High voltage takes center stage in this season of Hitachi Energy's Power Pulse podcast. We promise to bring you great content from the brightest minds in the business. We'll discuss challenges, opportunities, and all the hot topics any high voltage enthusiast or anyone interested in sustainability for that matter, is sure to enjoy. This episode of the podcast is all about quality and to tell you about quality in high voltage, we asked none other than the head of quality on as a guest.
He is Thomas Haas and will tell you about how digitalization is used as a tool to make high voltage safe and reliable. Thomas holds a masters in Electrical Engineering from the University of Aachen. He's worked in the power industry for just about 30 years and is very passionate about all things digital, especially the Digital Factory. But how digital can a factory working with equipment, from heavy machinery to small nuts and bolts, really be? You'll have to stay tuned to learn.
Welcome back to Power Pulse. I'm your host, Sam Dash, and today I'm speaking with Thomas Haas, Head of Quality for High Voltage Products. Hi, Thomas. Hi, Sam. And thanks for the invitation. It's great to be here. You're very welcome. Glad to have you. Thomas, you've been at Hitachi Energy for close to three decades. That right? That's about right. Yes, absolutely.
So when you reflect back, how would you describe yourself when you first entered the company, what your passions or anxieties were and then how do you compare that to now? I mean, I was a young engineer who was invited by a management member from ABB at that time to come to the first interview and then join. And yes, I liked it and I was just curious about everything to come. And for those of us who don't know the history of Hitachi Energy, can you say more about ABB?
Can you talk us through a bit of that history? Absolutely. When I joined ABB at that time in Germany, by the way, ABB as such has been formed out of two companies from Boveri Company and Swedish ASEA. And these two together formed the ABB that was in 88. Now I joined seven years later. Then after many years, there was the split into the energy sector of ABB – energy product related and the automation sector. And then ABB decided to divest the energy portion.
And that's what became Hitachi Energy- What became Hitachi Energy, fully owned by the global Hitachi Company. Yeah, yeah. Thanks for talking us through that history. That's really helpful. Oh, we haven't had anyone contextualize that yet for us. How does your particular working knowledge affect your daily life outside of work? Are you hyper aware of quality control on a daily basis at home with appliances or vehicles? Well, I wouldn't say too much in this way.
I mean, I have remained an engineer at heart. Yeah. Right, so I'm interested in everything that works. A machine that works, controlled by electronics, by digital mechanisms. And of course, it needs a certain quality, because I can become curious if things stop working as they are supposed to do. So that is with my car and my TV. I don't need a new one every day. I would like something that lasts for a while.
So I feel like, maybe after this podcast, I should get your contact info and give you a call whenever I need to buy a new TV or a new appliance because I'm sure that you do some really good research into what the best option is. Maybe, but not always. Yeah. I'm still also one of those who get what appeals me. Yeah. So let's get more into high voltage. I want to understand just how digitized a high voltage setting is.
We've learned about how parts get made and assembled, and we've also learned how a factory is set up. Can you explain this digitalization process at high voltage in layman's terms? Yeah, if you want. Digitalization starts with introducing a so-called enterprise resource planning tool. I should not name a brand name, most probably, but there is some worldwide known software which is used in many areas to control all manufacturing processes and all the material handling in the factories. Right.
This we have been using in our company for decades also as well. I grew with these systems, which basically take care that all your products follow a certain work step, a certain work flow that you have feel of materials, which materials go into your product, and that all these materials need to be purchased from suppliers, from other companies or other units in our own company? Yeah. And yeah, these parts need to be taken into the system, into the factory, but they also need to be registered.
So we would like to know at any time how much or how many of which product are in our warehouse in a specific box on a specific pallet. And this is why we need to register what comes in and, and increase the counter of the parts which are in a certain box, and reduce that counter once we consume out of this box. And in very simple terms, I'm sure it's more complex than this; is it something similar to how we use barcodes at a store or in the supermarket? Yeah, and now comes the point.
