John Krzesicki talks with Carrie Goetz about the trends in edge computing, a movement that shifts data processing from the cloud to local data centers closer to the applications.
Transcript
Welcome to the PSR PowerTALK podcast, produced by Power Systems Research.
00:06 John Krzesicki
Hello, my name is John Krzesicki with Power Systems Research. I’m a business development manager for the US and with me today I have Carrie Goetz who is a fractional CTO for multiple companies. She’s been in the business for many years. She was also voted as being one of the most influential women in the tech industry in 2020.
Today we’re going to talk about the Internet of Things and its impact on smart city, smart community, and industry and love to get her viewpoint on where she sees the market going.
So, Carrie, thanks for joining us and let’s start the conversation of, you know, what are you, what do you see in the next few years and what are some of your visions with smart city and some of the IoT technology?
00:54 Carrie Goetz
Well, hey John, thanks for having me to start.
So IoT to me is, first off, I hate the acronym. I’m the first to tell everybody I hate the acronym because it says absolutely nothing, and you can make it say anything that you want. But I think that when we talk about the Internet of Things, we have barely scratched the surface of where that’s going to go and…
There’s this guy named Parkinson, he had two laws for the computing industry. One was for storage and the other was for bandwidth and pretty much both of them say if you put it out there, they’re going to fill ‘em up. And I think we’re kind of at that point with IoT. Where, as more and more technology comes alive through the, you know, that’s automated by sensors and artificial intelligence and machine to machine learning, that we have no idea where all this can go.
I mean, we’re barely scratching the surface, and this could be all kinds of things. So, you’re seeing these sensors for smart water meters and smart power meters, and, you know, that gets coupled with leak detection in somebody house. And then from, you know, a smart city perspective we can use it for security and finding adverse events like gunshots. And, you know, having that control where the cameras zoom in and follow the noise. And precision agriculture and how we’re raising crops, figuring out the best water amount to put in the soil, that won’t deplete certain nutrients, and, you know, the best day to make that water happen. And maybe it’s not the same for the whole field. Maybe part of the field works differently than other parts of the field, and so it really allows us to get granular with some of the things we’ve been doing.
From an industry perspective and automation, I think, you know, robotics and what’s happening with industry 4.0 is great.
I think it’s a great way for rural healthcare and people in the middle of nowhere to get access to healthcare. You know there’s a, there’s a little device now, I’m sure you’ve seen the commercial, you can stick it on your iPhone and give yourself a medical grade EKG on the spot sitting at your desk.
So, you know, I think a lot of these technologies, you know, we, we just scratched the surface of where we can go with this stuff.
03:05 John Krzesicki
So, uh, again, there’s just, there’s a lot of data, there’s a lot of information floating uhm, and every day we’re hearing about new technology invading our cities and our communities. So, so, you know, one question is, you know, where, where does all this data go? How does, how is it managed?
03:24 Carrie Goetz
Yeah, so that’s a big one. So, there’s, there’s different places you can push data, right? So, some of, some data doesn’t need to be stored. If you, a good example, if you go to the grocery store and you buy 400 items. The credit card company that you charge that to doesn’t care what 400 items you bought. All they care about is that a transaction happened, here’s the approval code, here’s the card you’re using, that’s all that gets pushed.
But if you think about it from the grocery store perspective, they want all of that information. They want to know everything that you bought. They, you know, down to the granular. Not just for inventory, but sort of trends for the store. So, they’ll trend that. They’ll use your ZIP code to figure out the buying trends of people in those zip codes and, you know, different things to restock or maybe new products to bring on.
So, all of those are examples of that data being kind of all over the place, right? The corporate entity, if it’s a multi-site grocery store, probably just wants the result sets. They don’t need to know what time of day you bought or other things that might give information to other insights. And so, all of those are places for data to go, so it can either be stored at the grocery store, right, or at an enterprise level in the city, for a smart city, or the city controls the data center.
