Stan Schneider with Real-Time Innovations

SUMMARY KEYWORDS

IoT solutions, autonomous vehicles, teleoperation, data-centric technology, automotive systems, cybersecurity, software-defined vehicles, ECUs, distributed analytics, edge autonomic, real-time data, low latency, smart systems, industry disruption, RTI innovations

00:00

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. All right, thank

00:22

you once again for joining Industrial Talk, the number one industrial related podcast in the universe that celebrates industry professionals all around the world. You are bold, brave. Dare greatly, you innovate, you solve problems, you make the world a better place. We truly appreciate you here at Industrial Talk, and we are broadcasting on site. IoT solutions, World Congress, Barcelona, Spain, is the location. And I cannot stress this enough, you need to put this one on your calendar for next year. If you're not here this year, you're missing out be here next year. This is where all the cool kids are hanging out in the world of IoT. Can I have Stan? Can I say cool kids and IoT in the same thing? Whatever? Sure, there it is permission. There, there he is. Stan has given me permission, and Stan is in the hot seat. And Stan has been in this on this platform a number of times, talking about Stan stuff. So let's get cracking. Craz get cracking Stan stuff.

01:22

It's been a great thanks for having me. Yeah, Scott stuff. No,

01:27

you don't want to talk about Scott stand stuff. Stand so is much better. And you, you're far more intelligent than Scott stuff.

01:37

I didn't mean you, there's lots of Scott there's a

01:39

lot of stands. But no, this is specific to you. Stan, so you having a good conference? Yeah, it's

01:45

pretty good.

01:45

Yeah. Do you have a booth here now? Are you just sort of

01:48

wandering? No, I'm. I hosted this session, and I have a talk today.

01:52

What you talking on,

01:55

talking on data centric technology for teleoperation, we have recently been doing a lot of work with automotive systems. We've been working on autonomous vehicles forever, and one of the things you need to be able to do is get them unstuck. I'm actually talking with one of our customers, ESA European Space Agency, that has been working for a very long time on tele operating space rovers robots, yes, so it's, you know, it's sort of leading edge research robotics, and it's surprising that there's actually quite a bit of commonality between that and what we see with a lot of our customers doing similar kind of Vehicle stuff.

02:39

I was having a conversation with somebody specific to autonomous vehicles, and just the the amount of code, the sophistication to to make it autonomous, is just through the roof, yeah. And then, of course, we had the cyber conversation. And of course, that's a big deal too, as well, but, but it's happening, right? We're gonna have autonomous vehicles, yeah,

03:05

if you can define them, it's happening. So, I mean, you want me

03:11

to define it, I get in and I go, and I don't have to do anything, and it gets me to point B.

03:15

Is it snowing? Are you on dirt roads? Are you able to go through

03:19

See, I'm just a consumer. I just want to go to point B. I don't care what the conditions are.

03:24

Autonomies like happiness, you can never be fully happy. There's no such thing as fully happy. There's no such thing as fully autonomous. That's interesting. It's a range. Eric, actually, you've heard the level 12345, and I'm not wild about those levels, because it really, really matters about the environment, just like happiness, you can be happy at work and not happy at home, or vice versa. It really depends on where you are, what you trying to do, really thinking through

03:49

I mean is, let's say I'm all in there's autonomous vehicles. I'm a consumer. I'm not interested in knowing how the sausage is made. I just want to get into that vehicle and I wanted to go to be is that even plausible, or is there going to always be this responsibility on my part where it's sort of semi autonomous, but I better stay alert, because maybe that autonomous wants to not be autonomous, and I'm going to have to be called into action

04:19

there ever is a damn long time. Okay, so will ever happen? Of course, it will. I mean, might be five years, might be 50 years. It might be 500 years, depending on how you define autonomy, but it's absolutely going to happen. It's, you know, technology moves forward. It's an obviously compelling use case in many different kinds of things. Some things will happen earlier. The easy environment thing. So we've been doing autonomous vehicles. It's not all we do. By any means. We have over 250 designs of autonomous vehicles. Only about, I don't know. 10% of those are road vehicles that you could get into. Maybe 20% you. Yeah, most of them are we started out with flying things, or have 100 different kind of flying things, much easier to fly, by the way flying things, underwater things, surface things, mining vehicles, construction vehicles, forklifts, delivery robots, on and on and on and on. They're all autonomous vehicles. I just define a vehicle as something that moves and can operate for some period of time without help. A great definition of sort of academic definition autonomy is how long you can go without a greater intelligence helping even you are Scott, because you probably need help every now and then and so, yep,

05:37

don't talk to my wife. Yep.

