RTI Field Applications Engineer Patrick Keliher discusses the technical demands that IIoT developers face when building distributed systems.

In Episode 50 of The Connext Podcast:

  • [0:17] What’s a typical day in the life of an FAE like?
  • [5:00] What are some of the hottest IIoT applications in your region right now?
  • [7:52] What do IIoT developers typically ask you for in regards to their distributed system?
  • [10:52] How is Connext DDS being applied to the Oil & Gas industry?
  • [15:53] What is your favorite IIoT use case?

Related Content: 

  • [Blog] Connext DDS and the Industrial IoT: The Top 5 Things to Know
  • [Blog] Digital Transformation Revisited: The Promise of Things to Come in the 2020s?
  • [Blog] What is IIoT? The Industrial Internet of Things Primer
  • [Datasheet] RTI in Oil and Gas
  • [Podcast] How the Industrial IoT will Revolutionize the Oil and Gas Industry
  • [Product Page] RTI Routing Service  

Podcast Transcription: 

Steven Onzo:

Hello everybody and welcome to another episode of The Connects Podcast. Today I'm here with Patrick Keliher, we call him PK for short. He's a Field Applications Engineer. He's visiting from Texas and we're glad to have him. Welcome to the podcast.

Patrick Keliher:

Well thanks, Steven. It's good to be here.

[0:17] What’s a typical day in the life of an FAE like? 

Steven Onzo:

So today, the big thing I want to talk about is your role here at RTI, Field Applications Engineer. I would like to understand what a day in the life of an FAE is like, and for the sake of the podcast, maybe not a day, but 20 minutes in the life of an FAE. So I just want to kick things off by asking you, can you tell me what a typical day is like as an FAE? I'm sure all of your days are very different.

Patrick Keliher:

That is a great question, Steven. And yes, they absolutely are. The most important part of all of our FAE’s day revolves around our customers. I believe the FAE’s probably interface with customers more than anybody else in the company. So we start out by checking the support queue and make sure we have no urgent support issues that have shown up overnight or any existing support issues are being handled.

Even though FAEs don't typically solve the support issues, we make sure we coordinate with the support department who does an awesome job taking care of these things and make sure that things are moving along and escalate and push things if we need to. But most of the time it's just monitored, because our support department really takes care of that. So other than of course support, a large chunk of our days involve meetings, whether they are in-person meetings or remote meetings, we do a lot of those.

So we'll typically have two or three phone calls or web meetings or in-person meetings in a day, if we can get them. Those can be anywhere from a very short 15-minute touch-base meeting to a full two-hour, three-hour or longer or Lunch and Learn sort of meeting with new or existing customers, because there's a lot of information revolving around DDS - it can take some time to uncover exactly what the customers need to know.

Then beyond just that, of course we do trade shows where a lot of times we're off doing a trade show for several days in a row, meeting lots of other potential customers, etcetera. Then another big thing we do is we take care of all of our evaluation customers where we have many people starting to use DDS - they want to evaluate it. And so we make sure that while you're doing the evaluation, you get all of that personal one-on-one assistance that you might need. If you've got questions on the evaluation.

Steven Onzo:

That evaluation support, is that face-to-face or do they contact us via support portals. or how does that normally work?

Patrick Keliher:

We can do it face-to-face. That works out ideally. Many times customers are too busy and so they don't know when it's going to be, and so it's much more of an ad hoc. They call us up or we do a quick web meeting. But yeah, definitely the more successful evaluations we spend more time with. And if the customers can dedicate just a few extra minutes for helping us help them, then the evaluations are much more successful.

Steven Onzo:

Then there's also meetings too where maybe they're not available face-to-face and you have to do that remotely.

Patrick Keliher:

Yes - a lot of times our customers are distributed and so you may have customers in four or five different locations, in which case the remote meetings are perfect and that way you can potentially be onsite with two or three folks. But then I think the record meeting I had, there were 25 people dialed in remotely and so you absolutely can share a lot of information that way. It's more challenging, but you can also get a lot of good information distributed in a pretty quick manner.

Steven Onzo:

Right, and all of these put together. I could see how that certainly keeps you busy throughout the day.

