Gerardo Pardo-Castellote with RTI

SUMMARY KEYWORDS

data, dds, build, talk, people, latency, system, conversation, industrial, application, rti, software, listener, security, omg, securing, normalize, protocol, real, interface

00:04

Welcome to the Industrial Talk Podcast with Scott MacKenzie. Scott is a passionate industry professional dedicated to transferring cutting edge industry focused innovations and trends while highlighting the men and women who keep the world moving. So put on your hard hat, grab your work boots, and let's go

00:21

right once again, thank you very much for joining Industrial Talk. Thank you for your continued support of an ecosystem that is ever expanding and celebrating industry professionals all around the world. You are bold, you are brave and Daring Greatly you innovate, you collaborate you solving problems each and every day. And we truly hear on Industrial Talk. Appreciate that. And that's why we celebrate you, you are the heroes in this story. We are broadcasting on side once again, OMG q1 meeting, and it is a collection of problem solvers. You need to be involved, you need to consider getting involved go out to omg.org O M g.org. And you get to find people that are like and I'm gonna butcher his name. I've got gr G. Gr. Doe. You see, he's got the he's got the dunk and everything. And it's all good. It's all good. So let's get cracking. Good meeting, seeing, seeing how you've been a part of, of the OMG family for 20 years, since the beginning. 20. Did you have hair back then?

01:23

Yeah, I did. I still have some. Yeah, like

01:26

I have. Like, after a while, it's not even worth trying to protect. Yeah, so. So you've been coming to these quarterly meetings. I'm very intrigued. I'm intrigued by DDS and how specially how all of the conversations that are really happening today today that are around digital are bound the data around that whole data conversation, I need data, I want clean data I want, I want to be able to do it, I don't want to cause it's just non stop, because we're trying to be more connected. But before we get into that kind of conversation, give us a little background on who you are.

02:09

I work for a company called Real-Time Innovations. We are a spinoff from Stanford robotics lab. And we got started trying to build robotic systems like 25 years ago. And then when we were done with our, you know, different degrees, PhDs, and so forth, we kind of go looking for what to do. And we started a company trying to like, do the same things in the real world. And, you know, things kept going. But one of the things we did was build the software infrastructure to make the people lives easier when they were trying to develop these kinds of systems.

02:41

So I have to pull on the thread a little bit Real-Time Innovations. What is that

02:46

is the name of the company. I know that but what what do you

02:48

do what?

02:49

Oh, what I do, I'm the Chief Technology Officer there. And we build software, platform software, infrastructure technology, we build like a system based on DDS. That is a software kind of platform that other users would acquire and use to build their systems upon is like an operating system for distributed systems. That's the way to think about it.

03:13

Is it is? Oh, my goodness, you want it to be technical? I do? No, I can handle it. Sort of maybe a little bit. So for for the listener out there. It's DDS has been around for a number of years. And in fact, I think it's the 20th anniversary. Just explain to the listener what it is.

03:35

So like I said, operating systems are good analogy, right? People want to build application, but they want to deal be dealing with the low level stuff of like how to manage memory, how to manage access to files this concern, when you build a distributed system, you have very similar needs. You're trying to build your application, you're trying to create data, process data, and so on. But you have to get the data from here to there. How do you get the data from here to there is complex, right? So you want support in your software that makes the data move easy. That allows you to build an application that can communicate with her application without worrying whether those applications are, are they running on the same machine on a different machine, same operating system, or they're over a network, you don't care about that you're just focusing on your own logic and what information you want to exchange

04:26

does the CD DD DDS solution platform whatever. Also taken consideration, applications talking to other applications, that communication level or and of course, me, let's say I'm Scott MacKenzie manufacturing and I'm I just need specific information and, and I don't want to see how the sausage is made. I just want the data, right.

04:53

It's another layer right? Of course, there's a lot of layers of software. So you know, somebody's going to build a layer that deals with how to put How to send data but, you know, at the level of like 3g networks or Ethernet or something like that, and somebody then build another layer on top of that saying, Well, how is that going to be routed like the internet? Right? Then somebody else builds another layer that says, Well, you know, you don't really want to be sending bytes you want to be sending data? So, um, what kinds of requirements you have quality of service you have at the data level? How do you secure that, right, each one of those layers kind of moves the abstraction higher to what you care about, which is like, I'm producing some position information about my vehicle, I want to know, what's around, you know, is that a car? Is that another person? How far is it right, you're trying to access information from other systems. But to get information, you have to go through a lot of layers of software and hardware. So we're abstracting all that on top of other abstractions that, you know, the operating systems and the communication transports provide,

05:50

to help sort of somebody understand what you're talking about. Because these are, these are conversations that, from my perspective, require some sort of box and an arrow. And, you know, there's there's a visual representation, the only way that I know how you can try to truly visual visualize what you're talking about, do you have a specific use case that can help you help the listener understand with what what they're doing? So,

06:20

07:44

is there a potential for latency? Of course, how do you how do you address latency and in all of this, yet, very late, when you start talking about autonomous vehicles that you're talking about, I can't afford latency.

07:58

I mean, it building it on top of infrastructure that is trying to minimize latency there is like networking technology, like time sensitive networks, this operating system like real time operating system, right, which take very careful care of how long things take. And therefore you can estimate and bound that latency and know that you will be able to make a decision at a given budget of time. And then, you know, basically, we don't drive any faster than you can you can afford to drive just like a human wouldn't. Right.

08:27

A lot of the conversations also, I mean, evolve or revolve around security, take us through take us through that that focus in security on the in the DDS.

