“Prescriptive Maintenance, the next step to Predictive Maintenance” – Anja Vedder talks with Dr Johannes Pohl in his podcast “Business Unplugged”
In this episode Johannes Pohl talks with our founder Anja Vedder about the exciting topics: What actually motivates someone to leave a company and found their own start-up? What does successful implementation with a start-up look like? What do all these buzzwords like predictive and prescriptive maintenance actually mean?
Dr Johannes Pohl is himself is founder and entrepreneur of the management consultancy Peopex that deals with the successful implementation of operations and change management. For him, people play a central role in a successful company. He also devotes himself to this aspect in his podcast “Business Unplugged”, which focuses on the people behind these companies.
In the conversation with Anja, the central question is: what is predictive or prescriptive maintenance?
Predictive maintenance describes when machines fail and when is the best time to repair them. But what does it take to implement this?
“Not only predict, but also prescribe an anomaly. An event comes up, you need to be able to say, where does it come from, why is it there and what needs to be done?”, says Anja Vedder. For her, these are the necessary ingredients to run a reliable machine successfully. The understanding of the machines and derived from that, the recommendations for actions are the basis for optimisation and future automation.
One topic, many solutions on the market
The technical implementation of such solutions can have very different approaches. It is important not only to apply AI or static machine learning models, but also to have knowledge about the individual machines and to understand the use cases. Anja uses the example of pumps to describe what to look out for and why not all machines are the same:
“Nevertheless, it is the case that pumps can be found in different applications. And these large compressors, that we are looking at, are really tailor-made and unique. And that’s why it doesn’t help at this point to say: I have a large database. Instead, I have to be able to create a model that takes into account the individual physics. And in doing so, we do a model fitting, not based on the physical size of how the machine was designed, but we do a model fitting based on the data.”
In addition to all these questions, Anja also gives tips on what has contributed to the success of her company, the advantages of cooperation between start-ups and companies and why you also have to celebrate the small successes.