Welcome to Energy Bytes with John Kalfayan and Bobby Neelon. Your essential guide to the intersection of data and energy. This podcast dives deep into the world of energy, shedding light on how data, AI, and technology are revolutionizing this sector. Each episode equips listeners with insights into the most efficient tools and resources, paving the way for a data-driven future in energy. From technical nuances to broader industry trends, Energy Bytes offers an unparalleled perspective on the evolution of the energy industry. Join us as we decode the algorithms of energy, one byte at a time.
Welcome back everybody to uh another wonderful episode of Energy Bites. Uh rad here, John Calfian. My lovely co-host Bobby Nean is out today. His uh son woke up with a double ear infection and as a father of two, I feel his pain on that. So hoping his uh his son feels better. and uh but we're here today with Quinton Jones, senior manager at uh Accenture. Thanks for joining us, man. Oh, appreciate it. Glad to be here. I'm glad to uh you know, we we had a lot of communications and stuff at my when I was at my last company because it made a lot of sense and then we both went to new companies and haven't had a a lot of excuses to to hang out as much. So, I'm glad you could uh come on and chat with us. No, for sure. For sure. It's always nice to take a half day Friday. Absolutely, man. That's uh Yeah, that's that's why we we record every other Friday because that's when Bobby has his 980s off. So, one of the the perks of the industry, right? Let's jump into it. You know, I've known you for a couple years now, but uh let's kind of back up. We were just talking about this before we started, but you know, our first question that we normally ask people is kind of, you know, how did you get into energy/andor tech? Whether it's, you know, like for me growing up, I was destroying my parents' computers and, you know, hacking my way through stuff and and downloading, you know, key generators and all kinds of dumb [ __ ] that was breaking my my uh home P, my Gateway 2000, uh, to put a a time frame. on that. But, uh, you know, how did you kind of initially get into because you're super handy and I know you love tinkering and and messing with stuff, but how did you kind of get into energy and then andor tech so kind of backing up. So, we could go probably three hours long form podcast just talking about that. I've had a lot of jobs, a lot of jobs across years, but um, yeah, I originally had no plans to get into energy. So, I um, in high school was part of work release program. So, we had like a hospitality program. So, I went to work at the Sherin downtown Dallas and um ended up working in the kitchen and then ended up getting a job like after school working in the kitchen. So, I wanted to be a chef. That's that was my thing. That's awesome. So, yeah, I went to work when I was 15. Um worked there after school, but then also worked at a pizza place and they had a little Italian joint and so yeah, I did that. I just kind of grew up through that path like I mentioned. You know, graduated high school, had no plans to go to college, bounced around a bunch of different kitchens, um, lived in College Station for a while, went belly up, spent all my money, got kicked out of my apartment cuz with the restaurant life comes the other side of that life, right? So, anyway, worked my way around, lived in El Paso, Primed you for the Wheelfield, right? Yeah, it did. Uh, lived in El Paso for a while doing the same stuff. And then in 2000 I moved to Tyler. My dad was living there. So I moved down there uh with him. You stayed with him for a while and then was working in a a bar restaurant kind of doing dual duty. Um met my now wife. Um and so her dad actually owned an oil field services company. So I was going to say when you get to Tyler it's hard to get away from the oil field. Everything's tied in there. Everything. And so anyway, uh she ended up getting pregnant and so it was time to get a real job, make some real money. Uh so I went to work for him, started out, uh installing compressors, so like threading pipe and digging trenches and back breaking work, right? And I was like, man, no way. Like he paid good, but I didn't want to do it. So that's one of those it's like it's good at the time, but when you start doing it, you're like, I can't do this forever. Yeah. No, I mean I mean I was in best shape of my life. It was great, you know, but nah, it wasn't for me. And so I saw a guy putting in uh flow meter, right? And so this is 2000, but we were still on Bart and chart recorders, right? I mean, we weren't even digital yet, right? So it was all the, you know, bending the tubing and stuff and putting it on and calibrating. So I was like, I want to do that that that because they sat in the truck until he had to get out and bend some tubing, right? That's another fun thing about being in the field is you're always observe. Like if you're observant enough, you'll realize who has the really cush jobs that still gets paid either the same or most of the time more for whatever reason. Yeah. You're like, "Hey, how do I do that?" Exactly. And so I mentioned that I wanted to do that and that was an area that services they didn't have was measurement instrumentation. So we hired a couple of guys that knew what they were doing. I learned from them. Uh my first outing was actually running tubing for uh a CO2 plant and uh we ran thousands of miles of tubing, right? So did that for a while and then we ended up getting a piece of work from Devon where we went around replacing the Barton chart recorders with Bristol Babcocks and that ran on old Amco software DOSs. Right. So to calibrate it, it was like hold down the space button until it got to zero. Let go. Okay, it's calibrated. Right? It's it's accurate. So, uh we we uh all across from, you know, Nacadocious Lufkin area all the way, you know, almost to Austin. That's all we did. We would just run around and we had poles and solar panels and put those in. And so, I kind of cut my teeth on digital meters that way. Um then I got offered a job at another firm that was starting up and doing stuff in the field. Um and again, it's all about the pay, right? $2 more an hour. Here I go. Yep. Uh so I went over there and um we were doing a bunch of work. Um and these were our entrance into the first wireless instrumentation, right? So we Saturday and Sunday we would stay in a warehouse and build these big boxes, right, that had a little radio in there and terminal blocks. And then we had another side. So you go and you dig a hole next to the wellhead and you put your tub and case meter on and plug it into that. And then you would have another radio across the pad. So we would go around putting those in. Uh those were RF. Yeah. Yeah. They were just basic 900 megahertz pointto-oint MDS radios and then and then integrated into the flow meter, you know, via Modb Plus, right? Yeah. So did that for a while, and then, um, I ended up getting a gig with still same company but a gig with Encano Oil and Gas and, um, over in just, just a little north of Nacadocious and Travis Pete. They have a big field right there, and it's, you know, 150, 200 wells out there, and they were drilling a bunch and CPFs casual facilities out there, um, and it was great because you just turn in off the highway and it was just pig trap roads and, you know, four-wheeling all day long. And so I just was basically the the fix it man out there, right? I'd get 10 or 15 work orders, just hop around and then go if they were putting in a new well and meters. And that was my first foray into a true SKA system, learning about real comms, back all comma systems. Um, and that was in 200
six seven. And so at that time radiocoms were there was a company that launched weather balloons
and it had a frequency radio on it and so when that weather balloon would come over the field it would absorb right four hours worth of 15 interval information and then post it up. That makes me think of the, uh, there was some I can't remember Google's like skunk works, uh, division; but there was some, uh, I read some thing a couple years ago where they were, they were looking at, like, "how do we do, you know, global distributed internet communications," whatever, and one of the ideas was a weather balloon style thing, throw it high up in the altitude, and then Starlink came out and was just like. yeah that idea [ __ ] all over that right exactly and then Um then that company went bankrupt. So let's so staying with the the SCADA stuff. So your first entrance into SCADA. So back then how I mean how do you what's what are some of the big like kind of similarities between SCADA from then and now? And what are kind of the main areas of you know innovation or differences that you're kind of seeing. And I know it's a mixed bag because people acquire people and you know no one wants to pay to upgrade SCADA once they have it because of the cost and all that fun [ __ ] But yeah, so at at that time we had Signet, right? And so I thought that was the coolest thing ever, right? Because uh I worked for a guy gentleman Charles Lame who's been in the business a long time. He's