08:41

It's interesting, because you said earlier, you know, it's all about data today, and why is it about data is because data is very tied to the physical system, right? Like people or cars, you know, they have a position, they have an SPF, you know, a human, you know, in a hospital world has a heartbeat rate and so on. We can agree to that, you know, this is timeless, like things don't change, cancel the physics on the biology and so forth. How we use the data is very different, right? So there's all kinds of world pressures today, that is way less you to describe this kind of functionality. So, insecurity is also the same kind of thinking like, in the DDS approach, we focus on security data data. So if you're producing data, each type of data flow has its own security and gets protected by the by our infrastructure. The alternative to that is sort of securing the communications channel, you're not protecting the data inside, but while you're making sure that nobody sees what you're doing, so that's kind of the more classic, you know, HDP level security, where you're kind of encrypting a connection and then sending data on the clear on that connection. The problem with that is that everybody that that is in that connection, is to be able to like see the data inside, right? What is when you're securing the data, you can then send that data over any channel, and then it can go and flow through many hops and In a destination that has the necessary credentials to actually see the data. So that's what we call data centric security. And that's kind of the approach that we're taking DDS securing the data. So you can specify who has access to what, and then the data gets protected at the source, and can only be seen by the ultimate recipients of the data that have the right credentials. And the right case.

10:21

I can see when when it's an FIR, forgive me, for the term that I use, I use Greenfield Brownfield, I use old industry terms. If I'm an existing business, it's, I find it probably easy to build something new, and take in consideration what DDS provides, and then make sure that I architect My, my, my systems in a way that makes sense and was like I can, I can have that conversation. But I'm already an existing customer. I have systems, I've been in business for years. And and and I want to, I want to deploy DDS, I want to, I want to be able to start doing that. Because I know that I have, well, here's an interface here, here's an interface there it hits this table and snack table and pause. It's inefficient at best, you got

11:16

to normalize into something right? It's kind of like English. Why is that mostly spoken second language in the world right there. Because everybody said, Well, I can speak this. And I speak that and speak the other thing. But in the end, we all have to agree to something so we can all communicate. So you're going to have systems built on all kinds of technologies, if you want those systems to be integrated. You don't want to have to build bridges from every system to every other system, because that will be like an NS square problem, right? Yeah. But you can say, well, we're all going to like speak our own language, and also speak one other language that we can normalize to. And then if we don't speak the same language, we're going to translate to this common language, right. So VBS is a very good way to do that thing. And that's how people build those systems. We see in for example, a healthcare like people are buying, trying to integrate a hospital and get all these devices to talk to each other. They all have different protocols. So what they do is say, Okay, we're going to put an interface on top of that, that reads your own proprietary, whatever protocol it is, and he's going to normalize into this data centric world. And since all we agree on what this data centric representation is going to be, we can all communicate to each other, I'll translate from my own proprietary protocol into this that goes over it, which means you have a common, you have to have that common. And that's what DBS provides you a language in which you can provide define this common protocol, your common quality of service, and then everybody can go into that. So you can either be an application that is natively already speaking that, or, or translate into that, for the purposes of

12:39

so in a scenario where I have an existing business, and I'm getting ready to deploy a new new application, I just have to make sure that how that application communicates to this common source so that it then then that I'm able to pull everything.

12:58

I mean, if you can think of how the Internet has been successful, something called rest and HTTP became this lingua franca, like there was hundreds of other kinds of different protocols for you know, file transfer for this and the other thing mail and so in the end, they said, Well, you know, we can normalize everything into HTTP and rest. And so which is the communication protocol, you talk to, you know, your, your, your Amazon, store, everything, you know, your teleconference, and everything goes to that. Now, that has been great, but it's not for machine level, real time type of speed. So kind of devious provides the equivalent when you're trying to do like, real time performance, you know, latency sensitive type of applications.

13:40

Now, that makes sense. We're, you know, you, you, you've been involved for 20 years. So you've seen a lot of changes happen, where do you see it going now? What's what's sort of on the horizon? i You seen,

13:54

15:08

you want to be able to be able to pull in using the, the healthcare analogy, I want to, I don't just want this one device just be standalone, and I can't, I can't see how other devices, I can't start to pull in that data in a way that is meaningful to me or, or have access to it. And I might have multiple hospitals and, and so I want to have multiple viewpoints, but I just want to pull everything together and can't do it. That is it scalable. I like that. You were absolutely spectacular.

15:45

Thank you. Yeah,

15:47

I really enjoyed that. How do people get a hold of you?

15:50

Constantly, RTA, you can find me in the webpage there, get out of that rti.com or reach out to RTI anybody will be able to get you

16:00

rti.com, right. Not Real-Time Innovations. It's art is Real-Time Innovations. But it's RTI. Exactly. Make a note of that listener. Respect spectacular. Early. Thank you very much. I really appreciate the conversation. All right, listeners, we're gonna have all the contact information out on Industrial Talk, and make a note, it's rti.com ar ti.com Go out to also only m g.org. Get engage. You get to meet people just like the one we just had a great conversation. Thank you very much for joining. We're gonna wrap it up on the other side. Thank you. Stay tuned. You're

16:39

listening to the Industrial Talk Podcast Network.

16:49

His name is Gerardo. Make a note of that. So all his contact information will be out there on Industrial Talk. So if you're not you need to reach out. RTI is the company he is with also co-chair of Data Distribution Services, that is with the wonderful organization called Object Management Group or OMG. That will also be out on Industrial Talk. If you find yourself in need of an organization to get involved with, yeah, you need to connect with OMG to as well, great conversation. There's a lot to unpack there. We just take it for granted. We just and I guess that's really good. If we see how the sausage is made. Well then. That's not good. So kudos to the team at RTI. All right, we're building the platform. This platform is dedicated to industrial professionals everywhere you have a podcast, stick it on Industrial Talk. You have the need to sell something, stick it out on Industrial Talk. We're going to have another great conversation shortly. So stay tuned.