And that is about what is now changing over time and what we have now developed further. The managing of inflow and outflow of material and counting is something that can be done with the delivery sheets or with the purchase order of figures. What we want to know. We want to know more about each and every individual product, especially when it comes to really critical parts that really define the performance of our products.
Can you give us an example of what those critical parts are for a given piece of machinery or equipment that you're wanting to keep track of more? Yeah, I'm sure that you have introduced in some of the earlier episodes some of our products, and you may have come across the commonality about all high voltage products is that we need to insulate voltages, so we need to put insulating material between two poles of voltage.
And I would like to take the example of a disc space insulator or disc spacer or disc insulator, which is made from an epoxy cast. Yeah. This is a highly matured or highly developed and specific component of our products, and it needs a very decent quality control during mixing, during pouring and casting it into the final product. That epoxy. Exactly. It's a bit like, you can compare it with cookies where you have a dough. Yeah. A kilogram of whatever put together.
You have a block of dough and then you make a lot of cookies out of it. But you want to be careful that you're not trying to make that dough that has butter, which has a melting point at too high of a temperature. Or you want to make sure not to put that dough maybe in the freezer right before you're going to bake it. Right. So all of this depends on temperatures and conditions. Yes. And then the first thing is that we of course, in our own factory, we have a factory to produce that part.
First of all, we have all the controls in place to make sure that we have the best quality you can think of. And it has happened that we deliver a batch of these disc insulators. It's like a batch of cookies. Yeah. The entire number of pieces that are made from one mixture of this epoxy, for example, till reached our final assembly factory. Then maybe we find one of the occasional cases, that one of these pieces has not been ideally or perfectly cast.
Maybe it has a little air bubble inside or some impurity of the raw materials. For whatever reason, there's just some percentage of error that is unavoidable. That right? Yes. Yes. And then we would step in with our new ideas and say, okay, now we need to identify exactly the one that has this issue. It has a serial number. But we would like to know, are there any brothers and sisters of this product. Which means other components which came from the same batch of raw material. Right.
Which may be at risk. And this is where we step in now. Having each and every of these insulators marked with a dedicated serial number means they are all identifiable. All we have on stock, maybe several hundred. We can exactly tell which is which, and from where it comes, when it was produced, and which others had been produced at the same time from the same batch of material.
It reminds me of looking closely at DNA, what the make up is of a larger entity, by looking at very specific chromosomes that tell you how something is made up, what that recipe is. Yeah, and that's the point. You can also hear a bit, you brought the analogy of a supermarket. Yeah. And we all know in the supermarket the parts have their barcode which identify the material. So the barcode contains a material number. And this material number in a computer is linked to a price.
And then they scan it and the computer knows the price of it. Right. Now we would like to know even a bit more on our product. We want to know which product is it or which type of disc spacer or disc insulator, for example. We would like to know who manufactured it. We may have several, several suppliers. Yeah. We want to know when it was produced. So this is why we created something that is like the barcode in the supermarket. it’s a matrix code. Indeed.
It looks a bit like the QR code people know from their mobile phones. Yeah. And that contains a bit more information. It contains supplier of a part. It contains the manufacturing date. It contains the serial number which exactly identifies which is which. And that is our basic feature that we need later to identify material to relate it to neighbors, to capture it and to do and to work with it later.
And also, I would assume, to make sure you are constantly trying to evolve these materials and make sure they improve. Absolutely. I mean, it's all about data. Yeah. Once you come across an issue or a failure of a component, you would want this to be registered. You would want this incident to be registered to the material code, to the manufacturer, to be able to aggregate, to do analytics on this data, to find out, like a Pareto, what is the biggest contributor to issues.
And these are tackled first. I mean it's not about the analytics. It's not a goal as such. It's just to guide a bit. Where would you hook in and do your next innovation project or improvement project? Right, it provides a foundation? Absolutely. Yes, yes. So it's safe to say your factory is digitized. Does this mean that every single item in it can be tracked from the moment it gets there? And is this true for most production these days, not just at high voltage?