05:39

I mean, it's just you gotta, you gotta look at autonomy as a level of intelligence and a level of fit to the environment. And if you fit the environment well, and you're quite intelligent environment, you're you're more autonomous, but you still may need help direction, something along the line, and that we call teleoperation. It's not, it's not. That's another thing. There's many different kinds of right? It can be strategic teleoperate, you know, normal. Tell, if people think of teleoperation, you sitting in some, you know, office somewhere, driving a car, yeah, which is not a very compelling use case, because if you're going to drive the car, you need to have full attention on what's going on. You might as well be in the car. It's it's not that big a deal if the car happens to be on Mars, well, that makes more deal. Of course, that's true, but if you're, you know, if the car can handle many things on its own, every now and then it needs help. Or, you know, trucks, a better example, handle the freeway part well, but it can't do the last mile pickup and delivery. In the end, you can pick up and do teleoperation for those segments, and it makes more sense. The real hard part of teleoperation actually turns out to be the handoff. When do you hand off? How do you hand off? Where is that operator? Center, all those kind of things getting out.

06:59

It's happening. It is happening.

07:01

It's happening. It's been happening for a long time. It will keep happening. You know, when can you ride in a car on your own? You can go to San Francisco ride in a way more car. Right now, it's become the hottest tourist attraction in the city, really, yeah, it's on TripAdvisor. I think it's in the top five tourist attraction. Take a ride on Waymo. No way, but, yeah, but, you know, clearly I call is it profitable? That's long ways

07:32

from being but, but from an autonomous point of view, we have autonomous vehicles in production now, right in commercial operations,

07:41

there's a few in our profitable Why are

07:44

you in this? You're all into it. But

07:48

no so. So we do a lot of road vehicles only maybe five Buck five, and between five and 10% of them are shooting for high end autonomy. Most of them are building what's actually much bigger disruption in the industry called software to find vehicles. It's really a completely different way to think of how to make cars, and it's tearing the industry into pieces.

08:12

I don't understand the explain what? Why is that if

08:15

you open your if you have a reasonably recent car today, and you look inside, it's going to have four or 500 little computers called ECUs, electronic control units. These do something, usually some very well defined thing, and they're connected together over what I would call a stupid network. CAN bus. I'll get shot for calling it stupid. But compared to Ethernet, or iPhone bandwidth network. It really isn't very flexible with a very simple interface. So if you're doing, for instance, a smart cruise controller that's watching the car in front of you with a radar, all that thing is going to do is take a speed you want to go. So we'll go faster than it'll tell basically, the car speed up or slow down. It's the only interface it has, and that's great if that's all you want to do. But if you want to use that radar signal, which could have a lot of information, to differentiate the car and the pedestrian and something else, you can't really do that, because that's not what that little part does. So anyhow, if you look in your car today, you'll have four or 500 ECUs. They're all software that was built before the car went into production. To replace or fix a bug in the software, you have to do a recall. Recall is a dirty word, yeah,