Patrick Keliher:

Yep, our days can be pretty long. It is a fun job to have if you want to work from early until late in the day, potentially. Because our customers a lot of times have deadlines - and if you can't help a customer at seven o'clock on a Friday night, then they may have some serious problems. So we are a lot of times helping customers late into the evening or late into the week.

Steven Onzo:

I know you said customers - do you also work with potential customers, existing customers, or is it both?

Patrick Keliher:

For me, everybody I talk to is a customer, whether they're a potential, whether they're a past customer, existing, I have no differentiation between new ones and possible ones and existing ones, whether you've been with us for 10 years or whether you're just brand new to us, you're still my potential customer, my customer. And so yeah, we absolutely set up more time probably for future customers than we have for existing customers because as you know, RTI has got this really great infrastructure with services and support and all the other things that we have to take care of existing customers. So for new potential customers, a lot of that does fall to the Field Application Engineers like myself.

[5:00] What are some of the hottest IIoT applications in your region right now? 

Steven Onzo:

So I know you're located in Texas, but travel all over the South Central U.S. and even make your way down to Southern California every now and then. What are the big IIoT buzzwords in your territory right now? What are people getting excited about and what are some of the hottest applications we're seeing in that region?

Patrick Keliher:

So we are running into a whole lot of autonomous things, whether it's commercial things or Aerospace and Defense, autonomous vehicles. That's a huge growth for us and for customers, it seems. So this goes to stuff like FACE. So FACE stands for Future Airborne Capability Environment, which is a standard that has been developed by all of the services that fly things. We've got a lot of activity around FACE. It has really seemed like it has picked up some momentum, and all of the new programs are calling out FACE as a requirement for army aviation in particular, which we have a lot of in our particular territory.

And along with FACE, a lot of times we have got a certification requirement - so to fly many of these airborne systems in the United States airspace they have to be D0178C-certified, in which case our certification product that we have is a perfect solution as well. So you if you combine FACE and the cert product, you have a really great starting point for customers to get these very safety-critical, life-critical systems done.

Then of course we have a bunch of other really cool customers in our territory, Oil and Gas being one of them. And Texas is a huge business for us, and we've got lots of existing Oil and Gas customers and more showing up all the time. And Energy is also very important for our particular territory. We've got a lot of energy companies that are looking to adopt and adopting DDS as well.

Steven Onzo:

I just want to backtrack just a little bit. Go back to FACE. FACE and DDS isn't anything new but it is picking up traction right now. Why do you think that is?

Patrick Keliher:

I think like any standard, it takes time for the industry to then adopt it. Absolutely with airborne systems, they have very long lives. The B52 has been flying since the '50s, so absolutely these guys do not adopt things quickly - but when they do adopt, they really get on board. And so when FACE got to the 2.1 version of specification, I think that helped a lot of people get there. Now that the 3.X version is coming out, that really seems to be a driver for people.

So we are sort of in that awkward stage of...customers don't know if they need 2.X or 3.X. And the good news is we can help them with that and get them to the right decision so that they can meet the requirements for their end customer, the U.S. government, Army, etcetera.

[7:52] What do IIoT developers typically ask you for in regards to their distributed system? 

Steven Onzo:

That's actually a great segue and I love when that happens because the next question I was going to ask you is about customers and their needs. When you are talking to a customer and they say, "Hey, I need this from you." What is that most of the time, what is it that they need?

Patrick Keliher:

That's a great question Steven. We absolutely find with customers that they will typically ask for what they already know. And so I love the analogy that I always use, which is if your customer is looking to cut some wood, he will ask for a new improved handsaw. All he's ever had in his life has been a handsaw and he wants one that's got a better handle grip or something. And so we'll go in and say, "Hey, what about this really cool 9.6-volt portable saw that you can do 15 times as much work with and still cut your wood?”

So you don't ask them, “do you want a new improved handsaw?” You say, "What do you want to do?" And they want to cut wood. And so when it comes to DDS, we'll just say, "What do you want to do?" And they typically want to share data in their distributed system and so we can help them with that.

And so it absolutely broadens their horizons. And then as the discussions move along, we can uncover more and more use cases they'd never even thought about. DDS can solve a whole host of problems they haven’t even considered yet.

Steven Onzo:

Yeah, you tell them, "Hey, use this handsaw, and by the way, this handsaw also does 20 other different things.”