Well, I introduced the disc insulator. Yeah. There are also other parts which we don't track individually. I see Like nuts and bolts, for example. You get a box of nuts and bolts and they don't have individual serial numbers. And is the hope that you will start to track those items. Well, some is really not needed because their design is not decisive of the final functionality, or– They're not as key to the efficacy of the larger whole.
Yeah, because maybe they have so much safety factor where they are used that in the full tolerance range that they could be, they cannot create an issue. Right. Right. That makes sense. So and this is why, we have material that is only counted. So how much of this do we have on stock.
And there is this material which we are focusing on now where we really want to know what is the exact identification, what is the exact detail of what we have on stock in order to really track and trace it back to their manufacturing process. In this industry, with its very stringent safety measures, I'd hope faulty products don't occur very frequently. I'm sure we would all hope that. But if this were to happen, how quickly can something be traced back and fixed?
Yeah, that's a bit coming back to what I said before. We have the possibility, once we identify a part that has had an issue. Yeah. As you said, hopefully not too many. Yeah. Right. And we can associate it to a production time or to a production batch and then on a fingertip, find all the other serial numbers that belong to the same batch. And that's what is tracking and tracing. So first we track what is faulty and what are the others that might be at risk.
And then we trace where did the others end up. And so in your experience, what is the sort of range of time that a process like that can take when it's an easier problem to fix? How long does that generally take? Now in this digital world it takes five seconds And that’s across the board, even if it's a very complex, faulty component, you can do it that quickly?
Yeah, because it's a database behind and you can say, give me all components from the same batch and the database spits it out, if you like. And what if it's a component that hasn't been digitalized yet? You would take the elevator or the lift, go to the archive, find the right binder, do your thumb cinema and try to find all the other test reports that belong to the same batch, and then identify what might be the other components, which we also have to withdraw.
Yeah. So it could take quite a bit longer. It could take weeks. Yeah. Right. And in my experience I have had experience like this where we had an issue on the customer side and we had to identify which other products are at risk. And it took us really weeks just to find out which are these other products and where in the world are they installed. Right. This is, I think, a good opportunity for me to ask you about what a Digital Passport is. Yes. So I've heard the term used, Digital Passport.
What exactly is that and how does it work. You know, just consider that these very critical parts have a passport like our travel passport. Right. That is a data record. It's not a piece of paper. It's a record stored in a database that identifies the part, like our name and our number on our travel passport identify us. It may also contain a couple of attributes like, color of eyes or my height. Yeah. Right.
And it may also collect events during lifetime, like visas, or stamps, travel stamps when you use your travel passport. Yeah, right. So basically this data record accompanies the product like a digital twin a bit if you like. Accompanies through the entire lifetime. Sort of like an ID card that you're always holding in your pocket. Yeah. From the moment you're created to the moment you pass on.
Yes. And you can link it – the Digital Passport System is an add on, if you like, or complementary to what I introduced at the beginning, the enterprise resource planning, because we still go by the main requirement single source of truth. So each and every information should only be stored at one single point. So we have a bill of material that remains in the production planning tool.
Yeah. But we have now this individual tracking and tracing information for all the quality data we collect during assembly. We make a measurement somewhere. We see, does a product do its job? Is it fast enough? For example, the circuit breaker, is the opening closing fast enough? We may measure this. And by capturing all this measurement data in this Digital Passport close to the product, we can compare it to their siblings. We can identify.
The simple thing is, is a certain measurement inside or outside of the tolerance band. That's the usual. And you get a yes/no decision, which may allow you to move on or hinder you from moving on. But you could also see am I trending in a negative direction? Maybe it’s still good, but I'm already trending in a direction that is not good if it continues like this. Right, yeah. So I imagine you have to be very aware of patterns that may occur. Absolutely. But you need the data.
Yeah. Like with the cookies. Yeah. They may be still good. But you see they are becoming browner and browner. And at some point of time they will be burnt– So... If you don't control it. Right. And if you don't intervene. Absolutely. Yeah. So how much of that intervention is digital is left to a digital world or AI? And how much of that is human intervention at this point?
