industry, artificial intelligence, intelligent sensing, productive efficiency, sustainability, and industrial validation.

Would you like to join a leading, innovative organization strongly committed to research and providing value to society?

AZTI's purpose is to drive positive change for the future of people, contributing to a healthy, sustainable, and integral society. And you can be part of this challenge.

From the Efficient and Sustainable Processes (PES) team, we are looking for a person to carry out a Technology Fellowship at AZTI on "Advanced digitalization of food processes based on data integration, modeling, and artificial intelligence.

The selected candidate will apply for the IKERTALENT 2026 - training grants for research and technological personnel in the scientific-technological and business fields of the agricultural, fishing, and food sectors. In this case, it is a training grant program for technological personnel specialized in areas of the Basque agricultural, fishing, and food sectors (duration 2 years).

Project title: Advanced digitalization of food processes based on data integration, modeling, and artificial intelligence.

Project tutors at AZTI: Idoia Olabarrieta, Bruno Iñarra.

Center: AZTI Derio.

The project will develop and validate a comprehensive advanced digitalization strategy applied to the food industry, covering the entire data value chain and aimed at transforming data generated in the plant into useful information for monitoring, prediction, optimization, and decision-making.

The project will thus contribute to strengthening the digital and sustainable transformation of the food industry, developing solutions with transfer potential and contributing to improving the operational efficiency, quality, traceability, sustainability, and resilience of food systems. Likewise, it is aligned with the priorities of the PCTI 2030 and PEGA 2030 related to the digital and sustainable transition of the Basque food sector.

In a first phase, use cases, critical process and quality variables, acquisition requirements, and available data sources will be identified, including process sensors, non-destructive sensors, and advanced sensing, such as NIR spectroscopy-based optical sensors and electrical sensors, IoT systems, physico-chemical analyses, and operational records. From this, a functional architecture will be designed for the capture, communication, storage, integration, processing, and exploitation of heterogeneous data, paying special attention to its quality, traceability, consistency, acquisition frequency, and interoperability.

Based on this infrastructure, statistical and predictive models will be developed based on chemometrics, mathematical modeling, and machine learning and artificial intelligence techniques. Work will involve, among other approaches, PLS models applied to signals from sensors, temporal models for continuous multivariate data, and hybrid models that combine process knowledge with machine learning techniques. Special attention will be paid to the interpretability of results, robustness against variability in operating conditions, the ability to update with new data, and the generalization of the developed models.

These developments will allow for the monitoring of parameters and variables related to the efficiency, quality, safety, and sustainability of the food industry and its processes, estimating unmeasured variables, detecting deviations, predicting their evolution, and evaluating different operating scenarios. The models will be integrated into decision support tools with predictive and prescriptive capabilities aimed at optimizing quality, efficiency, resource consumption, sustainability, and process stability.

The developed solutions will be applied to relevant operations in the food industry, including thermal, fermentative, extractive processes and/or online quality control systems. Their robustness, generalization capacity, and operational utility will be validated in real AZTI use cases, under conditions close to real operation, evaluating their predictive capacity and their contribution to improving quality, efficiency, and sustainability.

Finally, the integration of data, models, and simulation will provide a methodological and technological basis to advance toward future digital twins applicable to real industrial environments. The modular and replicable approach will facilitate the adaptation of developments to different processes, food subsectors, and digital maturity levels, favoring the transfer of the developed solutions to the industry.

The selected person will have an active role throughout the entire project, participating in the capture, communication, storage, and integration of data, the development and evaluation of models, validation in use cases, and the deployment of digital tools. This will allow them to acquire specialized training and practical experience in digitalization applied to food processes, working at the intersection of sensing, data analysis, mathematical modeling, artificial intelligence, process efficiency, and sustainability.