Customer Showcase: Devil’s Peak

00:00
Speaker 1
We’re going to kick it off with Devilspeak. Very different application to what we’ve had previously. If you recall, last year we spoke. We had. What do we have? Last year we had Doula. Last year up in Johannesburg, we spoke with the Impala Platinum guys. They shared kind of Impala Platinum’s journey, how they transformed that. Devilspeak is of course, a favorite brewing company down here in the Western Cape, and not only the Western Cape, but in South Africa, founded in 2012, has grown from a small microbrewery into one of South Africa’s leading craft beer inside of producers. Amazing facility if you’ve been able to. To spend a bit of time there. So they have really become a big, significant player in the. In the local space.

00:40
Speaker 1
And what we’re going to be talking about this morning with Rob Aubrey, the director of Next Integration, and Richard Bettman, project director at High Tech Processing, who’s sitting in the front here is really kind of what that process looked like, utilizing a couple of these technologies and especially around kind of utility space was implemented over there. So, yeah, if you can join the stage. Gentlemen, welcome to the stage, please. Rob Aubrey and Richard Bettman. Thank you. Cool. Rob, are you well?

01:10
Speaker 2
Yeah. Hi. How’s that working? Yeah, there we are.

01:13
Speaker 1
Okay.

01:13
Speaker 2
Cool, Cool.

01:15
Speaker 1
Richard, you haven’t. You haven’t joined us at an elevate before. This is your first elevate?

01:18
Speaker 3
Yeah, that’s correct.

01:19
Speaker 1
Okay.

01:20
Speaker 3
Good morning, everyone. Richard, you’ve been far too kind to me. I’m actually just the. The project engineer.

01:26
Speaker 1
Oh, really?

01:27
Speaker 3
We have another Richard at the office who’s the project director, but we get confused quite often.

01:33
Speaker 1
But you’ll claim it for today?

01:34
Speaker 3
Yes, now I’ll claim it for today. And we like to joke in the office that I’m tall Richard and he’s short Richard.

01:39
Speaker 2
So there we go.

01:44
Speaker 3
Excited to show what we’ve developed. I think it’s really an interesting problem and we’ve got some great product that Rob’s developed and really excited.

01:57
Speaker 1
Cool. So maybe to kick it off with the. With context is always important or the most important. So maybe let’s kick it off with what you refer to the problem. What is the problem statement and kind of what was the need and the requirement.

02:09
Speaker 2
Should I jump in first? Yeah. So just a bit of background on Devil’s Peak Brewery. So it’s the company’s called Signal Ho Products and they are, as we all know, of course, with the microbrewery or this kind of smaller brewery sector, have been a company that’s been growing rapidly over the last couple of years from, I think when we first went there, they had a packaging line in the car park where a couple of guys are popping a few bottles onto a filling machine and kind of cranking it down, but manually, to now having opened a massive facility or a large facility in Epping and busy building another facility up north as well. So, you know, the company’s been growing really fast. We’ve been partnering with them from quite early on and it’s been an interesting relationship.

02:56
Speaker 2
I’m coming from a brewing background as well and an automation background. So they’ve kind of been utilizing all sorts of skills from us as well, which has been quite, you know, it’s, it’s quite nice to work in process engineering and so on, but they obviously outgrow, outgrew my ability to do process engineering. And that’s where the high tech and Richard came in. And, you know, we’ve started putting in really big, kind of largely sustainable systems. And that’s, you know, it’s a great example of scalability because now we’ve got to scale everything, we’ve got to repeat it all up north. We’ve got, you know, whatever you do now, you know that it’s going to be done again. And so you really got to have that scalability in mind. And again and again, they’ve got footprints in Kenya and all over the show as well.

03:44
Speaker 2
So you know, it’s not something that’s going to be small. So the challenge were looking at here is they first approached me a couple of years ago and said, kind of, you know, come and help us with our utilities, let’s look at what we need to do. And then we started talking about, you know, we’ve got a CO2 plant and we’ve got a huge opportunity. You’re spending a fortune on CO2, you’re generating CO2, but you’re venting that off. Because when you ferment, you obviously generate CO2 and you vent it off. But now we need to capture, we need to decide what’s the tipping point for collecting the CO2, capturing it, processing it and using it within the product. So that’s kind of.

04:19
Speaker 2
We did initial studies, handed over to Richard’s team to say, kind of, you guys, now it’s time to kind of implement and design the proper thing. Let’s, let’s get involved and let’s see where we are so Richard can take over from here.