09:34

when you said that, I just said the

09:36

only way they can fix is to replace the hardware part. It's sort of insane that. So that's sort of the way cars have been built forever, and it also is the way the industry supply chains are set up, right? Yeah, these tier ones and tier twos. And what they do is they take these specifications for a smart cruise controller, and they make a smart cruise control, and you can buy it from Bosch or Conti or Magna or whatever, and put it here. Are. That's sort of the way the world set up. But software doesn't work that way. So a software defined vehicle a huge contrast. Instead of all these lots of little computers, you'll have one or two, you know, few big computers, maybe one big central one, like what Tesla did, maybe smaller, big ones that was called zonal architecture, that really are their biggest goal really is to reduce the wiring. Wirings, third heaviest, one of the top most expensive parts of the car. I know that. And you can get rid of a lot of wiring just by putting compute power out in the in the zones, you know, the right front area, right and then connect to that with a high performance bus, and now you can update that software. So Tesla put everything in the central thing. They can update everything. It's amazing what they can update, but that they don't have this whole supply chain. And this the transition from this ECU based, which nobody calls hardware base to software based, software based design for the fine vehicles, what it's called is just absolutely transformative in the industry. And there's another huge disruption going out, getting rid of the gas engines to electric but this is at least the largest thing you know. Of course, if you're trying to make it autonomous and smart, you need high performance computing, and that's another thing you can do with a software defined vehicle. But most of our customers are just doing a new architecture called the EE architecture, for the vehicle to make a new kind of vehicle. And that's that's really what our company, my company, RTI, real time innovations. We're actually the largest infrastructure software companies, what we call the smart world, which I can define quite well for you. The smart world is taking large systems and making them much more intelligent. So

11:58

we do things like how to define large system like, what's a large system? We

12:02

run Canada's air traffic control system, NASA's Launch Control System. We actually run the power balancing North America. We those are very large. Yes, those are larges. We also do a lot of medical robotics, billion dollar programs. We do medical monitoring systems, imaging systems, lots of renewable energy defense systems for over 400 defense programs of record, all the data flowing through like an aircraft carrier, combat management system get shot down or, jeez, that's, that's what we do. We're trying to build these intelligent, distributed, very super reliable. Gets really hard when it has to not fade, right? Most systems, like even this, this show, which is sort of turned into what I call the monitoring, there's really three markets in IoT, by the way,

12:57

you gotta go down that I gotta go down that aisle.

12:59

But okay, with three markets? Let me go back to the other one later. Yeah, the three big markets not have great names for them, but we've been calling what you see here, where you have lots of little devices that are all connected to a single cloud service, right? Right? The monitoring market, it's the biggest market in a lot of ways, because it's the easiest market so your nest

13:22

Sure, sure that's exactly correct. Yeah, they're

13:24

really mostly just monitoring things. You can control things, but it's some one little device. Those devices don't share information very much, except with the cloud, and overall, it's really just making those little things smarter. Yeah, the second big thing that sort of analytics, distributed analytics, where you take a single thing, like a pipeline or a factory, and you sensorize it for 1000 or 2000 sensors in it, and you collect all that stuff and get it up into the cloud as big data, and think about it, and you're gonna think about it to make that factory run better, or decide what it's later. But it's not directly changing controlling the machines. It's just making it smarter overall. There's not a lot of that here, although you will see a lot of people like, you know, selling sensors that you would Yeah, choose 10,000 Oh, and put them in a fax.

14:17

Yeah, yes, that. But that's a completely

14:20

different market, and it's dominated by the cloud players that are trying to provide the data analysis and big data. And, you

14:29

know, right, right? And then

14:30

the third thing right there, AWS, the third thing you'll see a lot out there is we call edge autonomic. Oh, yeah, we're trying to make something work differently than it used, and typically smarter, more autonomous. That's where we play as RTI trying the autonomous vehicles, smart transportation systems, a Navy ship, whatever. Those are, all single things that you want to make run better as it is. But system very different than the other two things, right? Yeah. And then the biggest thing that you know, I can ask five questions and tell you, if you need our software and you don't have to know what I do and I don't know what you do, it's really