over at Waterbridge now, but grew up in Pioneer. So he kind of built one of the first early skate systems on his own because you talk about in-house. So we had Signet and he had some good people in there that build the screens, get the data and stuff like that. And so it it was cool because that, you know, I hadn't seen anything else. So that was that was kind of the thing. Um and then over the years you started seeing you know different companies kind of pop up right the Rockwells of the world and stuff like that and it was still like okay. Um, and then going to, you know, like y'all were talking about the other day, then inductive automation came, right? I mean, they blew out of the water the pricing arrangements, right? And then just there, you know, however, the other side of it was being fully customizable. You have to have a dev team, right, on staff, monstrous to support that, right? And so, I think they're they're they're coming along, right? There's getting more into the to the drag and drop, but, you know, it definitely has evolved, right? But today, you know, and and and I can already know people are going to hear this and start freaking out, but SCADA of old is not being utilized as SCADA as today, right? There is no real supervisory control, right? It's just data acquisition, right? That that plain simple data acquisition, pulling it here, pushing it here, sending it there, right? That's it, man. you know, I mean, you know, it everybody still relies on the old Purdue model in that polling response, right? I mean, there's a few out there that have moved to an MQTT broker, but that's not the primary, right? And and it's not it's not an enterprisegrade deployment either, right? So, but yeah, you you see it come along. So I think that you know another 5 years probably it will just be dashboards and then some other type of control component you know localized right HMI side. Yeah. And it'll be drag and drop and customizable without having to have knowledge of coding on the back end or of a video where everything's going with the code on on the coding side. low code, no code stuff is it's interesting. For sure. For sure. So,
so keep going because I interrupted you. So, so, so yeah. So, so right then kind of started my intersection of the tech that because I Yeah, I was same way, right? You breaking stuff at the house, pulling it back together, you know, sat in front of that first Mac and was like, "Oh, yeah, this is cool." Right? And so, I've always kind of been tech rel tech adjacent from that. But I did do a stint at uh at uh Comp USA for a while and so I'd you know build computers off to the side, made some money. RIP Comp USA. Exactly, man. Uh and and and yeah, and then it was the best too because at work we had a T1 line. So I was Napster all day. No one understands like just the the shift the jump in difference from going from 56 even to 256 on DSL. I remember how excited I was to get a DSL line. Oh yeah, that was at first and my parents were pissed because of how expensive it was. I was like, but mom, it's five times faster. Like I can load an image in two minutes instead of 10, you know? Exactly. Exactly. So, so anyway, so uh yeah, so we had that system with the air balloons until they went bankrupt. And so we actually had two weeks to build a new network out there. And so that's when I first learned about microwave, true microwave. And uh anyway, we put that out there. Um and then I became kind of the the guy that would go to the new areas of development and you know, set up all the infrastructure, find local teams to, you know, in techs, get people hired to work with at small sis. So, uh, we still were living in Tyler, but I was going to Tannesville and spending, you know, 10 days at a time, 14 days at a time, all that stuff, which was a complete shift, right? Because the typical instrumentation that we used on, you know, Travis Peak wells when you had, you know, maybe 60 pounds on Casey, right? If you had pressure, lots of vacuum wells up there to go into, you know, the first Hannesville well that we had and 10,000 lbs of pressure. Yeah. People don't appreciate uh pressure nearly as much as they should in No, I I'm lucky to have my hands and fingers, right? Because and I was trying to teach someone how to take a manual gauge out and put in an OEM tech wireless. You know, that was kind of when we moved into real wireless. And I'm, you know, sitting there and I'm like, "Okay, you know, close this valve, close this valve, and, you know, take this down." And, uh, they had just greased it. So, the grease was stuck in the autoclave fitting. So, it looked like it went to zero, but it still had,000 pounds. Anyway, so yeah, it shot out. So, lucky to have my hand down. Um, so anyways, went to Hannesville, developed the Hannesville, went to Mississippi, did the TMS wells, went up to uh high pressure, weren't they? Yeah. Yeah, it's crazy. They were nuts. Then went up to uh Oklahoma and Kansas to do the water oil wells uh out there. Uh spent a lot of time in Pahusca, Oklahoma. Shout out to Shout out to Pahusa. Shout out to Pahusca. Yeah. Love the small oil field towns, man. Yeah. and Bad Brad's barbecue out there, man. Kept us in in good shape. Um, so anyway, did that bounced around uh personal life, we now have two kids. I'd been on the road nonstop, right? And my second son, so there's five years difference between my older two boys. So I guess my my middle son now, my middle son now was about five or six. So my wife was like, "Enough, enough. It's it's it's time, right? Come off the road." You know, because when I would come in, I'd be exhausted, tired, or you know, whatever. And um you know, the life. So So anyway, um at that time, um uh Jonathan Klein as part of summation was calling on me to actually bring them in to do some work for for me. So when we were in the Hanesville, we had a joint venture with Shell, right? and Ken and Shell had those wells and so they were doing Shell's work. So they were calling on me to get my work. I ended up flipping that into a job interview. Came to Houston, interviewed, got hired on. Um and at that time we had moved back to Tyler and it was pretty much just going to be that, right? I was going to be remote. I'd come in if I needed. Um and then it became no, you need to live in Houston. So it was like okay, we'll do this five years, right? Come to Houston. 5 years we're out. That was 11 years ago that we said that. So came in to work with Simation. It was a coolest company, man. We were doing a lot of innovative at the time stuff um around upstream and downstream. Um so I started just as a guy. I knew Fisher Rocks really well. So I kind of started that guy. Then um we ended up doing a project for another client that wanted to move into, you know, autonomous operations into a CPF, right? so he could take itself down, bring itself back up. And so I designed that with another colleague of mine and and we put auto chokes and you spent a couple million dollars Yeah. getting it done. Um but along with that comes if you don't if you don't teach your people, it does no good, right? And so it didn't matter that we had all the automation in the world. They were still driving out there every day walking around manually looking at everything to the point where they built a 10-ft fence, you know, around it locked, right? Cameras knowing when who's going out there and doing what. And uh yeah, but then when you start talking about scaling that, right, money dries up pretty quick. So so anyway did that um and then uh I ran um what they called their industrial networking component. So which was like the network guys, and then cyber security. Cyber security kind of came out of that, um, and and and interesting enough, that was part of the reasons Accentra acquired us, right? So, um, so I guess that's the next part of the story. So, come up through the the data track from there, right? All I've worked all the way from the sensors up to the backall networks, the SCA systems, the historians, the kind of next generation of stuff, uh, then became IoT, right? Industrial IoT, u, and that was going to be the next craze. So, I did that for a while, and then, um, and then yeah, and then one day, uh, Jonathan called me up and said, "Hey, you know, come into the office." I got a project for you."So, I did and and he said, "Okay, come Monday because I need you to put some nice clothes on and go to this office over here and you're going to help this team that's uh called Accenture that's trying to do some work."And I was like,"Okay, cool." You know, great, whatever. So I show up there and and they were trying to basically take over all OTIT operations, right, from the the back end. So basically what they were going to do is company X would just fire all these people. Eccentric would just rebatch them underneath the project and all that. As part of that, they wanted the field component piece, the field networks and all that, which Eccentra at the time didn't have anybody that that knew that. Um they didn't even know, you know, putting on FRS what that meant, right? So, I kind of came in to play that role. And about six weeks into it, I showed up one morning and little