04:32
Speaker 3
So in the beginning of the year, were approached By Signal Hill, they had bought the secondhand CO2 recovery plants out of the Netherlands and they needed someone to make it work in their site. In Epping, we process engineering house specializing in system integration of various plant and equipment. So I went over there, inspected the CO2 recovery plant, came back and we needed to make it work in their cellar. One of the most important things with CO2 is you need to be able to harvest the CO2 that’s being generated in fermentation. And if you don’t do that, timelessly, you just vent it to atmosphere like Rob said.

05:17
Speaker 3
And so we put in a CO2 collection mains throughout this cellar and started working with their plant manager, Yuck as well on developing a tool which they can use to manage the collection of CO2 to make sure they are efficient in their cellar and hopefully get to a point where they are self sustaining brewery and they generate all the CO2 they need.

05:39
Speaker 1
Utilities in the brewing space, water, incredibly important aspects. Huge, huge benefits for cost savings and. Yeah, yeah.

05:48
Speaker 2
And sustainability and just kind of, you know, being in Cape Town, we’ve got to be in every way, you know, wherever you are these days with all utilities, we’ve got to be very conscious of these things. So, you know, I think that’s, that was key. We then kind of got tasked with. The project we’re going to talk about today is really just the harvesting and kind of optimizing how you harvest your CO2. Optimizing when we connect a tank up, being a bit. Trying to be as smart as we can about that because you can lose a lot of CO2 if you’re not kind of, you don’t know when. And a lot of the breweries just say eight hours after fermentation. We’ll connect the tank up and we’ll start collecting. But you’ve got to kind of.

06:25
Speaker 2
We really had quite a good scope from the clients as well as working with them to optimize it and with Richard’s team to look at what we can do. So. So yeah, let’s talk a bit about the architecture. I mean we used a very straightforward architecture for this in terms of an ignition server on site with a canary server and a flow server, both in the cloud. And we basically incorporated ignition with flow using Canary as our historian in order to look at the data and make it available. Very important for them was to make sure that we could have, you know, we could access it on all devices.

07:04
Speaker 2
Phones, wherever you are in the plant, they use a lot of cell phones and they use A lot of kind of, they’ve made their own, when they started out, they made their own Google app sheets application, which really has been wonderful for us. Very small brewery. But when you start scaling, we now have to move all of that information across to a sustainable solution. So we’re busy with a large flow rollout for them and so, and that’s for all KPI management. But the key was what can we do for this specific need? And we looked at, we can definitely make an ignition application for this. We can incorporate some flow dashboards and reports and tools into that which, you know, allows us to contextualize that data without having to go through a whole lot of, you know, extensive programming.

07:50
Speaker 1
Essentially ignition using the front end client serving capability of ignition 100%.

07:54
Speaker 2
So use the front end clients as serving application. And we didn’t do anything too smart there. We’ve just embedded some dashboards that we’ve developed from it and using some of the, some additional data from flow. So using the contextual and the power of flow to show what can be done, you know, to minimize your development time really in ignition as well as make it sustainable and scalable and utilize, you know, a lot of the time they’ll want to in their meetings, look at the dashboards and have that data in the various different formats and flow dashboards. So that’s what we kind of busy working on now as like a next step in terms of the extended flow application. So that every meeting they go to, we’re talking about, you know, numbers that they can make decisions with.

08:41
Speaker 1
Okay, so fairly straightforward kind of architecture, a little bit of hybrid with a cloud. Nothing, nothing too wild. Seems easier pretty much on paper until you get to the, to kind of these kind of things. Cool. Thanks Rob. That’s valuable. And this, sorry, this is the actual flow diagram from the CO2 recovery.

08:58
Speaker 3
So maybe I can talk a bit about this. So CO2 recovery, there are four major steps. You’ve got the gas collection compressor, compression and drying, liquefication and then purification and storage. So our tools looks at the collection, our collection efficiency as well as the efficiency of the recovery plant. So we’ve got meters installed on the various lines and we measure how much CO2 is generated. And if we see that the CO2 generated is lower than we expect, then we can go and investigate what’s wrong. Is there a pressure relief valve that’s venting? Have they left the isolation valve closed? Is the CO2 compressor running properly? And that’s insight that you can get from data which is pretty cool.

