Those who want to be able to improve their product must also know exactly what does not work. During her lecture at Advanced Engineering, the CEO of Effex talks about the importance of qualitative data, failed experiments and daring to innovate.
‘More and more companies know our message, but do not yet dare to apply it,’ explains Dewi Van De Vyver. ‘Companies in the pharma, biotech and chemical sectors, for example, often have a reputation for being innovative, but there is clearly still reticence and conservatism when it comes to Design of Experiments. These companies are consciously concerned with cost optimisation; things that take time and/or do not deliver immediately thus dare to be abandoned prematurely.’
During her session at Advanced Engineering, Dewi wants to inspire people to take the plunge into experimentation anyway. As an entrepreneur herself, she can draw on a rich experience grounded in innovation. She already won the ICT Woman of the Year award and was included in the Inspiring 50 Europe list in 2022.
3 reasons why you have to want to fail: Dewi Van De Vyver on Design of Experiments
“Companies looking to improve their product or process need to keep experimenting to discover which changes to their production can have the greatest impact.”
1. Lots of data =/= lots of knowledge
‘Almost all companies collect data today, but they don’t always know what to do with it afterwards. This is not surprising, because that data is usually collected during the normal, controlled operation of their production process. If something works as it should, you don’t learn anything from it. You have to step outside these predefined conditions to know how each element works in extreme conditions, or how they will interact with each other at that moment. Less, but qualitative data will teach you much more and only then do you have a starting point to work on.’
2. Fail must!
‘When we design experiments, we don’t necessarily start from combinations that we know will succeed. Only by also failing do you find out how different factors work with each other – and against each other – in different circumstances. That’s how you soon find the path towards the breakthrough you may have been chasing for a long time.’
3. Invest to reduce costs
‘With a good experiment, you can test several factors at once, instead of starting to test elements separately. This already saves you time and money. In addition, with good experiments you get a much more accurate view of what the entire production process will cost you after adjustments.
Experiments sometimes have a high cost up front, but you then know immediately what the further investment will cost you. In the other case, you discover by doing and each time you encounter surprises.’
‘Design of Experiments: how smart data acquisition and analytics accelerates innovation’ can be found during Advanced Engineering on Thursday 22 May at 11h30 on the Main Stage. Register today!