three-man team I was working with was like,"Hey man, welcome to Accenture. It's great."Da da. I'm like,"What are you talking about?"And they showed me. So, they had released messaging internally that Accenture had acquired Simation, but no one else. But, but no, but yeah, doing it nowhere else. So, I go outside and I call Jonathan. I'm like,"Hey, man. What is this?" It's crazy. And he's like, he's like, "Dude, do literally do not say anything to anybody. Like, we're trying. Like, the ink isn't even really dry yet, but you can't can't let that out, right? We can't have this mass, you know, thing happening, right? Like people are going to lose their [ __ ] So, uh, so yeah. Anyway, I did I I kept my mouth shut and just kind of watched as it as it happened. And luckily enough though, I'd already kind of been embedded into the eccentric world. So I made a pretty nice network and and met a bunch of people and kind of had developed myself with street cred. Uh you know, kind of know what I'm talking about, know what I'm doing. Um which was great. And so yeah, so then then all the real change at first they said they're not going to do anything to us, keep all our office, all this stuff. Yeah. About two months into it, they were like,"No, you're integrating."And how the how the world just changes, right? because we were a pretty flat organization, you know, none of us were, you know, better than the other, right? We had a seuite, but you know, after that it was pretty flat to very structured. Yeah. With inside of Accenture, right? The hierarchy. Well, I mean, even this morning we were talking about it. You I think you went to the bathroom, but Colin was like, just to put that in perspective, you know, Exxon has 70,000 people. I was like, and y'all have 10 times the amount of like trying to like the just conceptually thinking about how to try and manage an organization like that is insane. It's so many people. It it it it it really is in and it it's a shame in some instances, right? Because there's so much untapped talent, right, that you know to be on a project. But then again it also helps because if you establish yourself that way you're the first one thought of right so it it it helps in that way right so like you know I get pinged you know we have teams right teams is my the thing so you know I get pinged you know out of the blue somebody's like hey I talked to this person who knew this person and said you worked on something similar yeah you know can you share some information around that I was like yeah so uh so it's cool so we did and so once we kind of got in there. Um, you know, I kind of started seeing where it needed to go, where to jump. Um, which aggravated the crap out of my COO, which it was funny. Uh, but so one group became, um, the cyber security, so OT security group, which is a huge practice inside of Accenture now. Lewis has built that pretty large. Um, a lot of I'd say another, you know, 15% left pretty quick, right? Because they didn't they knew of that world. didn't want to be a part of that world. Um I actually moved into Accenture Technology a part of ecosystems and ventures which is our partnership all of our alliances right so our business groups are in you know VMware partnership HP partnership Dell partnership moved into that group um and became kind of the incubation lead looking at all the new stuff. So I worked with our investment side looking at startups and then bigger partners and the startups they you know that they were looking at Um, and then that became edge work, right? Because edge became the the next new thing and IoT was old. And so I started really understanding that world, a lot more of the IT stuff that's out there. Um, and then in resembling because that was another big thing that you see with a lot of the silicone startups, right? They don't have any of the energy background or the knowledge or you know what problem they're trying to solve. They can configure a Kubernetes instance like that. They have no clue what a wellhead is for sure. Exactly. Nor and they can spin it up, but they can't explain it to
to a normal guy, right, of like Cooper, what u and and and that. So, uh, so anyway, spent some time working with ecosystems, a lot with the heliler Packard Enterprise, um, and their, you know, kind of edge, uh, movement and that, you, the stories; they had lots of great people over there. Then, um, in 2019, Jonathan Klein, who was the original founder and Coation, his non-compete had ran out, so he came to me and he goes,"Hey man, what, what do you think about, you know, starting a product company?" I'm like, I don't have any clue about how that works, but in based on your history, you know, it sounds great. You know, I'm down. This was in August of 2019. And so, we talked it out. We got, you know, figured out what, you know, how it was going to work and, you know, kind of the the the package and stuff and what my role was going to be. And, you know, because we went through several iterations and different things. And then uh I turned in my notice to Accenture and said,"Hey, I'm going to go try this thing."They attempted to talk me out of it a little bit, but I was like,"You know, I'm going to do this." They said,"Well, okay, it's fine."So then I uh moved on to uh Technore. Um and so we started out as an AI ops company. Um and he still has the company and they're flourishing and they're more in the ops intelligence world. So uh very low code uh interface to create machine learning models and solutions right and run on small embedded devices. Um they've got a couple of clients and continuing to grow that put people on. Um wish them the best. I keep track of them and introduce them anytime I can. Yeah. Um so that was uh in 2019 was great. We were having great funding discussions and then 2020 hit co then co came and and changed the world, right? And so it was still pretty good for us because again it gave us kind of time to to build, right, without having a lot of back stuff on there. Um and then uh and worked with you know the first client that we had and on boarded around machine vision and um then we thought we this goes to another guest you had on. So we thought oh we'll just take offtheshelf components and just put them together and we'll have our own box right we're our own solution. We don't we don't need anybody to make anything for us. So we got, you know, cameras from Axis and, you know, Jets and Nanos were going to be our board and, you know, we just put it all together, a little modem in there and everything will work. Except it doesn't just work like that, right? Like you can't just take a telet module modem and stick it on a nano and make it work and the firmware and we had our own version of Linux. And so it became a lot of screaming matches between me and the dev guy, the software guy, which was in Sweden, because it was always pointing fingers, right? He's like,"Ah, Quit, your hardware doesn't work."And I'm like,"No, your software doesn't work."You know, back and forth. Oh, it uh yeah, it conversion at its best, 100%. And um and so anyway, learned a bunch of stuff there. Um and then um a little over a year ago uh CTO for uh Cloud First, which is a part of Accenture, said,"Hey, would would you ever consider coming back to Accenture?" And I said,"Well, you know, listening." And they were like,"Yeah, you know, when you were here a couple years ago and you were talking about edge, we get it now. We get that not everything's going to be cloud-based, right?"So, um so anyway, worked that out. Um and it really made sense. you know, I took Jonathan uh you know, he really should take my salary and go hire more developers, right? I really wasn't doing what I should be doing, right? My whole role was supposed to be like networking with the community, right? Bringing in new developers for solutions, taking it to market, and we just weren't to that point yet. Um, and you know, he he felt like, you know, I I was doing what I needed, but you know, I just I just didn't feel comfortable. Um, so anyway, came back to Accenture and been uh been doing this. I spent the first 6 months uh working in retail and even though I know absolutely nothing about retail and didn't want to know more about point of sale systems, that's for on the edge side. Mhm. Yeah. No, I think that's uh I mean my I'm kind of right there with you on the uh some of those auxiliary industries or whatever. I had no exposure to any of that until I started working at Hive Cell on the Edge side. And you're like,"Oh yeah, actually this edge thing makes a shitload of sense for retail, especially if you're doing millions, if not hundreds of millions of dollars in transactions and stuff."And then, you know, a network the the server or the server in your building or the even just the network, you know, the network cable gets cut to the building by some random utility person doing something completely unrelated. And then it's like, well, if you're Target or Chick-fil-A, like I know Chick-fil-A has a ton of they have Nooks. Nooks, they have Nooks in pretty much every store now. But, uh, you know, having those point of sale systems, like everybody's like,"Oh, yeah, it's just in the cloud."It's