09:50
Speaker 3
So what we did is we took their fermentation curves so we can go to the next one, took their fermentation curves and we put together a theoretical model of how fermentation runs on their plant. And from that we can generate a prediction of what the CO2 captured from ferment would be. And we use a flow meter and a purity meter to assess that afterwards. So just some pictures of the seller on, you can see there, we have certain tanks with like a changeover proximity switch and then also a flow meter and the purity sensor. So that’s the instrumentation we use. Very simple, straightforward. But we can get a lot of really cool insight and useful information from that.

10:37
Speaker 1
Yeah, it’s interesting, the journey from data to information or making data valuable. It’s not necessarily about the more data you can get. Absolutely. But very often it’s maybe a couple of data points. And from those couple of data points, once you’ve done kind of, I don’t know, regression analysis or just doing a couple of formulas and creating context between data points, you can create a whole host of KPIs from a couple of data points.

11:02
Speaker 2
Absolutely. With, with Richard’s models and with just one or two measure pieces, items of measurements, and, you know, we’re not doing anything fancy. We are using a couple of flow meters and a couple of proxies. That’s essentially all we’re using. And then a little bit of operator input signal from the Breen PLC and we’re a four away. So, you know, it really isn’t complicated. There was quite a lot of wiring which we’ve done and teams kind of through the plants as well, which was done very nicely. And then it’s really, you know, with all of these projects, what’s the most important thing is how the people use it. So we. The wonderful thing about Devil’s Peak is that the client’s really been engaged in the project. They’ve been engaged in making the operator the center of the project and really owning it as well.

11:49
Speaker 2
So we built a big kind of bus stop for them, which is basically what they call the operator workstation as well. And it’s kind of, it’s got all the interfaces, it’s got the right information for them, the right screens, the right kind of set points. And it’s really empowered them from having a little kind of, you know, how these kind of. When you’ve got a lot of growth in a facility, you’re sitting at a very temporary structure at the end of the day. So just rolling out a project with something that empowers the operators a bunch more has been really important. And you know that allows operator to kind of feel like they’re part of this journey as well.

12:24
Speaker 1
Yeah.

12:25
Speaker 3
And I think what’s also really interesting to see is that it’s a very simple plant. You know, there’s not a lot of automated valves, complicated controls, control room that’s, you know, air conditioned and everything. It’s, it’s, it speaks to the craft nature of how these guys operate. And you know, you can do a lot with very simple things.

12:47
Speaker 1
Yeah. Operator, you mentioned, operator, the importance of enabling those people, likely the most important people on our, on our operations floor, giving them pre warning, early warning of. You mentioned the alerts earlier. The ability for them to react appropriately, not having to rely on late data or very late data in certain cases too. Manual data that’s potentially error prone due to human capture of data. It’s all about these guys and really how we empower them.

13:20
Speaker 2
Absolutely it is. And also one thing that we’ve kind of. Now we need to get to the point and we have got to the point now where you can you see one little spike in CO2 and the operator needs to get onto it straight away. We’ve got a warning light there. So if they’re elsewhere in the field, they can see there’s an issue, go address it straight away. And that’s really all about why the why we’ve done this project which is really to optimize how much CO2 you capture. And realizing that a short component of missing CO2 data or missing collection on that can cost you a lot. I mean the cost of CO2 for them buying it at about 10,000 rand a ton from what I understand. So that’s correct.

13:58
Speaker 2
So, and you can, if you miss it for half an hour, you can easily vent a couple of tons. So it’s, it’s. No, you know, it’s well worthwhile having the right data and having the ability to make those decisions. And you can’t do that 247 as a management team. You need to have your operators do that.

14:14
Speaker 1
Yeah.

14:15
Speaker 2
Okay. So we’ve put together a couple of dashboards based on that. And you can kind of see that’s an embedded dashboard in ignition which makes it pretty easy for the operator. They don’t have to switch between applications really. You know, just some kind of flow measurement data, some KPIs for certain shifts. We’ve got a shift kind of column there per shift and kind of where they are tracking versus what they should be tracking so that they can optimize to the point of.

14:42
Speaker 1
Rob, I’m guessing this is on your laptop. They don’t have Netflix on the platform plant.

14:45
Speaker 2
Eh, you got the whole bank sheets on there. Yeah. Okay, so now you know all my favorites.

14:51
Speaker 1
So do you kind of access this from home with the appropriate security and VPN and.