I would say at this point of time, we collect a lot of data and then we make downloads and work with it, with analysis tools. But we are just at the beginning of a journey, and that may include more and more also artificial intelligence. Right. That we see the patterns in the recorded data, maybe with self-learning methods, maybe identified as going into the wrong direction.
Yeah. So we may find out, or the machine may at some point of time have better means to find out little deviations than our operators have that compare one data with the other., right. Right. It's hard to know at this point which will be more effective. Is that right? We don't know. But we know, example applications in other areas which we are now looking at and see, where can we start, where can we set pilots.
And that's a super exciting journey because it opens up absolutely new possibilities to be earlier in the entire case, to not clean up after something has happened, but to identify issues before they reach a customer. To avoid– Yeah. –catastrophe or mistake or... Absolutely. Because whatever we find in our factory, the repair of it is much more convenient and cheaper than if we have to do it at a customer side. Yeah. And before it's maybe had any interface with people and the public. Absolutely.
You've already touched on this a little bit just now, but why is it so relevant to go fully digital? And are there any downsides we need to be wary of? Are there things that should not be digitized in your mind? First of all, it's speed and data consistency. Speed is obvious. We talked about it, how fast you can get relations to other components to other material, for example.
Consistency is about whatever is linked digitally and like a measurement device that uploads the measured data digitally. You don't do typos. Right. Though, rather than typing what you measure with a ruler, you type in the measurement value. But if that ruler has a digital interface, that cannot be lost. So that's data integrity. But of course, you name it – is there a risk? Yes, of course. Big data always has a certain risk if it falls into the wrong hands.
Or if there is an opportunity to step in and manipulate. Right, because it is important to be trusted and to have trust in the data collected, is not manipulated, is really the true unmanipulated data– Is factual and not sort of– Absolutely, yes. Yeah. And– Is not sort of laden with bias. Yeah. And of course, we don't want to disclose to our competition if we have had issues with our engineering drawings and, and, and so on. There's a certain level of privacy you want to maintain. Absolutely.
And this is why we put a lot of effort on using industry standard basic tools like the database in the cloud. But, with the best technology behind and with the best in class suppliers that guarantee intrusion protection and all this cybersecurity activities so that we don't have to invest and develop this our own. But there we hop on and work with partners who know better.
Yeah. And so if you were to give us a takeaway on why you feel that Digital Passports are important for us, across the world, across the board, what would your simple way of telling us about why they're important? That's a good question. Often I hear this question, why do our customers need our Digital Passports? My answer is our customers need reliable products. They want to sleep well. They want products that don't cause trouble.
And the Digital Passport is a tool for us to improve quality, to get feedback for new development, for improvement of products and by this, make sure that we can offer the best reliable products. So it's not about disclosing all the data to our customers because we use it, we improve, but we are sure that we can react faster on issues that the customer might have. We can do better for our product development and so on and so forth.
So it sort of sounds to me and correct me if I'm wrong as though the Digital Passport really provides more support to the consumer. Is that right? Yeah. First of all, to us as a producer, to make better products and to be faster in case help is needed. Yeah, yeah. We have some side effects. You can download the product documentation from this system. So there are few benefits also for customers. Right. But basically it's for us to be a trusted quality provider. Right.
To be recognized as a quality leader and also be transparent and demonstrate we can show you all data through the production process. We don't hand it out because it's also sensitive. Right. But we are transparent. You can see it and you can trust that we work with it and that we work on improvements. Right. There's transparency within the company. And then that has a onward effect to the consumer and the client. Exactly, yes. Yeah. Yeah. Thomas, thanks so much for joining us today.
You’ve really helped us understand the ins and outs of maintaining quality standards at various stages in high voltage products, and how this process continues to evolve. Thanks for tuning in to this episode of Power Pulse. Until next time. And that's it for today. We'll be back soon with some more great content. But before you go, remember to give us a follow so you don't miss an episode. Thanks for tuning in. See you soon! This episode was brought to you by Hitachi Energy.
Created and introduced by Bárbara Freitas-Daniels. Content and script writing by Cassandra Inay. Guest speaker, Thomas Haas. Hosted by Sam Dash. Produced and edited by Creative Chimps.