Patrick Keliher:

Exactly. We've got so many Swiss Army knife tools in our toolkit, and in DDS itself, that it's amazing - all of the value we can help customers uncover, and how to make their lives much easier as they do their next project.

Steven Onzo:

So you ask them what is it you're trying to do? You kind of walk them through that, figure that out. There is one thing though, with a lot of these older distributed systems that have been running for a long time, DDS is best used when it's built as the foundation of a distributed systems architecture. What are the options for these customers with legacy systems that want to use DDS, but it's not possible to retrofit that into their system?

Patrick Keliher:

I think the number of customers who said, "Hey PK, we've got a brand new system we've got no code for,” is basically zero. So absolutely, every customer we go into is brownfield where they have got existing technology - and if we say you have to rip and replace everything in your existing system and put DDS in its place, we would be thrown out the door pretty quickly. And so DDS is really ideally suited for going into these brownfields systems where we can hook up to your existing technology, your existing field buses, whatever you may already have. And using something like our Routing Service or some other technologies, you can now make the system much more modern, much more data-centric and still keep all your existing hardware.

You don't want to replace tens of millions or billions of dollars of hardware on these systems. You just want to make the software much better and make your system a data-centric way of operating.

[10:52] How is Connext DDS being applied to the Oil & Gas industry? 

Steven Onzo:

So, you mentioned Routing Service, maybe they just need Routing Service. Can you talk about maybe a use case where somebody didn't need routing service and how they benefited from that?

Patrick Keliher:

Oh absolutely. Yeah. Routing Service is a huge advantage in oil and gas customers, because when you got an oil rig that's got technology from the 1970s on it, a lot of times this technology has got four or five different communications mechanisms inside of it. Customers have not even thought of the possibility of being able to see all of that data and use all that data in real time without having to do massive changes.

And so by using Routing Service and having adapters to talk to all of these different existing field buses that exist, we can drop that in - just put Routing Service on that new piece of hardware and software they put in the system, and now they have access to all of that data that they've never even seen before. And it's amazing to see guys working on these oil rigs for years whose eyes light up when they can say, "Oh wow, this is what's going on with the system."

Because up until now, all that data is gathered and sent up to the cloud someplace and ignored for all of eternity. Now they can actually operate on that data and improve things dramatically in real time, which is a really cool thing for them to do. And Routing Service can enable that for that and a whole bunch of other technologies. But for us, we see it a lot in Oil and Gas, being in Texas.

Steven Onzo:

Yeah. And I'd like to pick your brain on that just a little bit more, since it's a huge industry in Texas and you have some experience with that. Besides Routing Service,what other changes are the big ones in Oil and Gas that people are seeing?

Patrick Keliher:

Well, what Oil and Gas is going through is a fairly fundamental shift, because a lot of the technology that they have developed into their platforms is now going obsolete. So they know they have to change how they do business and how they are operating with this with a lot of these industrial control devices in their system. And so because of looking at that, they can look at changing everything around. And so one of the big things that a lot of the folks are doing are called rigs of the future, and a rig of the future says instead of having a crew of around 40 people working on an oil rig, we can cut that down to about eight by automation. So we can automate this stuff, use robots and take all of the very manual activities on an oil rig and turn it into a very automated rig of the future thing.

And DDS can do a whole lot of that. So we're very big into robotics and so a lot of even the robotic stuff they're using, they can use DDS for that along with all the data. So it turns into a much safer and much more efficient way of exploring for oil, which is important and still is today.

Steven Onzo:

So I know you just touched a little bit on this, but I think we can get just a little bit deeper. From there, once their technical needs have been met, they can start seeing the benefits of DDS. What are those benefits?

Patrick Keliher:

What it sort of turns into a two-phase benefit they get. Initially they're going to see a whole lot of productivity gains. All of those users that are having to do a lot of writing of code that didn't really add any value to their system will now just be able to use DDS for all of that. So they can concentrate on what their value is and let DDS do all the heavy lifting.

And so they'll get productivity there - they'll get a lot of performance gains, because DDS can make the performance better and make sure their qualities of service are met without having to do any unnatural acts. And then all of the tools that we'll offer as well will help the customers analyze their system. So that's the first initial benefit they get.