14:55
Speaker 2
No, of course. Yeah. So you would have your VPN and you know, we. Well, when you’re walking around the plant, we obviously can access it on any device. But yeah, you know, we obviously tunnel.

15:06
Speaker 1
From a mobility point of view. Did you utilize the epital or is it just the. The. The mobile URL web client?

15:13
Speaker 2
So we actually haven’t utilized the app on this. We’ve used it generally just through the web client. Yeah, you know it’s a. It. It does the trick and it does the trick more than enough, more than suitably for us. And maybe that’s an opportunity for us to.

15:28
Speaker 1
I think the app will be useful if you’re considering for example using the kind of auxiliary functions of the phone, taking photos, using the location of the phone, then the app may be 100%.

15:39
Speaker 2
So yeah, it might well evolve into that. It might well evolve into more utilization. We definitely want to evolve it towards collecting more of the data from various sides of the plant, especially further utilities. They’ve been upgrading refrigeration and various other sectors as well. So we need to get that on and then, you know, having a platform and now they’ve purchased the license for this tool, they can. The wonderful thing about this is that it’s so expandable and scalable to all of the other sectors of the plant. That’s really where we working now do problem solving on the pasteurizer. We can. We start adding all the trends to that and start kind of each time we touch another machine we really get involved in kind of incorporating it into the project as best as possible into their infrastructure.

16:27
Speaker 2
And trying to do that with a unified namespace is the challenge, as we all know. And trying to define these standards as we grow, as they grow with the business as well is really important. So you know, that’s. To me that’s a management. It’s not a management. It’s a. It’s our responsibility as well as their responsibility to help find that and standardize that as best as possible. And with you know, process engineers involved too. It’s wonderful because the resources are all in at play.

16:58
Speaker 1
Yeah, all the SMEs interesting about the. You said the scalability. You, you’re probably only limitation you will be is kind of the resources that you can throw to that server over time if required.

17:11
Speaker 2
Yeah, I mean, yeah, exactly. It’s resources and then it’s kind of just making sure that, you know, everything’s up to date. That’s as simple as that really. For, for a scalability perspective and then making sure that, you know, the, you know, the equipment and whatever’s rolled out at the various different sites is of the same level and standard so that we can talk to it. Yeah, cool.

17:39
Speaker 1
Scatter views.

17:40
Speaker 3
Okay, these are some scatter views of the plants. So we’ve got a view of all the tanks and this is quite nice because the guys can use it to see what products are in which tank, what state are they in. Are they venting, filling, collecting, emptying. And this is I think something that you can grow and build on. You can start using this information further on in the process. You can put in the flash pasteurizer you could put onto this and it’s really cool.

18:17
Speaker 1
And from a design point of view, were there any, Sorry, we didn’t say we’re going to chat about design philosophy. I was just thinking design philosophy. Was there any kind of discussion around use of kind of the CI more situationally aware kind of graphics of use?

18:34
Speaker 2
So yes and no. There hasn’t been a lot of specifics from the client on that. We are aware of that and we have tried to do that within reason but at the same time it’s not a SCADA as such in terms of this application. It’s more of a user interface. And that’s kind of where were working towards making it as user friendly. And it was also based largely on their Bruin system. So.

19:10
Speaker 1
The balance is important. It’s not necessarily kind of 50 shares of Grand Christmas Tree, you know, it’s, it’s kind of the balance between the two.

19:20
Speaker 2
Okay, so we’ve got a couple of different pop ups and alarms that will come up through the process and you know, they need to understand. And we’ve also had to incorporate various different commenting systems and so on. So what happens, how does it go, how does it, what happens in this scenario? And you know, inevitably with these sort of projects you always find many different little nooks and crannies that actually arise. And one of the ones that We’ve found is certain one work processes that kind of arose and kind of highlighted major losses as well. So you know, you always kind of, you scope a project, you think it’s going to be a couple of, you know, days or weeks worth work and then you know, it starts dragging its heels in terms of finding other nuances on how to deal with these.

20:02
Speaker 2
But that’s how the, how the cookie crumbles. And I think we’ve all found that over the years. So you know, we need to incorporate them. So you know, various different ways of updating after a server crashes or PLC goes off and you know, making it as reliable as possible has kind of been the highlight of the day for us or the priority for the last you, maybe couple of weeks just in all scenarios because they’re growing so fast, especially before peak season now they’ve been doing a million times a day, they’ve been doing maintenance of various sorts and switching off this and resetting that and the really kind of testing it to the nth degree. And that’s, that’s been, you know, it’s been good to do before peak season comes up.

