Belgian Factories of the Future and Machine Learning: What is possible and what is required for the manufacturing industry? | Agoria

Belgian Factories of the Future and Machine Learning: What is possible and what is required for the manufacturing industry?

Published on 22/10/18 by Sibylle Dechamps
Machine Learning is a section of Artificial Intelligence that addresses the development of the learning abilities of manufacturing machines based on data, algorithms and technology. But how can this be converted into specific applications? What is possible and what is required? We brought technology providers, innovative manufacturing companies and cutting-edge start-ups together to exchange ideas and insight. It was no coincidence that this took place at Van Hoecke NV based in St-Niklaas, a company that, as a kitchen industry partner, is committed to the digitisation of manufacturing processes and solutions.

Keep calm and pioneer on

Let's begin with the bottom line: There is a lot of pioneering to be done in the manufacturing industry. The road ahead is not clearly visible, which is normal. Technology providers not only provide solutions but also act as consultants. The following are some of the pitches made by these providers:

At Ometa, Radix and Yazzoom, the digital twin returns as a specific application of 'learning' for a product, process or the complete project life cycle. This digital twin is used to trace glitches, opportunities to make processes even more efficient and when maintenance or repair work is required (predictive maintenance).


""What would you like to monitor for machine learning as a company? The product, process or machine? This initial decision is essential and many companies decide to focus on the machine."
Mathias Verbeke, Sirris

Start-up Factry offers its data processing solutions and return structured data to the production department of manufacturing companies. The data is used to create a predictive model that visualises the production process one to two hours into the future. Faktion, another start-up which should not be confused with Factry, not only provides 'pure' machine-learning solutions but also invests in chatbots and consultancy to provide optimal support to their customers. Sirris guides companies during the first data collection phase providing both an AI on-demand platform and custom-made master classes for their customers.

The focus is predominantly on technology, but your staff needs to be involved in it from the start too."
Frederik Van Leeckwyck, Factry

Tackle problems one step at a time 

Technology providers can, therefore, provide Machine Learning as a fully fledged package, but how do manufacturing companies perceive implementation in their daily processes, products and machines? Companies raised a few uncertainties and questions during the discussion. How do you structure the large quantities of collected data and, at the same time, obtain a result fast? Which infrastructure is required to implement Machine Learning? What limits the options? Is the effort put into the planning worth it when you consider the output and especially when companies are facing labour shortages? Or the basic question: How to start working with Machine learning at your company?

"The manufacturing industry should not be scared to turn to AI and Machine Learning. Not getting involved is not an option." 
Geert Engelrelst, TE Connectivity

Have you asked yourself similar questions? Read the key takeaways of the session below!

Key takeaways: 

  • Start small: Setting up a large ambitious project may block the progress of this project and it is impossible to know today where you will be tomorrow. Take small steps towards your goal to allow for the monitoring, assessment and steering of the project as it develops.
  • Strategic decisions: It is still very tempting to try to link everything to machine learning, machines, processes and products. Once again, select one aspect to address and focus on that!
  • Quick wins! Analyse where small changes can already lead to specific results and, therefore, success!
  • AI and Machine Learning do not replace people: Success will simply depend on the strength of the combination of data science and technological machine and process experts.
  • Involve your employees and stakeholders from the start
Was this article useful?