15:14

so right off the bat, no, I failed number one. So file,

15:19

no question number one, is it a big problem, like maybe front page news or a complete, complete failure of the system, if it stops for one minute, one minute. So if you look at most of the sensors here, where they're doing Power Monitoring or temperature thermostat, or it doesn't matter if it's down for a minute, you wouldn't even notice it, right? But most websites, you know, they don't. They don't care if they're down. That's the first question. Second question, you know, you don't have to know what I do to do that. Have you said the word millisecond in the last two weeks? Because if you haven't, then you're probably either at human speed or below, and human speed or below. There's lots of ways to solve things don't require this kind of data centric model. I haven't got to talk about yet, very fast communications and information between devices. Third question, and you started with this with the code. Do you have at least 10 people on this project that write code for learning software engineers? I don't care what you call them. All they do is they write, and as soon as you have that, as soon as you get to about 10, it's really before 10, but your cost and your success is driven by software integration. People don't see that in this industry. They do many other industries, software integration becomes absolutely the finding characteristic of your success and your cost and everything going on, and what we fundamentally sell is software innovation technology, which is why we ask that fourth question is harder to ask, but really, is it? You know, are you doing more than just collecting data? Is data something static that you can collect if all you can do is, if you want to just collect data, there's other easier ways to do that. If you want to act on it right now, get the right data at the right place, at the right time and the right many. There's many places typically, I usually just ask is, are you sending data on one place? Knowing what that meant, but really, if you collect data, you want it all in one place. If you're using data to do something in real time, like, control a vehicle going down the road, or, you know, keep airplanes from crashing, then you need to have it. Now, it doesn't really matter where the pedestrian was.

17:30

You don't want latency, right? You don't want that.

17:34

You have to have low latency. And you have to low latency to multiple points, very rare. You want just one thing with low latency, whatever. And then the last question, which is, why we're, we're pretty good sized company for the largest software coming in. Protect yourself, really. You know, we're about 300 people instead of 30,000 people. Is, are you willing to build a new architecture? Because if you ask the four first four questions, yes, you can't do it with your architecture. It's something most people build these system focused architecture. Where am I going to put the server? What networks do I need when they really need a data center architecture? What data do I have? Where does it have to get put that together first and then decide if you're going to have servers at all, which is usually a problem on the edge,

18:19

I have to ask this question. It seems to me that it would be easier to achieve what what RTI provides at a beginning of a project like you have this conversation, and you're you're designing it in such a way that achieves what needs to be achieved, and you It satisfies all of the questions that you but, well, how do you retrofit something that has been in production for like, let's take an aircraft carrier that's been in, you know, out there in the ocean for a little while, and you're trying to rip all

18:56

the electronics out. Yeah, our customers have three things in common. First thing we have in common, we sell in architecture. It's very difficult to retrofit an architecture not that it hasn't been done. We're running Grand Coulee Dam was built in the 40s. Yeah, there you go.

19:10

That's a great example.

19:11

I can tell you a whole story of Grand Coulee Dam, if you want. It's fascinating, but they, they ripped out all the electronics. It really changed its mission. It used to just produce energy. That's all it did today. Its main mission is actually balancing the grid with all

19:26

the solar. Sure, sure, sure. So

19:28

frankly, is the largest power plant in North America. Everybody knows that. It's also the fastest power plant of any size in North America. It can take its seven gigawatts of power on and off the grid in about 10 minutes. Oh, yeah, there's nothing else you can't do that with it. With a oil fired plant, it takes many hours, yeah, to safely shut it down. You can't just stop putting power out. It'll it'll blow up. You know, nuclear plants can take days to safely shut down. The gas peakers, you. They're faster, very fast, but they're very small, and they so whatever they're using that the hydropower portfolio now to balance the grid as wind and solar, which are, you know, challenging clean energy, but dirty power? Yeah,

20:15

they suck bars and they have, they have some problems,

20:19

right? They're clean energy, but they're dirty power, because you can't count on them at any time. You can count on things like solar going away at night. Happens every night. Yeah? Guarantee it, guaranteed. Yeah, right, right about the time when all the power demand is going to peak, yeah? Is when it's shutting down. Yes, another whole problem. Yeah, exactly.

20:44

They got a, got a friend going on over there? Got

20:46

a bug? Yeah, software,

20:51

yeah. So, so it, it's hard to, it's a challenge to retrofit, right?