20:45
Speaker 3
No, for sure. And I think maybe one of the pop ups is quite interesting to talk about is the operator comment pop up here.

20:52
Speaker 1
Good content.

20:53
Speaker 3
It’s, it’s actually very valuable, such a simple pop up to be able to provide information to the following shift that you may have missed in your handover.

21:02
Speaker 1
And I’m guessing this was a paper process previously.

21:04
Speaker 3
Yes, yes.

21:05
Speaker 1
So kind of a whiteboard, maybe a clipboard, a couple of notes.

21:10
Speaker 3
Exactly. And now we’ve got this digitized and the operator can see exactly when an issue arose. What was the problem? A breakdown. Then it was fixed but now we’re experiencing another problem. And yeah, quite a, I think quite a cool system.

21:27
Speaker 2
Yeah, I think it’s good. I think. Look, let’s have a look at that next screen. So the next screen is kind of, we’ve got various different brands and we’ve got kind of different fermentation profiles. So it’s not anything that’s going to, you’re not going to be able to go and repeat Devil’s Peak Brewery based on that. But it’s just interesting to see, you know, this is the data from the collection and how we’ve incorporated that from the model and how we’ve kind of modeled it and to kind of fairly limited number of set points that the operator or the brewing manager can kind of adjust. As needed. But with this modeling has also come the ability to really see the fermentation curve in action as well.

22:09
Speaker 2
And the lab team, what we’ve been looking at now is how the lab team, when they’re doing their samples on the tanks, can enter those values and then we can kind of calibrate accordingly. We can update the model so that it’s even more efficient. So that’s on the fly. So, so that’s quite useful. And it’s all part of that digitizing the facility so that we can kind of, rather than having the clipboards with the fermentation profiles, we can go towards digital systems like that as well so that everyone can see it straight away. We’ve of course incorporated it. As I said, we’re rolling out a large flow application for them as well.

22:45
Speaker 2
So we’ve incorporated all of the data from this, the collection, the harvest efficiencies, the CO2, plant purity and the kind of the collection of the plant efficiencies and then the overall efficiencies into their dashboards on daily basis as well. So, you know, they’ve started using that this week, which is really exciting as well. So we can kind of see the integration from the SCADA through to flow, you know, really efficient for them, which is great. And yeah, the power of using flow on that at that level is wonderful in terms of getting the production meeting off Excel and into a sustainable solution that everyone can kind of have one version of.

23:30
Speaker 1
It’s not a day late now.

23:31
Speaker 2
These guys were pretty good at Excel. I must say. Most people are pretty good at Excel these days.

23:36
Speaker 1
It’s kind of hairy when those Excel files get to like the 120 meg kind of size. Then it becomes a little bit heavy. Yeah, no, fantastic. Great stuff. Thank you. Any kind of lessons learned things to note?

23:50
Speaker 2
Yeah, look, I think lots of lessons learned in terms of. But the main lesson for me was just how important and how empowering it is to have a client that really knows what they want and wants a solution as well. That was kind of the crux of making this a large success is having the buy in from the clients and having the right team around the table as well and trying to get to that point that your client knows what they can get and knows what they are able to get. That upfront work is important.

24:25
Speaker 1
Absolutely. Yeah.

24:27
Speaker 3
100% agree. We were quite lucky in this case that were building off the back of a similar system that was implemented at another brewery. But I think we closed some good gaps and improved that system.

24:41
Speaker 1
Lovely. And there’s the ability to repurpose a lot of that work. Fantastic, guys. Thank you very much. Is there. We’ve got about 30 seconds left.

24:50
Speaker 2
Any.

24:51
Speaker 1
Any questions for Rob? Richard. Just a congratulations to Robert and Richard. Richard won’t know me, but I know the Bettmans from Sasselburg days when I was a kid. His father. His aunts also know his mother very well. She worked for us for a while. Wow. And we’ve been following you, Richard. We’ve been watching you closely.

25:18
Speaker 2
No pressure.

25:20
Speaker 3
It’s a small world, eh?

25:22
Speaker 1
Very small world. And congratulations.

25:25
Speaker 3
Thanks very much. Brilliant. Thank you.

25:28
Speaker 1
Absolutely great stuff. Thank you. Thank you very much. Thank you so much, guys.