It can also be at the edge. And at the edge that can be a smaller data center, say, you know, sub, sub one megawatt or somewhere in that range. Or it can be the larger enterprise space that we talked about, or the cloud, and that can be the public cloud, being AWS, Azure, Google Cloud Platform, any of the other IBM platforms. And that could be where you’re using somebody else’s hardware, you’re just doing software licensing. And normally what’s going, what you’re going to find is it’s going to be some combination of those. There’s no one size fits all. But we do know that moving into the cloud is very, you know, pretty much nothing other than spinning up your instance.
Moving around in the cloud and moving data out of the cloud is expensive. So, if you try to figure out, from an application perspective, what data are we gathering, what are we going to do with it, and where does it need to live, then that starts helping feed your decision about where that data is going to be.
Because, you know, these sensors put off continuous data. And it might be that you don’t need to know that the temperature was 78 degrees every second throughout the day. You might want to know the average temperature, so that’s two very different amounts of data. And so those are the kind of things, and the decisions that I think are going to drive, you know, how much is going to be at the edge and how much is it going to be at a bigger data center.
And at the edge it might just be an all-inclusive device, like a little hive cell, or it might be, you know, one or two cabinets out there. It might be a self-sustaining data center with immersion technology. Sitting out there that, you know, people are writing data to. It can be any combination of things. So, I guess edge just really defines the location of the data being closer to the application and users that are going to use it. And I think we’re going to see just large combinations of all of the above moving forward.
06:32 John Krzesicki
So, give me just a little more of a, you know, yeah, I certainly understand what the cloud is, but can you kind of just give me a little more definition around edge, edge computing and why is that important?
06:44 Carrie Goetz
So, I think a great example is content delivery networks. All these streaming providers, you know, there’s, it would be impossible for everybody to stream the latest, best, Netflix movie, for instance. If we all jumped on the same Netflix server, there’s not enough compute to make that happen. There’s not enough bandwidth in that pipe to go in there to make that happen. We would all get that little spinning wheel continuously and, you know, our frustration levels would increase. And so what happens is they take copies of that movie and they spread it all over the place. And that’s why you see, like on Netflix, you’ll see trending movies or trending this or that because they’re keeping up with that themselves and they’re figuring out what they need to push out to all of these other places.
So, it’s like, it’s like Santa’s helpers. I guess it’s probably the best way to put that, right? You’ve got the main, main dude in the North Pole and he’s got all these helpers that are running around listening to all these little kids. And so, it’s very close to the same kind of theory, right? As, as we have this data all over the place we want to process what we need to at that location and then you know the rest goes somewhere else.
Police cameras are a great example of that, right? You know, you want to know when that policeman has had an interaction; but there’s a lot of data that’s captured during the day that means nothing. So why do we want to send all of that to the cloud when we can process that locally and send the result sets to the cloud? Because now it becomes a very financial decision as opposed to a convenience decision.
08:08 John Krzesicki
So you touched down a little bit about smart cities. What other industries do you see moving towards the edge and this type of infrastructure?
08:16 Carrie Goetz
So, I think rural healthcare is a big example. Industry 4.0 is, you know, and what’s going on with industry and robotics and, and how we, the amount of information that we want to capture for every single manufacturing line, it’s, it’s pretty insane.
I think that education is kind of ripe for this, even if we do, you know, smaller groups of schools, maybe. Doing some of the edge compute there.
I think bridging the digital divide is, is a good one.
I think that smart cities and smart communities, smart transportation are all part of this.
Drone deliveries are certainly something that’s going to use edge compute. I mean, you know, like I said, we’re just touching the iceberg. If you could order a pizza and have a drone show up at your front door with that pizza, that’d be pretty cool.
And, you know, the Jetsons when we were growing up. But a lot of these young kids probably don’t remember that show. But, you know, you had the the maid that was really smart. You had meals instantly delivered. You had all this kind of stuff. You know, that was in the Jetsons house and they were not, not too far off from some of those things coming true.
Autonomous vehicles are a good example, and probably one of the ones that I think drive. Between content delivery networks and autonomous vehicles and, and fleet management. Those definitely are, are big drivers too. I think especially for a lot of the areas where there is no connectivity, and, you know, no signal, right?
09:45 John Krzesicki
Yeah, you just, you just touched down a couple points I thought were interesting; in terms of a, you know, a lot of data, a lot of connectivity, rural areas.
