 Live from Las Vegas, it's theCUBE covering HPE Discover 2017 brought to you by Hewlett Packard Enterprise. Hi everybody, welcome back to Las Vegas. My name is Dave Vellante and this is day three of theCUBE's live wall-to-wall coverage of Hewlett Packard Enterprise HPE Discover. This is theCUBE, the leader in live tech coverage. We have a little reveal here. JR Fuller is here. He's the Global Business Development Manager for IoT Edge Line at Hewlett Packard Enterprise He's joined by Doug Smith, who's the CEO of TechSmart Gentlemen. Welcome. Thank you. Thank you for having us. All right, lay it on us, Doug. What is TechSmart all about? We're gonna have, like I say, a little virtual reveal here. Sure, and first of all, thanks for having me here. You're very welcome. And TechSmart Chemicals is a 50 year old company located in Glean Park, Texas, which is right on the Houston chip channel outside of the city of Houston. We are a manufacturer of specialty chemicals, one being DCPD, which stands for Dicyclo-Penedaine. We have been making significant capital investments in the physical plant over the last 20 years. And about two years ago, we realized we needed to move forward in a control system, a new control system initiative at the plant, as well as a baseline mechanical integrity initiative. And so we're a small organization of 53 people, and we look to our contacts and got in touch with HPE and started a conversation. We don't have a normal client-customer relationship. We have a partnership of people, HPE people, TechSmart people, so. So, Jared, pick it up from HPE side. So, you guys have made a big push into this whole IoT business, and you need partners like that program. So, it's kind of interesting the way we got started. You'll probably remember last year we had the big pump, the pump demo, the flow surf pump demo. So that was a project of mine, and Doug had heard about that from a mutual friend and gracious, very gracious of him. He invited us to come out at TechSmart and actually install that at his facility. And he said, I got this bug pond over there, you can put that in there, and then you have a production version of that because we had the proof of concept version in our lab. And I said, that is really nice and very sweet, but no, let's figure out what we can do that will really benefit you because that won't really benefit you. And that started a dialogue that's been about a year. We've been talking about this, and I think it was in August, I proposed to him and said, what do you think about doing a refinery of the future? And his words to me were JR, I don't know what it is, but I love it. And I said, well, let's figure out what it is for TechSmart and let's go from there. And that's kind of how we started the genesis of this entire journey of what we're doing. So you kind of laid out the vision, which is fantastic, so that you're North Star. And then just for the audience's benefit, we weren't here at Discover, there was this amazing floor exhibit, and it was pumps and tubes and pipes. Machine learning and yeah. And it was all kinds of data that was flowing through there, and sort of, I guess a digital twin, if you will, of the factory floor of the refinery. Well, of a plant, yes. And that's a great segue into TechSmart and how we have synergy between our two organizations is that TechSmart, although a small chemical process facility, we have all the equipment that the huge companies have. We have boilers, we have pipes, we have distillation columns. And we need to move forward with our people to instrument, to gather data, to do data analytics on the edge, to have a connected facility with Wi-Fi capabilities. So that's where the conversation started. So much of the data, maybe most of the data today, or historically anyways, analog data, is that correct? It is a combination. So what we are doing, once again, we're a small organization, we have one IT person, and that person is contract. So how we are approaching it is TechSmart stays in the chemical, we use the analogy of swim lanes. We are swimming towards profitability in the chemical business. HPE is swimming in the lane with technology, and then we're working together on this voyage of discovery out here that we're figuring out along the way. Yeah, and for sure, you're not IT, you're operations. Yes, sir. And you guys are IT. Exactly. And so talk more about the partnership. What is that all about? People. It's totally about people. It's totally about people, and it's interacting with each other, it's showing up every day. It's working towards things. It's when you do run into a problem, and Doug's got a great story of when we had a problem. When you do run into a problem, you have the mutual goal of how to solve this problem together. In a typical customer-vendor relationship, there's some kind of built-in tension that's there, and you're worried about, oh, the vendor's trying to do this to me, or oh, the customer's trying to get something from me, and we don't have any of that. We actually have a very solid partnership. And occasionally, if one of my team, or one of his team gets off track on that, we bring them back to the fold and said, no, no, no, we're plowing road here. We need them to cut trees. We need us to cut trees. We all need to be heading in the same direction. You can't stop and go, how come this isn't paved? Because it's never been done before. And it's that shared objective of the refinery of the future that you're working towards. Can we describe in a little bit more detail the refinery of the future? Sure, let me just jump in on that, because in this voyage of discovery, with these conversations, we talked about what do we need to achieve the goals that we want. And so first there is the hardware component. What do we need here to achieve these goals? We'll just take the example of the pump. The pump is the heart of any process facility. If you have a critical pump go down, it can put you out of operation. There's a cost associated with that. And so what we need to do, there's a cost associated with putting wiring from our control center to a national pump. If we can have a wireless network and a sensor on a pump, we eliminate the cost of physical wiring. So the wireless network was provided by one of our content partners, Aruba. And so that is installed. We are working to- Do you know those guys? I do, I do, they're great. Then we're talking about, well, what do we do with that data when it comes in? So we have two edge line servers in there and we have one in our control room and then we have one in, it's super, they have one here on the floor here at the- Discover, yeah, the micro data center, which is for our place, everybody's like, oh, it's fantastic. So yes, sir. And what that does, and so we have the, I'll just give you an example. So we have our old system, the old server over here, size of a refrigerator. And I have used this numerous times when explaining the project to people here at Discover is that I have to explain what we're doing to my 81 year old mother. And when I say we have a refrigerator over there that used to run the plant and now we have this one little thing the size of, you know, a little tablet. She goes, and it saves money and it increases efficiency, she gets that. So those are some of the phases of the project and now I'll pass it over to JR because we've then identified