like, well, what if you can't access the cloud, then if you're Chick-fil-A, you have hundreds of pissed-off people, whoever it is, right? And so, it's uh I like I said, I had just never I had no reason to think about it, but I had never thought about it. And then you're like,"Oh man, that actually makes a ton of sense, right?" Yeah. I mean, exactly. You use Chick-fil-A. Chick-fil-A is perfect example of QSRs, right? Quick serve restaurants that have it figured out. Yeah. But then they also, you know, recently, you know, so because there was been several articles, right, of the Nooks and the way they have that all set up and they actually have two for redundancy, but they even looking back at themselves and saying,"Okay, yeah, we're good. We still could be better, right?" Chick-fil-A's operational, like, efforts around"how do we make this as efficient and, like customer, experience focused process as we can"is just, like, it's topnotch, you know? And, like, because that's the thing people don't realize: that, like, these guys are use, you know, you've got your phone, right? So then you submit your order on your phone. Now they can monitor when you leave to go pick it up, so now they know,"Okay, submit that into the queue wherever it needs to go,so"so that the guys can cook the food because it looks like, based off the route, you're going to be five minutes away. Then you've got the, you know, cameras around the building that are pulling data in and in the Chick-fil-A app, you tell it, are you in a car and what color and all that stuff. So now they can start identifying, hey, this person is probably in line over here just based off the computer vision. So make sure that [ __ ] is is in production at at very least by the time they pull up. And there's all these things that just, like, it's very operational efficiency driven that you know the average person would have no clue about, right which? is I think that's how you know technology is, like, really good is: the consumer doesn't even realize it, right? Like, hey, it works, and it's a good experience; that normally means that the [ __ ] on the back end is doing its job, right? that's that's exactly it but it's it and it's funny with my wife right because my wife's not tech I'm right there with and doesn't get, you know, it's fine. You know, it's just not our world. But anytime something happens, right, where I'm trying to do, right, I'm making an order or whatever, right? And I'm like, they're [ __ ] blah blah blah blah blah, right? Is it's broke, man. They need to fix like, yeah, come on, like trying to call their customer service and tell them how to fix their stuff, right? And my wife's just like,"How do you know that's how that works?"And I'm like,"I just do. I just pay attention, right?" And it's I had the same conversation with my wife last night on Netflix actually cuz I was I pulled it up and I was trying to find like my watch list or whatever. You know, normally like historically I know they've been messing with how the main home algorithm displays certain things or whatever, but uh like normally it's you know a few rows down on the main page that you land on. And I couldn't find like I scrolled like where the hell is it? And she's like just go to the menu and go down to it's the last option my list. was like, "That's dumb. It should be the first one or the second one at minimum, right?" And she's like,"Well, I mean, I feel like it's pretty accessible."It's like, I know you do, but just from like my perspective, like I'm so into this stuff and just the little things of design are what make or break things so so easily. Um, let's keep talking. I want to I definitely want to talk OT and security and I definitely want to get into the edge stuff. Um because again I think like edge is edge I think is probably coming you know you've got your hype cycle then you've got the trough of dis despair or whatever and then it comes back up. I think we're probably coming out of the despair, right? We had that hype cycle coming out of 2021ish and there were all these edge companies and then of course now with the economy and everything and tech is hurting or whatever. But uh, you know the, the edge stuff to me is so fascinating because there are so many use cases, but damn near nothing out there in most instances is really even like there's not as many people. The bigger software companies or people, like, they're still catching up to the edge stuff, right? Like there, I can't tell you how many times we had customers who like,"This is what I want to do, these are the softwares I want to use,and"and we go, we start down the line and then we get to whatever software and they're like,"We don't have, you know, we don't have a containerized version or we can't run on ARM." And it's like,"Well, you've got a client that's telling you that they want to do this, so can you figure it out?""Like, no, well, maybe," you know? Like, it goes into their queue and it's like,"Well, [ __ ] But so tell, me tell me a little bit more about how you kind of, both just from personal experience and just your thoughts on, uh, how you kind of see see edge playing into the the energy space, or if you guys have done any interesting stuff or looking at any interesting kind of projects in that that realm either currently or in in your past. So I'd say a lot of it is kind of came in two phases, right? Two buckets I'll put it in because again when we talk about edge, right, there's lots of edges, right? Defining the edge is another huge problem, right? So just even that when you come into a, you know, a meeting is a fun thing because if you don't then everyone has their own idea of what it is. Yeah. So, so my go-to decks is I just start there, right? Is from my viewpoint and from an Accenture standpoint, and kind of how we kind of mark, is: Edge is not a technology, it's a topology, right and? And and it's the same way that you know you talk about it's not energy transition, it's energy addition, right? It's the same type of thing, so. I very quickly, depending on who's in the room, if it's OT, I very quickly go to,"Hey, I'm not talking about ripping and replacing, right? I'm talking about landing and enhancing, right?"And that's kind of the... And so, you know, again, you know, traditionally it's been distri...you know, centralized, like from the core, right? Going back to how we were talking about earlier, from a ska system, right? It's all about,"Hey, give me a piece of information, thanks, give me a piece of information, thanks." all goes back to one server whether that's in the cloud or onrem or wherever wherever their core is at right it's going to the core and then it sits there to die yep And then, you know, kind of moving into this distributed fashion, right? As we talked about coming in the data lake, MQTT, C broker, where more things are being shared, um, and and that kind of comes into the edge. But really want to call about, you know, an optimize, which is a combination of both, right? It kind of work those. So, I kind of level set there. And then I go into I've got another slide where I talk about the edge is kind of coming up, right, from a gateway and what a gateway can and can't do on up to a micro data center, right, and kind of how that stands. So then, once you get to that level then, then it becomes,"Oh okay, well I already have the server out there doing this thing or or PC doing this thing,"and it's like,"Well no, now you got..."Now we're just talking about the infra layer right now, right? Then you go into kind of up that stack, right, of the software and the Kubernetes containers, low code environments, all that stuff. so I'd say still it's all about education right and getting teams because from there then I really talk about you need to have a full edge strategy right don't go deploy a pilot out in the field with a gateway to do this thing, right? Cuz again, where you going to go from here, right? You you don't have all the components. So I will say now finally clients are starting to get that and understand that and so is the rest of the market, right? So, as you've mentioned, right through CO, a lot of the smaller companies kind of faltered off to the wayside. But those that are left now the bigger companies are going, hey" we, can't can't build this fast enough internally; let's go partner, right? Let's go white label, let's go put these people under the hood."Right? And that's where you see the Zedatas of the world coming from, right? I mean, they've got a solid solution, they've got a solid kind of ecosystem that they work with with other smaller perspectives, and they got the runway to, you know, to start doing some bigger stuff now that they've partnered up. Um, and then you kind of see that I'd say the other side of the equation I'd say is the from the cloud providers themselves, right? I mean, right, you know, Azure Stack, it's pretty solid. Azure Stack Edge, right? And they've got some new stuff coming out in November that makes them pretty sticky, right? And and that's an easier conversation too to a lot of the IT side of the clients, right, of hey, it's just the same look and feel. Um, you know, AWS, they have their stuff, you