Enhancing efficiency in food processing - Smart sensoring and data analysis
Practical information:
Want to register?
- Prerequisites: A technical background is highly recommended.
- Price: €650
Georganiseerd door:
Goals
This course offers a comprehensive exploration of cutting-edge sensor technologies and smart data analysis and how to apply them in food processing. Participants will:
- understand the latest sensor technologies and their role in enhancing operational efficiency and adaptability;
- learn to select and implement sensor systems that align with specific production needs and cost considerations;
- develop skills to integrate sensor data with AI-driven analytics for informed decision-making and process optimization;
- recognize how improved efficiency and adaptability contribute to sustainable practices as a natural outcome.
Programme
1. State-of-the-art overview of sensor technology in food processing
Ensuring product quality is a top priority in the food industry and advances in sensor technology are enabling faster, more reliable and non-invasive quality assessment methods. In this session you will gain a broad yet practical overview of sensing technologies relevant to food processing and quality control. We will begin with fundamental sensing principles and then explore more recent technologies that can rapidly assess key quality attributes such as composition, freshness, contamination, and texture. Through real-world examples and industry case studies, we will discuss the strengths and limitations of different sensor systems and their suitability for various food applications. To conclude, we will introduce a technology-application matrix, a practical decision-making tool to help food companies identify the most effective sensor solutions for their specific challenges.
Lecturers:
- Bart De Ketelaere, Research Manager, Mechatronics, Biostatistics and Sensors (MeBioS), KU Leuven
- Tim Van Meer, Technical Director, Pomuni
2. System integration & data infrastructure
Collecting data in rural and harsh environments has quite some challenges. Mostly, data captured at sensors have to use non-traditional means of communication to be collected at the server, where analysis is done. In this session, you will learn how to define data, how to format it and how to send it over a very constrained environment and limited bandwidth to the server. We will discuss some common wireless communication methods and give insight into their advantages and disadvantages. Further, we will dive into the concept of edge computing, where data is analysed at the edge using embedded computing. We will conclude with the contemporary trend of machine learning at the edge and how complex analysis can be done at battery-powered constrained devices.
Lecturers:
- Prof. Hans Hallez, Distributed and Secure Software (DistriNet), KU Leuven
- Jonathan Kesteloot, Co-founder, Captic
3. Data analysis techniques
In this session, we will delve into the complex task of transforming raw data into meaningful information, which must be delivered to the right user at the right moment. This phase of data analysis and processing presents its own set of challenges. Often, decisions need to be made swiftly in real-time during production processes, requiring the use of sophisticated, real-time AI solutions that can process data quickly and accurately. Despite these technological hurdles, when effectively implemented, sensors offer substantial advantages by improving quality, safety, and efficiency. Techniques for data analysis, including artificial intelligence (AI), machine learning, deep learning, and decision-making processes, are essential for extracting insights from sensor technology. During the session, concrete use-cases will be presented in which AI is applied to interpret acoustic or radar signals for food quality prediction and for estimating the remaining useful life of processing equipment.
Lecturers:
- Matthias De Ryck, Innovation Manager, Declarative Languages and Artificial Intelligence (DTAI) ,KU Leuven - Brugge
- Prof. Peter Karsmakers, Declarative Languages and Artificial Intelligence (DTAI) ,KU Leuven - Geel
Practical
All sessions will be in English, including lecture materials.
Teachers / speakers
Bart De Ketelaere
Bart De Ketelaere is Innovation Manager (IOF) at the MeBioS division of the Department of Biosystems. He combines a Master in Bioscience Engineering (KU Leuven) and a Master in Statistics (UHasselt). During his PhD, he worked on the broad field of industrial quality control with applications in agrifood. His main interest is in combining novel sensor technologies and data analytical tools for product and process control. His AI-related interest relates to building and maintaining models in case of limited data, as well as to the analysis of (hyperspectral-) images and spectra. He is (co-)author of +150 peer reviewed papers and several patents. Besides, he is co-founder of two spin-off companies, both active in the field of data science.
Keywords: industrial quality control, Agrifood, Image Analysis, Applications of statistics, machine learning & AI
Hans Hallez
Hans Hallez is a lecturer (Docent) of physics and informatics to 1st year (freshmen) academic bachelor Industrial Engineering students. This is within the faculty of Industrial Engineering (FIIW) of the KU Leuven (TechnologyCampus Oostende). In the final master year, he also teaches "optoelectronic communication" to electronic engineering students.
His research interests are within electronic implementation for medical and industrial applications. More specifically his interest lies in the implementation of signal-processing algorithms within programmable wireless sensor networks. He is also interested in new sensors and measurement techniques.
Hallez is proficient in conducting research concerning electronics, physics and ICT with medical applications. From the more fundamental research on EEG and ECG signal processing at the UGent, he's gone to research on a demand-driven and practical basis at KU Leuven. There he is responsible for teaching and conducting research.
Matthias De Ryck
Matthias De Ryck is an AI Solution Architect at Superlinear, driving strategic AI adoption and innovation for industry-leading clients. With a PhD and postdoctoral experience at KU Leuven in Computer Science, specializing in multi-robot systems and machine vision, he bridges academic expertise and industrial deployment.
He has successfully led VLAIO-funded innovation projects, translating cutting-edge research into operational value. At Superlinear, Matthias designed and led the computer vision roadmap for the Port of Antwerp-Bruges, tackling impactful use cases including debris detection, quay wall inspection, berthing validation, and train wagon and vessel identification.
He also spearheads the development of Superlinear’s open-source computer vision accelerator SuperSight, enabling scalable and efficient CV prototyping across sectors. As a trusted consultant, Matthias guides companies in prioritizing AI efforts and leveraging state-of-the-art technology for measurable impact.
🧠 AI Strategy & Consulting – Solution Architect at Superlinear, advising across logistics, maritime, and industrial sectors on CV adoption
📊 Applied Computer Vision
– Led implementation of CV tools at Port of Antwerp-Bruges: improved
asset monitoring, detection accuracy, and operational efficiency
🔬 Academic Foundations – PhD and postdoc at KU Leuven in robotics & machine vision; published on distributed multi-agent coordination
🚀 Innovation Leadership – Coordinated successful VLAIO-funded R&D projects to accelerate industrial AI innovation
Peter Karsmakers
Dr. Peter Karsmakers received the M.Sc. degree in artificial intelligence in 2004 and the PhD degree from the Department of Electrical Engineering, KU Leuven, in 2010. From 2010 to 2013, he was a post-doctoral researcher in the MOBILAB research team from Thomas More. From 2013 to 2018, he worked as a post-doctoral researcher at KU Leuven where he co-founded the ADVISE research team. Currently, he is an Associate Professor within the Computer Science Department in the DTAI section at KU Leuven and is a member of the Leuven.AI institute. Since 2022, he is a PI of Flanders Make@KU Leuven. His research interests include designing machine learning algorithms that consider application-specific constraints. These can, for example, relate to the computing platform on which the machine learning algorithm will be deployed on, to the need of physical consistency between model variables, to the lack of data annotation or to other application related specifications. He worked on diverse projects, mostly in collaboration with industrial partners, that involve monitoring of both humans as machines using sensors such as microphones, accelerometers and radars.
Related courses
Artificial Intelligence in Business and Industry
Postgraduate - Kortrijk - KU Leuven, PUC - KU Leuven Continue, VAIA