Episode 4: Engineering the Food Industry Forward: a Challenging but Limitless Evolution

By 28 March 2023July 26th, 2023Quality Leaders Podcast

The way our food travels from field to fork has changed dramatically over the years. Bram de Vrught (managing director) and Teun Keusters (deep learning engineer) know all about it, thanks to their experience at Qing Engineering. In this episode they discuss the specific challenges the industry is facing, such as a growing demand for high quality food and the shortage of labour. They also tell Maxime which technologies enable tackling these challenges, and explain how the combination of Vision AI and machines is pushing the industry forward.


The sky is the limit: technology that seems impossible now, might be possible tomorrow

AI is evolving fast. Every week there are new insights, new methods and new tools. “What seemed impossible a couple of months ago, might be possible tomorrow”, says engineer Teun Keusters. When it comes to technology within quality assurance, there is only one limit that comes to mind for Keusters: defects need to be sensed in some way. “Some customers have asked if they can look inside their product, with deep learning and two dimensional cameras. That’s impossible.” – For now.

With AI it’s not just about the “how” but also about the “why”

Many companies that want to implement AI in their production process are cost driven. They want to replace a couple of people with a new installation and think about the costs and the ROI (Return On Investment). But, as Qing Engineering points out, implementing AI should also happen for value driven reasons. “By integrating an AI model, you can see if a product is okay or not. On top of that, you also learn to understand why it is okay or not. You can use that information to prevent producing bad products in the future”, explains Bram de Vrught.

You don’t need to build a whole new system to use AI: “You also wouldn’t build a new Excel if you want to create a spreadsheet”

In order to start using AI, you shouldn’t necessarily build a whole new AI system. Bram de Vrught compares it to Microsoft Office: if you want to create spreadsheets or documents and update them, it’s not necessary to build a new Excel or Word. You can simply use the platforms that already exist. It’s the same for AI: it’s a completely different thing to develop a new system or a new tool, than to just use an existing platform. “With Qing, we use the platform of Robovision. That way our clients can use and maintain an AI system, without needing their own data scientists”, he concludes.