know, got a lot of cool stuff as well. Um, you know, Google just come out with a newer version of distributed edge, you know, that kind of addresses some stuff. They've done some alliances, um, you know, in the manufacturing space around with litmus automation and stuff. Google will never have a place in energy. They've already shot themselves in the foot with that. I think it's sad, but it is what it is, right? Uh, you know, this uh even even clients today will say that, hey, we were doing stuff with Google, we liked it, but no. Yeah. I mean, they destroyed any any confidence that anybody in that space could have had in their uh in their ability to uh build something out and then not Yeah. It's just a I mean, as you and I both know, the energy industry is very much so driven by risk. And if you're obvious about the risk, it's going to be a lot harder for someone to be like,"Yeah, go ahead and put all of our stuff on GCP and hope that Google doesn't get mad about it and shut it all off and then we're effed."You know, like that's uh No, I agree. I use I mean, I use GCP for for some of our database stuff internally even just because we're we're a Google shop and it integrates really nicely with a lot of stuff. We can automatically back up, you know, our Google Analytics and our ad accounts and all our Google services directly to BigQuery. Yeah. Natively, which is sweet, right? But if I was, you know, if I was building something that ran on it, which we are, uh, it's not on Google, I guess. So, it's a it's a weird it's a weird dynamic for sure. It is a weird dynamic. So, so anyway, so I'd say a lot of that still in education mode, right? We've seen a lot of pilots, a lot of, you know, PC type stuff, a lot of stuff that it's been called edge. It's truly not. Um, Where do you see people kind of either tripping up or making kind of either bad strategic decisions or bad, uh, moves kind of in that edge space, whether it be, you know, security or, like, one of the big things that I never thought about even when I was there, um, until it started kind of popping up, but was, like, you, the CI/CD part of Edge, right? So, the whole concept for, if you don't know, I've got a great, personally, uh, I personally think it's a great pitch, uh, from one of the Energy Tech nights you can go find on on the DW channel, but, uh, talking about edge, but the whole concept is: everything doesn't need to go to the cloud. The cloud has become I use the dumpster fire uh gif in my presentation for that because that's basically what everyone has done. Hey, just put it in the cloud and then we'll figure it out. I was like, well, you put a bunch of shitty unstructured, you know, data in the cloud, that just means you have shitty unstructured like it doesn't make it any better. Yes, you have it stored somewhere, but that doesn't m make it actionable or easier to use. Whole concept of edge is being able to move the compute closer to the source of the data, do more of the processing at the site of the data. A lot of people have strug like trouble kind of comprehending what that is. that uh one of our potential customers at at Hivesell gave us this use case and it was Walmart and you know Walmart has dozens of security cameras at every single Walmart and they've got thousands of Walmarts, right? And so it's like they have to record all this video because of liability, security, etc. But it's like but what if they don't like when do you cut off storing that video? Like why do you need to store all of the video? Like you've got, I can't imagine how many hours a day a Walmart is damn near empty, that it's like, why are we even storing this, right? And so the whole concept would be, hey, we can put these edge servers in the in each store. They can all run the same software; they can be managed from one location from our IT department somewhere else; but they can sit there and process and do computer vision and say, hey, if there's video without humans in it or people in it, don't save it because we don't need it, right? Like, just the storage costs of that become incredible because it all compounds, right? Like that's I think that's the big thing with edge is, you know, you're talking about hundreds or thousands of locations whenever you're talking about what a good edge use case kind of think looks like really in most situations. There's still others where even a handful of locations edge can be very useful. But um you know just being able to store data locally process it and say hey I only need out of the thousand hours at a of video that we recorded this week I only need five of it because only five of it has people in it or whatever it may be. Um, but you know, where do you kind of, where do you see people kind of unsure about things or, you know, maybe even making bad decisions today, because that sets them up for, like, if this is coming which I think you and I both are in the camp that it is then that's going to screw them essentially in the future, right? Whether it be again security or even just like the frameworks, right? Like that was a big struggle for us running virtual machines versus containers versus Kubernetes. No one is experienced with Kubernetes on right most energy side of things yet. So that's still a whole new conversation. But what are Yes, it's know that's a huge question. It's it it very much is right. So what I have seen is very single point specific solutions being rolled out because it was to solve an immediate need, right? or or a thought of need, right, of of issue. Um, and a lot of those times are happening because of individual business units or have a problem. Somebody locally comes to them with a solution. Hey, we can put it in and all this, fix their need, and then they come into their midyear reviews or whatever with their other colleagues and they're like,"Hey, we fixed it this way."And somebody in South Dakota says,"Oh, yeah, that's cool and all, but doesn't work for this."Right? So again, it's not scalable in any instance. Um, and then to your point too as well, there's frameworks, right? So nobody had that conversation upfront of what all is it involved when you say an edge, right? An edge solution, right, from an operating system across, right? Um, and the and then the last thing I would say is nobody ever talks about how do you stack use cases, right, with the same infrastructure, right? Right. It's always about, hey, here's 10 use cases, but each one of them require a different component, right? And so it's like, okay, well, I'll start with this one. Oh, that one worked great. Now, let's stack this one. Oh, well, not so much. we should have, you know, so that stacking instance is is new and definitely then you get people's eyes that light up when you say that, right, of saying, hey, if you just had cameras out there, you could see one, do they're wearing their PPE, right? Do they have the things that they need? Two, did they fall down? Are they on a ladder? You know, all that stuff, pose estimation type of stuff. Uh hazard location areas, blockings, all those type of things, right? So that again has kind of been a missed opportunity I think in in in moving edge forward. Um and then when you talk about the security aspect of it as well is something that I I'm getting more and more into conversations about right so you know a lot of OT security is going on it kind of fits in that framework but some of when you talk about CI/CD stuff right doesn't really play nice with the security framework. um you know and and and and large companies have been trying to figure it out for a long time. What does you know a a a trusted IoT gateway? How does that fit into the architecture? How does you know XYZ, you know, that's where all this gets I mean that's what people I think don't appreciate when they say edge, right? Is it's it's not just software. It's the underlying framework, the architectures, the hardware that it's running on, and then on top of that, the networking and security aspects of all of that. And then you've got, oh yeah, we need a camera that pulls 4K over here, and we need this sensor from that company, and this one's Modbus, and this one's TCP IP. And like, it just becomes a rat's nest very, very quickly. And I think that's uh, you know, because that's it's one of the weird things with the edge propo proposition at least, right? Is it's like, hey, you know, it theoretic or it the whole one of the big pitches in my mind benefits for the oil field, which led me to kind of take that job with Hivesell, was:"Hey, this can run in a disconnected environment,"which is great. But then you've got and you've got a lot of use cases for edge which are disconnect plants, except power plants, all things like that, right? Where it's like,"Hey, nothing externally needs to be able to connect to this on purpose, because it's just that's the most secure thing we can do is completely detach it," um, from the network. But then it's like, well, [ __ ] Well, how do you push updates to it, right? So, it's like it's this big dichotomy of like push and pull and you know the the IT networking guys are like yeah you're not going to have access to this thing behind our firewall. I've got to open that