Second order is when they start using DDS, now they can actually start looking at what data-centricity can bring to their system. So, when you start looking at a system from a data-centric way instead of a message centric way, all you have to concern yourself with is what does the data do? How does the data use and how can I make my system better use the data.

So I don't have to think about passing here and there. I just say, "I need this data at this sort of rate and I can do this sort of algorithm or control with it.” And so they can now literally change how they architect their system by going to this data-centric mindset, which is a huge advantage and usually we see that in a sort of a second wave of productivity from customers.

First wave, they start using the tools - the system’s easier to do the second wave - now they actually get far more performance out of the system and get even unthought of benefits than they would consider when they started this process. Some customers get it really quickly and some customers take a little bit of time, but there's almost always that light-bulb moment where it just flips on and they say, "Oh, well all I have to do is say, “I need this data at this data rate,” and it shows up. I don't have to know where it comes from or how it's generated or anything.

I just get the data I need. And then you can see the light bulbs go off and so it's pretty cool when you get to that stage with customers and some get it really quickly and some it takes a little longer but they all get there eventually, which is nice.

[15:53] What is your favorite IIoT use case? 

Steven Onzo:

Right. When you say it like that, it sounds a lot more simple. So I'd like to end the conversation by asking you if you have a favorite use case that you've encountered, maybe one that potential customers alike could learn something from.

Patrick Keliher:

That's a great thing. I've always thought about that when people ask about new, exciting customers. And one of my favorite ones we came up with was actually one of the oldest industries in the world, which is the Rail industry. The Rail industry has been around for hundreds of years - what can you use DDS for in a Rail industry?

Patrick Keliher:

Well, this customer came to us and they had a problem where when a locomotive would pull into an area with connectivity, the locomotive was supposed to have a parameter file downloaded which would be very large, explaining what the route was. I don't know enough about locomotives and what it was. I just know it was a large file that was sitting in a database and the locomotives would have a requirement to get this file downloaded and verified, and make sure it was working in less than 15 minutes - which seems like a long time, until you realize that you may have 7,000 locomotives pulling into a station all across the country or the world at the exact same instance.

So there could be a problem with resources and communications. So it was typically taking them sometimes as long as 45 minutes, which meant that those locomotives couldn't leave the station and they'd be 30 minutes late, which obviously no one wants. And so we looked at starting with DDS for this, replacing that particular part of the system that they deployed, and got an initial field trial going in a remarkably short amount of time.

I don't want to say how short it is, but it was astounding to me. And they deployed this thing in the initial field trial to I think 15,000 locomotives or some crazy amount, and it really reduced the amount of time it took to start as the locomotive guide into the station. You could then download these parameter files much quicker than their requirement was. And that was a challenging thing as well, because they were dealing with their end customers, the rail operators, I guess they'd be considered, where they would have all sorts of different databases and formats of these files.

And so they used our database integration service, and they could then attach our DDS in these databases without a whole lot of changes and adapt to all these different operators of these locomotives. So it was a hugely successful project for them and for us. And after less than a year of dealing with them as a customer, we went back and saw them and they had deployed this on three or four different rail systems, which was pretty impressive to us and in less than a year of total usage of DDS.

Steven Onzo:

Wow! Since then, have we had other locomotive companies evaluate DDS who we could share that story with?

Patrick Keliher:

We have got a bunch of activities in that space. Locomotives are once again another non-fast-moving, not fast-changing industry, but yeah, we're absolutely talking to a bunch of those folks in similar sorts of applications. So you can see, for this particular use case it's not going to be as locomotive-specific - that could be any number of train or transportation-specific ways of doing things. So yeah, it's an absolutely interesting use case, and this particular customer is now going to start broadening its use of DDS in a whole bunch of different other locations in their back office, making it much more of a data-centric system they have today.

Steven Onzo:

Well, PK, I want to thank you again for coming onto the podcast and shedding some light on what it's like to be an FAE, and hopefully we can get you back on the podcast soon.

Patrick Keliher:

I'd love to be back. Thanks a bunch Steven.

Steven Onzo:

Alright thanks, PK.

INSIDE THE PLAYBOOK

Get the latest updates and insights from the RTI newsletter

Subscribe to the Newsletter