Stephanie Wiechers: You're listening to the data edge where we talk about what actually happens behind the dashboards. I'm Stephanie. I'm the founder of Pear Stop. And today I'm joined by Erwin from our team and Martijn from SPIE. And Martijn is working hands on with asset data in the real world. And so honestly, most companies think they have their data under control, but they don't. So today we're getting into what is actually going on. Welcome Martijn.

Martijn van Balkom: Thank Thank you for having me in the podcast.

Erwin De Werd: Welcome.

Stephanie Wiechers: Thank you for joining us. So there's actually three things that I think are very interesting to discuss today. The first one is what's really going on, what's actually broken. Then we're going to go into a bit like, why does that matter? Because we can fix everything in the world, but there needs to be some purpose to it. And then what happens when you fix it? So Martijn, your specialty is anything around asset management and asset data, or that's the the topic that we'll be talking about today. So I would like to ask you what kind of assets and what kind of data are we even talking about here?

Martijn van Balkom: Yeah, so at SP we work in ⁓ multiple domains with the physical assets of our customers. So for example, it could be high voltage lines, could be industrial buildings, utility buildings like offices or hospitals. Actually any kind of thing that is ⁓ man-made, SP has something to do with it.

Stephanie Wiechers: And so an asset is ⁓ something also as small as a bumper or are we always talking about whole buildings?

Martijn van Balkom: Yeah, it can be both. but we talk about the asset at speed, usually the pump or boiler or part of a high voltage circuit or it can be all kinds of things. But we usually bring it back to a physical component that you can see and touch that has a purpose for the users of the objects.

Stephanie Wiechers: Yeah, so the actual real, the real things in the real world.

Martijn van Balkom: Yeah, exactly. Yeah.

Stephanie Wiechers: Yeah. And so you have a lot of these assets. What do you do with them?

Martijn van Balkom: Yeah, so our main activity that we do with these assets is called lifecycle management. So what we do for our customers can be creating new assets, so newly built assets that also can be maintaining those assets over the whole lifecycle. And at the end of the life, see that they are replaced in a renovation or replacement project and that they can keep on giving. Yeah, giving what the customer wants from it. So it can be heat in a building so that the users can enjoy a work workplace or just powering a whole city with high voltage grids so that everyone in the city can enjoy electricity at their home.

Stephanie Wiechers: Yeah, okay, so basically what you actually do is you make sure that we all can live in a pleasant and comfortable world.

Martijn van Balkom: Yeah, that's what we do.

Stephanie Wiechers: Okay, cool. So let's go into those assets. into these heating systems, electricity systems. Why is it important to have, because we're talking about data quality here, Why is it important to have high quality data? And also, what does that even look like? What does that mean?

Martijn van Balkom: Yeah, so data quality is basically the start of our work ⁓ in maintaining those assets or in doing renovation. So as you might know, it's one of the big trends in our market is that we have a very big shortage of personnel. So it's really hard to find technical people that can do these kinds of renovations or maintain the assets. So we want to use the workforce that we have most efficiently. So we don't want to send the technician to a customer with the wrong equipment or with the wrong parts so that they have to come back. The parts have to be reordered and they have to go back to the customer again. So for that, we need to have the right amount of data quality. What asset is there at the customer side? ⁓ What is the manufacturer? What is the type? ⁓ So that we can actually provide technicians with the right tools to ⁓ do maintenance.

Stephanie Wiechers: And what?

Erwin De Werd: Basically, you have then two different goals at the same time. So it's an efficiency thing, but at the same time, like you say, it's solving the problem of a shortage in, well, the right people that you have, right? These both things together.

Martijn van Balkom: Yeah, exactly. exactly. Yeah. Yeah. So the data quality is one of them. And, and then so far we've talked about, let's say basic data quality. So knowing what type of asset is there so that we can do the right action on it. But if you look a bit further and then you can also think about ⁓ trying to analyze asset behavior to optimize the maintenance strategy on the assets. So you think about predictive maintenance, for example, that you can try to predict when an asset is going to break down so that you can go just before that and do the right action so that we don't spend time doing maintenance that is not useful but doing it at the exact right moment when it's needed.