port up for you if if you want and you need to ask me to do that because that's their role, right? And it's like, well, no, I just need to push this little software update. It's like how do we do that? You know, yeah, that that's a big a big area of discussion right now is is is what is the best practice for that, right? So, where let's use your video analytics use case, right? You're 100%correct, right? None of those images should go to the cloud unless the algorithm isn't doesn't believe what it's telling you it is is, right? So, whatever that threshold may be for the accuracy of that model, then yes, you send that, right? because it needs to be labeled and reran and retrained. Um so it's okay you it's easy to set up rules based on that and say okay here store locally 7 days with everything but these bucket over here at midnight every night send it off right and then you can set it up on the cloud side that says hey every Friday somebody's going to go in and annotate it rerun the model and and push it back down right and and how does that that happen but but that's the key right is is is one how do you transverse the layers in the network to get that out. Mhm. And then get back in, but then also trust it enough that that okay, it can't just throw a control right right out of that. Um and and and I'm having real world conversations around this because even then just basic Lauren sensors to a Lauran server. Do can you name a tool right now that can actually look at that network? Right. I couldn't I like I it says it's encrypted. What what does it matter, right? Like why do we care? Um but that's the thing, right? And when you start talking about security visibility on top of, you know, the the the plans that you want to do with stuff, right, the updates and all that that comes with it. So I'm definitely seeing that somebody needs to come up with an answer for that. Yeah. Uh and and so I think they'll start seeing some movement. And I've posed that question to a lot of, you know, companies that I work with in in and especially in the startup space of like, hey, can your tool do this, right? If it could, you know, I mean, and and I think that's how we'll see the change, right? I mean, that's, you know, and they should be looking out for that as well, right? I mean, the clear blade story, right? I mean, I mean, how smart was that to you to figure out, hey, you know what, what if I built something that was an easy migration pan for IoT, right? Oh, you know, their business quadrupled overnight. Yeah. Um, so I think that's where we would start seeing some some things come in. Um, I'd say, oddly enough, too, I'm starting to see a lot more power um in Service Now that they've got some real extendable capability in the OT space, right? So, Accenture worked with Service Now to create an OT module. So you have your stuff like uh Draos and the ZOMI that does all that asset discovery. Then we'll push that into service now and then you can set up the you know the what you want to do rules based from there and kind of push that down. Yeah. I'm like well why can't you now extend that to cube clusters right and and the software revisioning and all of these. So we'll see you know I mean it's it's an everchanging world. No, that's the cool part about it being so early, right? Is there's just I mean, it's the same way I felt when I first kind of got through some of the initial hive cell pitches when I was interviewing with them. I was like, damn, there is a lot of opportunity here. And I didn't even know all the problems and [ __ ] we were going to run into in phase at that point. But I was like, man, this is this is slick. Uh but yeah, I mean, there's there's so much, right? Like uh CI/CD for example, right? So like, you know, uh, what happens to your edge device when you're pushing an update and you lose network? You have to account for that scenario because that h it's going to happen, right? It's not if, it's just a win, right? And so, it's like you have to have not only software on the backside to say,"Hey, this didn't work until there is connectivity; validate that that connectivity is strong enough to be able to push this; and then start pushing it again,"but then also on the edge device that says,"Hey, we didn't get the full download."Kind of like on your iPhone, you're right. It says, "Hey, do you want to download it now or later?"Okay, we'll do it while I'm asleep so I don't have to worry about it. Whatever. But then it's like if it doesn't finish and your code isn't written correctly and it started to update, well [ __ ] you're you're you've just blocked you completely locked up your server and it is a nightmare that I still live today, right? So, as we were rolling out our devices, right, we had, you know, a fleet management screen, right, where you could see last poll came in, everything. And we someone pushed an update to the model and you could sit there and watch them just go red like like exactly like cancel cancel.
And we had scripted in there basically like if it didn't if it didn't have hash back to the MQTT broker, right, it would power down the modem and basically reset the modem. And you know, eight out of 10 times that that corrects things, right? A little power cycle. Mhm. So I'd say 10%of them came back that way and the rest did. All of these units were not in Texas. Yeah. So that's the big thing that, you know, people don't even think about is; it's like, okay, the benefit of edge is I don't have to have tech or, you know, a server manager IT guy at every single location because it's distributed and I can manage it centrally. But then, if I [ __ ] it up and I brick my my server, I have to have someone go out there. So you have to build that pipeline so it's robust enough and that but that's again that's where all these opportunities are in my opinion right like yeah It's uh it's, it's kind of a wild west world, but I think there's a lot of, you know, it's not new the, like the axe, right? Like, we've done all of this in the cloud, we've done it with on-prem before; it's just now we're taking pieces and learnings from those and figuring out,"Okay, well, here's all these nuances that we don't have to account for in my, you know, AWS cluster because it's got its own dedicated lines and backup power generators and all this other [ __ ]right? Yeah. I mean, exactly. I mean, we PNID loops have been around forever, right? We've done it in some form or fashion. I've we've remote shut wells, we've changed choke bridge, all this, but you're correct, right? It's now moving away. It's it's just again giving it much more capability and and and really dialing stuff in is when do you when do you think uh how far away are we from seeing because I know going back to your the kind of SCADA automation side. I know we're starting to see
call it pseudo automation. Some things basic things can be automated. Again we've talked about risk before. So the risk thing is always one of the biggest hurdles you're going to run into with automating field equipment and the oil field. But where do you see, kind of, how far out are we from having some of these things, these lower hanging fruit things not a fully automated drilling rig, but you know, some of the opening and closing valves, chokes, you know, distributing? I mean, because there are companies that are doing that, right? Like,"Hey, here's the your automated chemical distribution for your um production on all of your wells,"and then you can look at them and change this one, and,"Hey, it's going to release at 0.1 gallons per minute for five minutes until it hits this volume and then it's going to shut off,"right? Where do you see that kind of how do you see that evolution today and and kind of moving forward? So, yeah, you're right there there are there are team there are fields that have that today, right? That you can change choke flow rates, all that type of stuff. You still have a guy that's going out there, right? I mean, the end of the day, still got to have a measurement guy calibrate and change orifice plates and stuff like that, right? It'd be nice that that would be very slick, you know, maybe that's the next business idea, right? Is create an an AI orifice plate, right? Where it just can, you know, die for whatever. Um, so anyway, you will start seeing more and more of it, right? So, I'd say actively, not personally, but a I know a team that's engaged with an operator to figure it out for the perian, right, of what that looks like and how to get there. um you know for a lot of their assets. So I'd say that's you know probably a year in the think tank banking right and then you know probably go from there. Um I will say a very large uh UAE company is on the path. They've already done one of their fields is fully autonomous. Uh that's awesome and u they're going to continue to move. So so it's coming right. So I think yeah it definitely will happen. I think the next area that that is kind of an interesting and and and we'll see it play out more in the utilities before we see it in oil and gas, right? Because usually they're the first to adopt, but truly virtualized PLC's. Yeah. Right. They're coming. Um they're coming. I know my diehard in