AI PathFinder

Why this project?

Food companies increasingly recognise that digital transformation is essential for growth and competitiveness. However, many face challenges such as:

• capturing the right data,
• identifying their most important data related needs,
• processing and structuring data effectively,
• and navigating the overwhelming range of available AI solutions.

Except for the largest organisations, most food companies do not have sufficient in house expertise to evaluate or implement advanced digital technologies. The ambition to invest in AI is there – but the key question remains: how to begin?

Methodology and results

Through AI PathFinder, the project partners guided food companies in understanding the steps required to implement AI successfully. Working together with research institutes, food manufacturers, machine builders and technology suppliers, Flanders’ FOOD delivered the following results.

AI Checklist and step by step roadmap

Based on the Industry 4.0 Maturity Index (see Figure 1), the project developed:

• an AI readiness checklist, and
• a practical roadmap.

These tools clarify the conditions needed for successful AI deployment and offer guidance on how companies can strengthen these capabilities. The maturity phases help businesses improve specific processes without needing to reach the highest stage immediately.

Pilot Demonstrations at VEG i TEC

Two demonstration cases were developed at the VEG i TEC pilot facility to showcase the potential of AI in food processing:

• Case 1: Optimising the blanching process. AI was used to automatically adjust key process parameters, ensuring stable sugar and starch concentrations. This resulted in constant product quality and improved process control.
• Case 2: Predicting odour nuisance. This case explored whether AI could help predict unwanted odours during production. Although an AI model was developed, it became clear that no suitable sensor currently exists that can reliably detect the specific odour under investigation.

Inspiring Examples from the Sector

The project compiled a set of real world AI applications within the food industry. These cases aim to inspire companies and demonstrate the broad range of opportunities AI can offer.

Access to Project Results

The AI checklist and roadmap are not publicly available. They are used within the project consortium as core content for masterclasses and tailored guidance. What is available to everyone is the AI PathFinder Guide, which provides:

• an overview of the six steps towards successful AI integration,
• and practical examples from the food industry.
  Download the AI PathFinder Guide (NL)
   

Two webinars are also available, offering deeper insights into the VEG i TEC demonstration cases:

• Webinar: The power of AI and digitalisation at VEG i TEC. Explore how AI is used for process optimisation in the food sector.
  Order the webinar here! (NL)
   
• Webinar: unwanted odours in production processes. An update on the search for suitable odour sensors and the potential of AI for odour detection.
  Order the webinar here! (NL)
   

Target group

AI development depends heavily on data availability. Machine builders and technology providers play a key role in generating, structuring and supplying this data, which is why they were closely involved in the project. They form an important target audience for the project outcomes.

Project partners

Flanders’ FOOD managed and coordinated the project. The project was executed by:

• Imec - Kris Van de Voorde
• Sirris - Christophe Michiels & Jakob Kesteloot
• Ghent University | VEG-i-TEC - Imca Sampers
• Howest University of Applied Sciences - Bart Leenknegt