Stephanie Wiechers: Can you walk us through what is the current state of that data quality? In the optimal scenario, I'm hearing is, we know exactly these are all the assets we have and this is when we need to service them and with what equipment. We spend all of our time perfectly. That's the perfect world. Where are we now?

Martijn van Balkom: Yeah, of course, we're not in the perfect world at the moment. So what we see is that the data quality sometimes is lacking or it's so divided over different systems that we cannot find it at the right moment. So even though, let's say, the initial phase when a new building is constructed or when a new power line is put up, at that point, we know exactly what is going going in. and what materials are used. But then as further you go along the life cycle of the asset, then that data gets lost. ⁓ It may be not handed over properly from the people who do the construction to the people who do the maintenance. ⁓ There might be different contractors involved over the years. So sometimes we are doing the job for some years, then another contractor comes in to do it a few years at the customer side. So there's all kinds of handovers of data. that more or less, yeah, break down basically the data quality along the way. And then in the end you are left with, well, we hope barely enough, let's say to do our regular work, but we want to improve that to also improve the services that we can give to our customers.

Stephanie Wiechers: Mm-hmm.

Erwin De Werd: Amartain, maybe I can ask you because you mentioned a couple of times the part life cycle. Can you explain a little bit more where does it start for you? And well, and end is the end of the part life cycle that is just replaced. But where does it start? So what does it include this whole life cycle for you?

Martijn van Balkom: It can be really broad. We like to also be ⁓ involved in some way at the really early stages of an asset so that we can even be thinking about what kind of functionality should be provided. But basically for SPIE, usually it starts with engineering and then from engineering to construction and then to the more or less maintenance and ending up with the renovation or demolishing of assets.

Erwin De Werd: Interesting.

Martijn van Balkom: Yeah. And when you think about asset management, then we always like to focus over the whole life cycle of an asset, because the things you do in the construction or even in the engineering can have a big impact, of course, also on the whole life cycle of the the assets.

Stephanie Wiechers: Yeah. Well, what we've of course been working on together is what are the actual assets under management and basically cleaning up that part of that involvement you were talking about. So all the mechanics come in and a lot of like clutter comes into the database. ⁓ There's a lot of jargon in there, which, you know, doesn't reflect the actual asset in a consistent way. ⁓ which just makes maybe like 20 to 40 percent of the database kind of unusable. Does that resonate with you?

Martijn van Balkom: Yeah, so it is of course what we are cooperation of these past months as focused on. So on the one hand, of course, ⁓ we very much like to instruct our technicians to handle data properly and to register what they see. But ⁓ on the other hand, we also want to help them by not giving them too much administrative jobs and help them with this kind of AI technology to improve the data without human interference. So ⁓ really try to see if we can improve the data using AI and therefore making life easier for our technicians.

Stephanie Wiechers: Yeah. And one thing we also talked about a bit is ⁓ once this asset database is cleaned up, then that's a really like nice thing to have, but that doesn't directly bring the value because then you might know what you have, but then that doesn't mean that anything is done with it just yet. What do you believe is like the biggest opportunity that's there in the market for companies like SPIE? surrounding an improved asset data quality.

Martijn van Balkom: I think most of it comes down to the way that we can advise our customers on how to continue basically. So if we know from a different, some kind of brands of heat pump or some kind of brands of lighting that it will break down more often than with other brands, then we can advise our customer either to do a different maintenance strategy. So maybe to replace it before it breaks down, or we can advise to move to a different brand. that will last longer. That kind of advice ⁓ is only possible if you have that kind of data that supports that. So we really want to first of all get the basic data of what assets are present at the customer. And then of course add it on top that we register all the malfunctions, all the replacements, all the things we do on the assets, make that connection. And with that package, let's say of ⁓ what we call the static asset data and the dynamic asset data. So the dynamic can be our maintenance, our technicians going there and doing an action, but it can also be sensor data that we collect from the assets and combine it all into an advice we can give to our customers.

Stephanie Wiechers: Yeah, so basically that makes me move from just a trusted maintenance partner just to ⁓ a partner who also advises and helps with cost saving, helps with efficiency, just helps along the whole line of the business.