20:56

So, first thing, they're all doing something new. Second thing, to build that new architecture usually costs $100 million or more, 200 million. We have customers who build an autonomous vehicle. It costs billions, I'm like kidding, billions. So making that decision is risky and it's scary and it's big project, and because of that, it's new and it's expensive and risky and scary, and most people are betting their careers at least on it. The third thing they have in common is they're all terrified, every one of them scared. And our role in this industry is sort of be the safe you know, we sell trust. We Well, yeah, try and provide the trustworthy partner that's been through this literally 1000s of times. You know what you're likely to run into? We can help you out with it. You know? It's, that's our, that's our, yeah, see,

21:52

I, I, it's a brave and exciting new world,

21:59

smart New World.

22:00

Smart New World. Smart New World. I do all the smart world. Smart. We've been

22:04

looking for a name for it forever. Maybe you can help us out. You're me first. No, the smart world, to me, is defined as smart systems that work together to do something better. And we need a smart world, because today's world is pretty stupid. Yeah, it truly is. I gotta get rid of this.

22:22

I know, yeah, get that fly out of here. The reality is one that everything now, now smart, the term smart, is becoming that miscellaneous file again, because now all of a sudden, everything wants to be smart. And it's not bad. I have no problem with it, because I'm all into smart, right? But it's then all of a sudden, it becomes sort of hazy on this intelligence. There you go, right here, right now. Industrial Talk intelligent. Well, I don't

22:56

know what else to call it. No more autonomous. It's also more intelligent. Another, another definition.

23:01

Stan, we've already burned through a bunch of time. See, you never discipline. Even

23:08

need you. I could talk.

23:12

I

23:15

I'm sorry I didn't know

23:17

it was great. I knew I could just sort of pull the string and push play or whatever you want to do it and everything would have just it. I don't know. It's just an interesting time. Your perspective is so valuable. I really, I really appreciate it. Yeah, when you walked up and said, Hey, I'm here, great. Check it out. Scott,

23:39

you've been a real you've been an icon at this show for a long time lots of other places. I love your podcast.

23:45

Oh, thanks. This is the great Hey. How do they get a hold of you? They want to get a hold Easiest

23:51

way is connect with me on LinkedIn. Stan Schneider just got a number, but pretty obvious, RTI or send me an email. Stan at RTI time and rti.com and I promise to read it in about two months.

24:07

Or you can just go out to Industrial Talk, type in Stan in his other broadcast, his other interviews will pop up, and he'll have all the contact information to stand there, because we're gonna have it out

24:19

there. Dan Schneider, LinkedIn, easiest way to connect.

24:24

Yeah, all right, listeners, we're gonna wrap it up on the other side again. Connect with Stan. You can tell he's we're just just barely scratching the surface on what Stan knows and and he's there to help you. He wants to collaborate. Make it happen. Captain, we're broadcasting from IoT solutions, World Congress, Barcelona, Spain. It is a wonderful event. Put this on your calendar demands. Stan has been going here for how many years I

24:46

was on the I was actually the Vice Chair, that's right, I see steering committee that founded this thing. I was about 10 years

24:53

ago, yeah, well,

24:56

it's been through a lot of through a lot of changes. All right? And the pandemic. And don't forget the pandemic.

25:03

All right, we're gonna wrap it up on the other side. Stay tuned. We will be right back. Yeah,

25:08

you're listening to the Industrial Talk Podcast Network.

25:17

Yeah, Stan is the man that is a download of Stan's stuff. You need to take a pencil paper, write that stuff down, or go back listen to it, because Stan stuff never disappoints. Stan ever disappoints. He's a he's amazing. He thinks through all of this stuff. Cool. It's really cool. I really enjoy IoT solutions, World Congress was the event you get people like Stan. As you go to that event, it makes sense, you're in the world of digital transformation. Are you in the world of just all of that? Yeah, that's an event for you. Make it happen. Captain, all right, I have to apologize because we had a video glitch, so you're not seeing Stan as much as you should be able to see Stan. But we prevail. The talk is good. We are building a podcast. Do it right here. You need to tell your story. You come to Industrial Talk. You need to highlight your technology. Come to Industrial Talk. It's all here. Be bold, be brave. There greatly. Hang out with Stan. Change the world. We're gonna have another great conversation shortly, so stay tuned. You.