how are we going to use this cool hardware to achieve objectives? Yeah, so when we look at the refinery in the future we actually have a three phase project, right? So everything, you don't boil the ocean, you bring it down into things. So phase one for us was putting the Aruba Wi-Fi network out in the entire refinery, in the entire facility. So we've done that and because it's a petrochemical plant it needs to go into a special enclosure. So we had a partner with Xtronix out in the UK that creates this protective enclosure. Like militarize them. Yeah, well it's actually even beyond that. It is, okay. There's in type one, div one environments there is the potential for hazardous gas to be out in there. And so electronic equipment with sparking and things like that and gas that can explode not a good combination. So these div one boxes make it so that if there is an interaction with a spark and some flammable gas and there's an explosion it's contained within that box and not contaminates to the whole factory which would be that plant, the whole plant where it would actually create problems for everybody else. So that first phase was putting those div one compliant Wi-Fi APs out there from Aruba. We're also putting our beacons with our location-based services, the meridian system out there so they can do wave finders so they can get to the right pump to fix it and also their clear pass. So putting clear pass out there so it's a secured network, right? So we don't want anybody to be able to go in there and mess with anything. So the basic connectivity, the security to allow that, all that basic infrastructure to connect to the- Exactly, so that was phase one. Phase two was they had a rack of other people's compute in there and we replaced all of that like Doug had said with two of our edge line EL-4000 converged systems. And so one of those we actually mounted on the control room floor so right out on the edge not in a data center environment not in a temperature controlled space per se, right? And what we consider a data center. And then the other one, we actually did get an HPE micro data center and we put the other one in there. It's secured, it's badge access, only a couple people in the text mark have badge access to actually be able to get that. And when we look at the compute needs growing, that's where they're going to probably grow into is that data center. So phase two was bring the compute. So I called those two, phase one and phase two my infrastructure phase because now I've got what I need to do. Now phase three is really interesting because that's where we're going to start doing IoT stuff, right? So there are five projects that we're doing on IoT. So the first one is predictive analytics. So this is both at the discrete and the process level. So when we talk about that pump that we saw last year, that's a discrete machine. We're doing predictive analytics on that machine. But that machine feeds a process. So how can we predict what's happening on this machine? What's the impact of that to this process? So that's the first one. Can I hop in for one second? So JR is using the example of the pump. And I mentioned the pump earlier, being the heart of the organization. So it's been interesting being it discover for the first time for me. And the way that I have been talking with people, you have people that are extremely interested in the human component and how is it affecting people. And then also there is the critical bottom line. How is it gonna make me money and save me money? So this pump is an excellent example that addresses both of those. So if we have a pump fail, there is a significant cost if it shuts us down for the day. So we're a seven acre facility. And let's just throw a number out for easy math. Let's say it costs us $100,000 a day if that pump goes down. If you have a facility that's 100,000 times larger, just let me pull out my calculator and your math can tell you this solves a problem. From a human perspective, it's just like your heart stopping. There's a risk associated with that pump going down within the facility. Okay, so we're very tight on time. Sorry. So that's okay. So you got five phases or five IOT projects within phase three, predictive analytics. Let's run through them. Second one is video as a sensor. So this is using video to detect things that are going on and using the edge analytics to be able to power that. The third one is safety and security. So these are things like man down, directed response, those types of things. The fourth one is connected worker. And I define this as location based context aware content. So just very quickly, if you have three different people at the pump, one's a operations person, one's a maintenance person, one's a finance person, and they're all using that augmented reality that we saw, they're going to see three different dashboards. Location based context aware content. And then the fifth one is, we're going to tie into the two sister projects that are going on out there with the DCS upgrade and the NEO's Palladio mechanical integrity program and do a full life cycle asset management. Okay. So these are big projects. So now you've got the fully instrumented refinery is kind of where you're at. Now you've got all this data flowing. What happens to the data? Where does it, where does it get analyzed? Where does it, where does it end up? Where do you go from there? Sure. So of course, having the edge line servers there, we're doing data analytics on the edge so we can have real time right there information to help our workers operate safely and efficiently. And then we have this wealth of historical data that we can start analyzing either on premise or off premise to get, to help us do our Probably the models. Better. And then also this is one really cool aspect from a TechSmart perspective is we do a significant amount of toll processing. That means somebody comes to us and says, here Dave, make this for us. And we will run it through our equipment and give them an end product. If we can improve the way we cook whatever or process whatever it is that they want, there is a significant value added to that. So. And that historical data that's in the lake, if you will, lives on-prem, it lives in the cloud, or you don't know yet. Everything is on-prem. The cloud applications that we'll probably use are around safety and security when we start talking about weather and humidity and wind direction and those sets. So bring in some outside data or models that you like. So TechSmart is a single facility. So leveraging the cloud to communicate to other locations and things like that isn't really a necessary driver, although it would be, completely would be for some of the target customers that we want to sell this to additionally. But the vast majority of the data is staying at the- On-prem. Correct? At the edge. The firm's assumption that we've been making that 90% of the data in this world is going to be analyzed at the edge, maybe a trickle some stuff back, some nuggets back to the cloud. But guys, we got to go. That was a fascinating story. Thank you so much. As you can tell, I can yammer a lot about this. Thank you, Dave. I really appreciate it. My pleasure. Thank you. We'll be right there. Everybody will be back with our next guest is theCUBE, we're live from HPE Discover in Las Vegas, 2017. We'll be right back.