Martijn van Balkom: Yeah, exactly. Yeah. We want to be that, as you say, a trusted advisor that can, yeah, advise our customers for the, let's say, the coming 10, 20 years on what's the best way forward with these assets.

Erwin De Werd: I'll come on the mic.

Stephanie Wiechers: Yeah. Do you have a... This might be a bit of a hard question, but in the end that's going to save them probably cost and time. Do you have an idea of kind of a ballpark where that would land?

Martijn van Balkom: Yeah, that can, you mean the cost savings where they would land.

Stephanie Wiechers: Yeah, yeah, like for your clients, you, you know, when you're able to do, give them actually this advice and help them choose better assets and plan the maintenance more efficiently.

Martijn van Balkom: Yeah, so we always talk about the relation between ⁓ cost performance and risk. So we want to optimize, let's say, the things that a customer wants to do with a building, for example, he wants to have 200 people working there under a nice environment, so nice climate conditions, nice lighting. That's basically the customer's job. And we want to help them to do that cost effective. but also giving the performance that is required from the asset so that it can be a comfortable indoor climate or something like that. And not having huge risks of breakdown and that they, for example, have to close the building because the air conditioning is not working. That's of course is very unwanted for a customer. So we want to advise them on the ratios between cost but also performance basically. and what risks they are taking with us.

Stephanie Wiechers: right, well then in the meantime, I do have question, basically, ⁓ so data quality, this has been the case for many, many years, right? So ⁓ why fix this now?

Martijn van Balkom: Yeah, that's actually a very good question because you could say it's our job for many years to manage the assets of our customers well and to have this data quality up to standards. But what we see now, especially with the development of AI technology, is that the data quality standard that you need to achieve for AI to assist us in our work needs to be higher. So, for example, for automatic handling of maintenance reports or for to existing technician in how to fix something that goes wrong. We want a higher standard of data quality than what we would have wanted five or 10 years ago. So there's really a big focus now in the company on improving this data quality ⁓ in order to build on top of that all kinds of ⁓ services or things or tools for our employees that can really assist them in their daily job.

Stephanie Wiechers: Hmm. Yeah. Yeah. So actually really enabling that AI transition and getting all those benefits from it.

Martijn van Balkom: Yeah, DataCore is really the foundation on top. can build all these kinds of solutions, either, yeah, let's say really functional for the guys in the field that they can look up ⁓ what can be common malfunctions on a certain asset and they help to fix up to more or less management reporting and dashboarding that says something over the whole scope of work that we are doing.

Stephanie Wiechers: Yeah, yeah. And if, well, without you now asking you to give all your secrets to the competition, but if you could give one piece of advice to someone who also wants to do more with their asset data, what would it be?

Martijn van Balkom: Yeah, so I would say, let's say aim both for let's say, involving the employees in the company in getting the data quality better, because it starts with that. If you don't know anything about an asset or if the information is wrong. So it's a totally different asset that has been replaced in the past and nobody wrote it down. Then you really need, let's say the eyes in the fields and the people being involved in improving data quality. And seeing, okay, if I report this now and say that the data is wrong, that will help myself a year from now or to help a colleague a year from now. So that's really important to focus on that. But besides that, yeah, more and more technology becomes available to help with that, to clean up data without human interference. So it's actually, I think a good strategy to aim for both at the same time. So to get employee involvement, but also to look at more or less technological solutions to improve data quality.

Stephanie Wiechers: And basically those will then amplify each other.

Martijn van Balkom: Exactly. Yeah. Yeah.

Stephanie Wiechers: Yeah, amazing. ⁓ Well, all right, I think that's a really, really good check actually on asset data, on the quality. And ⁓ I believe you've given us some amazing insights in what can be done with it. ⁓ companies think they have it sorted, but clearly there actually is a lot still under the surface. So Martijn, I want to thank you for sharing your insights for joining us today. And Erwin thank you for ⁓ in and being part of the conversation. ⁓ For everyone listening, well, I think we found that if you're dealing with messy data, ⁓ might not be the only one. ⁓ we're looking forward to seeing you next time on the Data Edge. ⁓