guys out there are going to be like, "Nope, never."You know, it has to be this way. Um but yeah, I mean it's coming and you know it's coming it'll get there. Um I interested to see how many sensor companies really start going 5G enabled. Yeah. And getting the price down. I think that's one that that'll be to watch. Mhm. Um and or um truly just coming with MQTT out of the bat, right, without being proprietary. So that's a perfect segue because I wanted I think this is one of the first times we've talked about MQTT on the show. So I wanted to unpack that a little bit. Tell people about what MQTT is and then kind of go into why MQTT came around, what the benefits of it are for especially around the kind of SCADA IoT side of things. Yeah. So MQTT right is just a message bus right telemetry. It's not new, right? It was back in the late 60s7s. I didn't realize it was that old. It was came from IBM. Uh Arlland Nipper is one of the the forefathers right from Sirrus Logic that actually works hand inand with inductive automation. And I guess that makes sense though, right? A good 40year adoption cycle for oil. Yeah. I think if if I'm not mistaken, the first IoT MQTT message was a toaster. I want to say it might have been a refrigerator. I might have that wrong. But but anyway, home appliances. Yeah, the home appliance, right, that went through there. So, um, but yeah, the whole point of it is is again going back to where we were talking around, right? Today, everything's pulling response, right? And and it's it's you got to configure it and it's again, a data is coming whether you need it or not. Yeah. For those of you who have never had to deal with that, um, Bobby and I both dealt with that at our the company we both worked with or both worked together at RDS. But the problem with that is if coms go down or uh even if there's just some lag, right, the server that's receiving the data will have nulls or blank spaces or even the the time series coming in out of order. Uh all of these things end up screwing up your database very easily and very quickly. Um especially if you're not watching it, that can happen. And then it just keeps going, right? And it it never it's not intelligent essentially, right? It's just kind of a dumb, hey, here's the data. It's available. Push it over here and then it's just going to go over there, right? And and it's it's impossible to troubleshoot some things when you have a bad actor like that, right? Because well, and the data does not stop coming in either. That's the other part. It's like, okay, well, I've identified where it is, but now I've got to go figure out how to rewrite it to correct it while I'm still getting new lines of data every however often. Right. Yeah. Yeah. Exactly. Um, and the other the other piece of that is is is the way that the data moves, right? So it's from the sensor pointer, PLCRT, whatever up to the SCADA system over to the historian. Yep. And it goes there to die, right? And so, if you're a reservoir engineer, or you're facilities engineer, or just whoever right, lost, you know, whatever you're trying to do and you have an application that wants to utilize that data, it...I have seen personally, at a large producer, it take almost two years to get the right approvals and the right person to unlock the data out of the historian. just to transverse into the other side. Right. And so so yeah, so it's miserable. So you know about I don't know four years ago maybe it really started popping on the scene. Um inductive automation was kind of the really the first ones to push it was this idea of having a broker right of how you you have everything published data to it. You you you have it in a JSON format. You have it rules-based, right? It doesn't have to be time based. And then anybody that wants to have that data can subscribe to it, right? So your other if you want to test out some new machine learning algorithm company XYZ, great. Open them up a pipeline, let them do what they need to do. You don't like it anymore, shut it off, right? It's it's it's really tight and easy integrations. Um and and and having it in a in a kind of stairstep manner as well even makes it even more better, right? to where you have it from your core, whether it be cloud or or on prim. You have your master spark plug B broker, then you can have smaller brokers doing what the things it needs to do. And again, serving right multiple applications um on that and it does bring a much more secure environment as well, but you have it set up in the in the right way. Yeah. No, that's uh I'm excited to see how how that kind of grows and helps make the oil field data just one more robust, but two much more accessible because that's what we've talked about plenty of times, right? Like yeah, IT departments in most operators, I think we're even having a a topic on this uh at Fuse in a couple weeks. Go buy your Fuse tickets if you haven't. Um, but it's like it should be almost embedded in the teams, not this overarching power center of, hey, I'm an engineer and I need this data, but I can't get it because it has restricted it or whatever. And it's like that is just killing all of the internal innovation that you could possibly have because the other part of that is they're still engineers. So, they're going to find a way to get the [ __ ] data like I'll work around it. I'll dump it to a CSV. Like, whatever. We're going to use the data how we're going to use it. uh just make it easier for us and we won't have to call you all the time. The number one perpetrated lie between it and OT is OT thinks it doesn't really work 247 and it thinks that OT just has no clue about anything. Right. Right. And and and that and that's not the case. Right. I mean not at all. If they would just work together easily and understand it's for the betterment of the daytoday, a lot of cool stuff could get done. Yeah. Absolutely. That's uh Yeah. So, we've got a few more minutes. I did want to uh I wanted to ask just quickly, what do you what kind of cool stuff do you see coming on the kind of SCADA IoT side of things? Like you mentioned Laura Wan, which is not new by any means, but I adoption of it I would say is newer, at least in the oil field. Um for people who don't know what Laura is, would you give them a a quick blur on Yeah, it's just it's just lower lower in the frequency, right? It's just, you know, meant for short distances, small burst packets, right? Um it's just a different, you know, think about it from like Bluetooth. Very similar, right? It's just a different different bands in the frequency. So, good for like on pad location or on pad data distribution or can we It can go a little bit further. I'd say anything under seven miles is where you want to do it. So you could, you know, you could do, you know, a tiein, right? Pressures on the tiein on the pipeline back to a CPF kind of in there. Um, a lot of times the the kind of the preferred methodology is is wherever your network tower is, right? So not like your your AT Verizon towers, but your your actually network field telecom network tower. you set your your server gateway uh in there and they kind of let multiple plaids have it versus having a gateway in every RTU box that's at every pad. No, I think that that was one of the the things I realized when I was at H Highell talking to this a lot of skater guys like yourself on the on the edge stuff is. It's like, you know, the company thought,"Oh well, we'll have, you know a, hive cell on every pad or on every well,"and it's like, no, that's not how this right. this will be at those kind of like uh near edge locations which I the cell phone tower is a perfect like analogy for normal people right your phone is the edge device. your cell phone tower is this me middle ground which a lot of people don't realize have they basically all have data centers at each cell phone tower now in some capacity and then that gets pushed up to the cloud right it's going to be the same situation you'll have all these IoT devices on each location on all the pads those get pushed to this kind of regional level field level Hubs and then that does all the compute processing, computer vision, ML models, etc, and then pushes what else is needed up to the cloud. Um, I think it's uh, I think that model makes a whole lot of sense for because they already that's the other big thing is they already have infrastructure there, right? Like, most people in my experience with edge want to put it it's just this, you can't break them of how they normally, how things have always been done, so it's like,"Okay, well, we'll just put it in this ska cabinet with the other servers, right?"Like, or in the, you know, the server room with all the other servers on them, right? right um I think it's going to be I think it's going to open up a lot of really interesting stuff. Make things so much safer. Um, and also make the field guys lives just that much you're more empowered because it's like, hey, I know what's going on. I know the tools I need. I know where it is and I'm going somewhere with a purpose instead of just driving my same route every two weeks or whatever. And yeah, you're 100%. It's all about
