A man working in machine fabrication standing in front of a large component and using augmented reality functions on a tablet.

“I like to quote Sesame Street here.”

MISUMI expert discussion on artificial intelligence in mechanical engineering: episode 1

 

Is the sector fit for the future when it comes to AI? What role does digitalisation have to play? And what about the human component? Jens Neumann (Manager Production Planning, Supplier Management & Digitalization at MISUMI) took the opportunity to speak to experts Markus Jung and Norbert Lukic from the X-Reality and digitalisation specialists XRGO about the status quo in the sector. The focus is on a pilot project that highlights the relevance of machine learning in combination with innovative technologies such as augmented reality – and also what can happen when companies miss the boat.

 

 

 

Neumann: Before we start talking about your pilot project, I’d like to hear your general opinion on how X-Reality is currently being used in the mechanical engineering sector. Are technologies like augmented reality already being used extensively?

 

Lukic: I have to say, the answer is a clear yes and no. Alongside the pioneers who have already completely integrated X-Reality into their processes, there are of course several sceptics, for whom this technology is just a gimmick. In between these two camps, there are many companies that are cautiously getting to grips with this topic, consulting with us and are interested in simply trying the technology out. During the pandemic in particular, the interest in AR glasses – such as Microsoft’s HoloLens – for use in customer support or manufacturing has grown rapidly.

 

Jens Neumann

Manager Production Planning, Supplier Management & Digitalization
„Tailor-made digitisation and automation strategies will become even more important in the future in order to remain competitive.”
Copy link:

Jung: The general feeling after the first comprehensive tests is that X-Reality provides a boost to efficiency that users no longer wish to do without. “I cannot understand why we didn’t try this earlier” is the sentence we hear most often from decision-makers at our customers. Without wanting to give too much away, we also achieved this efficiency increase in the pilot project we already mentioned.

 

 

Two people use a tablet to view a virtual prototype of a machine component.

Whether in training, manufacturing or sales: X-Reality boosts efficiency in many different aspects.

 

Neumann: Well, let us jump right in here. What was the starting point for the project?

 

Jung: One of our partners from machine production was wondering how to improve the efficiency of his machines – and thereby the manufacturing processes of his customers. Following initial talks, we had a clear plan to use dynamic maintenance intervals. Basically, this means that we no longer follow predetermined cycles, but adhere to intervals that are aligned with the wear of the machines and individual components.

 

Neumann: That sounds quite logical. What benefits do dynamic maintenance intervals provide exactly?

 

Lukic: The goal of dynamic maintenance intervals is to maximise the intervals. Let’s look at a part in a manufacturing machine, for example. Previously, this part was replaced after, say, 20,000 operating hours – in keeping with the principle of “better safe than sorry”. Dynamic maintenance intervals allow operators to identify that this part only needs replacing after 27,000 operating hours when production intensity is low, or after just 16,000 hours when the loads are higher. This approach not only optimises component utilisation, but also prevents unplanned downtimes.

 

Jung: Of course, by optimally utilising the life cycle, we are able to reduce the use of replacement parts throughout the machine’s service life – simply by not replacing these parts earlier than is necessary. This lowers costs and is in line with a concept of sustainability. A further benefit is the ability to minimise storage of these parts. By linking to a well-organised parts catalogue, like the one from MISUMI, the required parts can be ordered automatically and delivered in time.

 

Neumann: Did you discover any additional benefits during the project?

 

Lukic: Absolutely! And our partner was delighted. In addition to preventing unplanned downtimes and being able to predict maintenance requirements more accurately, we also noted a reduction in unforeseen technician deployments and issues in the supply chain. Making maintenance intervals “dynamic” had a positive impact on the planning and execution of activities.

 

Neumann: We have heard about the technological aspects. But what about the human component?

 

Jung: That’s a good point. People will continue to play an important role in maintenance, as well as in the production sector as a whole. This becomes especially apparent when you are faced with an acute lack of specialist staff, or when long-term employees retire – you can’t simply replace knowledge and experience with nothing. In our example, we are focussing on conserving expertise and knowledge, in order to then use this to train staff in an optimised and targeted manner. X-Reality, or to be more specific augmented reality, plays a key role here.

 

 

An engineer is using VR glasses which display a step-by-step guide.

It is clear that the human component will continue to play an important role, despite the technical innovations.

 

Neumann: Perfect transition: which technologies are used?

 

Lukic: When it comes to the human component, i.e. the staff, we are using augmented reality to create interactive methods for learning, for example step-by-step guides or guided or even assisted approaches. This not only makes a significant contribution to increasing staff efficiency, but also occupational safety.
When it comes to predictive maintenance, we use technologies from the areas of artificial intelligence and machine learning. We rely in part on standardised software, but also on solutions we developed ourselves, depending on the use case.

 

Neumann: This is often the part where the theory is put forward that machine learning can only be as good as the data with which it is fed. Is this justified?

 

Jung: Yes, very much so. Without data, neuronal networks cannot form the information needed to recognise patterns and make predictions based on these. This is not just the case when it comes to machine learning, it also happens in our brains.
Collecting and analysing machine or production data will therefore have a significant impact on the futures of many companies. I like to quote the German intro for Sesame Street here: If you don’t ask questions, you won’t get answers – and if you keep on practising, you can do anything.

 

Lukic: (laughs) That’s a great comparison! And as machines can usually only do what we teach them, it is up to us to “train” them and feed them with the data they need. This won’t work without empirical values and this is what we derive from the collected data.

 

Neumann: So can we say that digitalisation is the key?

 

Lukic: You’ve hit the nail on the head there. Digital technologies have long been part of our everyday lives, but only very few companies have made them part of their strategy.

 

Jung: The terms “digitalisation” and “digital transformation” have been around for years, with some sectors now dismissing them as buzzwords – despite them describing a vital and ongoing process of change in companies, both technologically and economically. I would even go so far as to say that if a company does not have a digitalisation concept today, it may be missing a decisive piece of the puzzle to remain competitive tomorrow.

 

 

Do you want to find out more about the topic of X-Reality? Then read our supplemental article on X-Reality: four technologies that will only benefit you!