the the level of being comfortable, right? I mean, and I and I think that it will start to get much faster as you know, the field guys are younger, right? They they're used to they they're they're cloud native. They're used to having their things. Um, so, you know, like again, like if you were an operator and I said,"Hey, let's quit wiring up all these sensors. Let's go to wireless."Right? You're going to be like,"Okay, cool. Do it." Right? Versus when I was telling them,"Hey, we were going to go wireless."They're like,"Are you stupid?"Yeah. Like, no. Right. Um, and and and and I did take one bad look, you know, on that. I mean, I'll never forget sitting at at this manufacturer and telling them they were lying about their battery lives. But, you know, that was the only thing that was bad, right? If something would happen and it would continue to, hey, I'm 15 pounds, I'm 15 pounds, I'm 15 pounds. You know, you're going to drain that battery. And so, when you have to go replace batteries, then y kind of bad luck. But, but I think all of that is changing. So I think that's definitely kind of where it's moving to is no longer will it be thought of okay yes I have seven wells on this pad so I need to have seven right different meters and all of this stuff right I think all of that will start getting condensed which I mean for some part you do for measurement you have to right because of the lease agreements and all that so that's got to pay those royalties and those taxes got to pay the royalties so I think that's what it'll start coming up and you'll start looking at how you know how my whole field interacts and is there information that I can share to another field. Um, and I'd say then other companies, you know, starting to share that data as well, right? So, you can really get into, you know, what we were talking about with the mud logging, right? Where you can see, hey, I did this type of work over on this type of well and got this type of results. Now, I can replicate it. Oh, I'm seeing these early signs from the the pressure trends that I'm seeing, so I need to do this. Right. I think that's really where the the f the next generation is coming. Yeah. No. Yeah. I mean like just the little stuff, right? And I think that's one of the few areas that we have actually seen edge deployed in the oil field specifically is like preventative maintenance type stuff, right? Hey, I think early on a lot of that was super hyped as AI when it was just conditional statistical analysis. It's obviously getting better, but um, you know, I think the the pie in the sky thing with with edge is, you know, the uh distributed machine learning models, right? And so, uh, federated learning, you know if, and I...I don't...I don't think there was a really good way to describe that to the average person until chat. GBT to, me is like the perfect example of that. People don't think about their using it on their computer as like the edge instance, but every person that uses it is a unique instance, right? And so you're having but just look at what's happened in a year on that model with you know you it's the fastest growing app. You get millions of people using it now pretty much on a daily weekly basis. And so look at how much better it can get in such a short amount of time when you do have that ability to feed one giant model from all these subm models and all their unique instances and nuances and then just create this giant loop. New models, new learnings, feedback gets up to the mother model. It gets retrained and then pushes that back down. And that I mean I have no idea how far away we are from that in the oil field. But that gets me super excited about like the potential because there's just so much data and there's so little time and there's so few people still in the industry that it's like that's going to be a really really cool thing whenever it does happen. Yeah. Secondary production is the number one use case for edge, right? And and and that version of AI, right? Because you're 100%correct. is there are plunger and pumping left companies that are out there that do have solutions and they they fit a need. Um but yeah, that's not true AI, right? And and once it gets to that point where you can have that, you know, millisecond reflex loop. Yep. Can you imagine how much production we would get out of older wells? It would be huge. Yeah. It's going to be it's going to be crazy. [ __ ] man. We're almost we're we're up to an hour now. Um so at the end we normally do a uh little speed round. Um what were we watching? I was watching uh I was watching watch what happens live with my wife uh last night. It's on We watch a lot of Bravo shows at my house. Um but I'm going to I'm going to kind of piggyback off of something he does. So instead of asking these open-ended questions, I'm going to ask you just some multiple choice questions. So just give me your uh Mac, PC or Linux.
That's so difficult, right? So it depends on what I'm doing. I'm on my tablet when I'm cruising and just doing like that. Apple. If I'm on doing work stuff, it's got to be a PC. Yep. Has to be a PC. Anything that I'm going to do for home automation or anything like that, it's Linux. Yeah. So all three I kind of uh I kind of feel like that's probably going to be a pretty common answer here. Uh because I I think a lot of my coding frustrations come from the fact that I'm on a [ __ ] PC and not a Mac or Linux uh out of the gate. Uh Xbox, Nintendo, or PlayStation. So, I have the PS5. That thing is a bulking beast, but I like it better than Xbox.
Firefox, Chrome, Edge, or Brave? Brave all day, every day. Although I'm upset with Brave because now they've had an update. So the crypto Mhm. wallet extensions broke and now you have to go one of two. They've like reduced the the options, which I'm like, you know, I don't care anyway, right? It was, you know, talking right, right? But when the Brave browser comes up now, you have this like 404 error box, right? That's just glaring you in the face. Fix it, Brave. We like I I need you to fix it. Yeah, we like your [ __ ] Yeah, just fix it. Uh, okay. Here's here's a fun one. NBA Jam, Call of Duty, Doom, Warcraft, or Oregon Trail?
None.
Not a gamer. None. No. No. I'm not a gamer. Just that. Uh, no. PGA. Okay. Golf. Yeah. All day. The Tiger Woods one. The OG where you could spin the [ __ ] out of ball. I played so many hours of that game. Yes. Yes. I I um I Yeah, I tried playing Fortnite with the kids, but that's just so hard. It's just too much. It's just too much. I'm like, wait, what? There's that's a whole another like level of skill I feel like that generation has that was really into Fortnite because the shooter alone in Fortnite is hard enough. But then you've got these kids that are just building while they're shooting and I'm like I can't even my brain can't work that fast. No, it's the same way with Roblox. Yeah. Oh, dude. Rob, what a great like my six-year-old uses it. You know, we've got the restrictions and stuff set up, but still it's like I people don't people sleep on they're like,"Oh yeah, it's a video game."It's like, no, it's literally every single video game that you could ever want in one spot, right? Like, it's such a great platform. GCP, AWS, or Azure?
Azure first, GCP second, AWS third. I like that. GPT, Claude, Bard, or something else? Jury is still out. The juryy's still out. I'm right there with you. Uh, Postgress, SQL, or Snowflake? stuff like another good answer. So, Bobby and I'm gonna I haven't even told Bobby that I'm switching the uh the style of questions, but we're gonna we're gonna keep doing that because this is I think this is a lot more fun to uh Yeah, it is. So, I I kind of, you know, as mentioned, you know, I keep up with your podcast and I watched a couple more just to go, okay, what do I expect, right? It's something that's instilled from an Accenture, right? You have to prep six weeks before you before you do the thing, right? Whatever. So I had to was thinking I was like all right I need to know a book and I need to know Yeah, the book one will be tricky. Well and, that's it's even a tricky question to ask, right? Because some people are like,"For the longest time I didn't read books, I just listen to podcasts or audio books and all that stuff."But it's, uh, man, it's been fun. I can't believe the hour's already gone. Yeah, I know for sure the on books. So the last thing next five moves Patrick David bet. You follow him? Yeah. Great story. Great story. Nice. I'll I'll add that to my uh my list. But no, I do. I appreciate I appreciate it. Yeah. How do people How can people find you? LinkedIn. LinkedIn. I uh I I have a Twitter Twitter Twitter or X I was going to say it's not even Twitter anymore. Yeah, an ex troll profile, but I don't I don't I don't really Yeah, that's a bad one. Uh, and then Instagrams, you see a bunch of kids stuff, so Oh, for sure. Yeah. So, Quinton Jones on Twitter. It's Quinton with a Q. No, Quinton Jones on LinkedIn. You'll never find a Quinton Jones on X. If you do, it's not me. We have butchered uh that part. Made it more confusing. Quinton Jones on on LinkedIn. Yes. Not on the non-existent Twitter platform. Um, appreciate it, man. Thanks for coming. Yeah, see