Now we also understand this is going to be a stress on power. What is, what is your vision on how to support all this data of the data centers, the edge and what do you see kind of happening in the future?
10:07 Carrie Goetz
Well, I think that edge data centers have a really cool advantage when it comes to power; because it’s a lot easier to do something at a small scale than it is a giant scale, right? So we’re not building 50 MW data centers at the Edge we’re building sub one MW, which is relatively easy to power off of solar, renewable natural gas or, you know, one of those other wind farms, some of those alternative methods; because solar and wind take a lot of land, and if we’re trying to do this at a smaller scale, it’s easier.
You know, you can put the solar panels on the roof of a smaller data center. You can use submersion cooling to make sure that the power demands are lower because, you know, the heat projection is happening through liquid as opposed to air, which you know brings down the power requirements.
So, a lot of those things happen that I think are really cool and we have to realize that a) the grid is taxed and a lot of places where these data centers go there’s not enough power to support that data center sitting in the middle of nowhere.
So, if I’m a farmer and I’m working on my precision agriculture snd I’ve got, you know, sensors all over the fields, and, and I’ve got data gathering requirements and I’m trying to sort out how to run my farm through all of this technology, I’ve got to have something that’s going to process that for me. And I think that’s where these edge, either facilities or even compute devices, really are going to come into play and be able to support that.
But from a power perspective, really driving down the need to where we’re not adding in things that draw power unnecessarily, that we’re being the best stewards of power that we can. And I think the edge gives us all of those opportunities. Again, because it’s a smaller footprint and I think it’s easier to “lather, rinse, and repeat” and do some innovation at that smaller scale and, you know, make those to be tried and true methods.
12:00 John Krzesicki
OK. Great, great. Thanks for the insight.
As we move along with the series, or, this is the first series of hopefully many, but we’re going to take a take an opportunity to address smart cities. We’re also going to have some conversation around agriculture, smart, smart manufacturing, smart healthcare as we roll up these, these podcasts.
Now the, you know, to, to finish off this podcast, you had mentioned the digital divide. I thought that was really interesting. Can you give us what’s your viewpoint of the digital divide and how edge computing could help or could actually close the gap?
12:38 Carrie Goetz
Yeah, so the digital divide I think is massive and people don’t realize. And it’s not just rural areas, right? There’s certainly rural areas where they’re dead signals. We, we’ve all driven through those, you know, when the phone doesn’t work, “Oh no, got one bar, I’ll have to call you back”. That’s the digital divide and not just there, but even places where there may be Internet, but it’s insufficient to support things like streaming and actual Internet use. So, people having true broadband connections.
And I think COVID really brought to light how big the digital divide is, and it certainly made it worse; because a lot of people lost their Internet plans, you know. If they’re not working, they can’t pay for their Internet plans. So, what used to be one in 10 households, now they estimate to be closer to one in four households in the US is living in a digital divide, which is pretty crazy. And that’s people that don’t have access to reliable broadband.
And I think too we sort of highlighted how the over-subscription model of a lot of these providers is woefully negligent when everybody all of a sudden works at home.
And so, not just in the US, but if you get outside US the digital divide is pretty massive. There’s entire countries that have maybe 4 Internet cafes sitting in cities that you go to and nothing outside of the main populace of a city.
And so being able to bring this, I mean, I know we’ve gotten to the point, I think, with Internet and information that this is really kind of a human right. You know, you should, everybody should have access to be able to learn new things and everybody should have access to be able to schedule critical appointments and those deals.
And so I think edge, as we start building out the edge, that gives us this huge opportunity, because all of those edge data centers are going to be connected back to the Internet. And all of these edge devices are going to be connected back and as we start opening these up, which I think is absolutely going to happen, then we start healing some of this digital divide and people have access that didn’t have access before.
Microsoft is doing a big push right now putting containerized data centers out in the middle of nowhere with satellite communications going back.
I mean there’s a, there’s a lot of ways for the bits to get from point A to point B; but certainly this digital divide is massive. I think that it changes the way people work. I think it changes the way that people go to school and learn and just the quality of life across the board. And my favorite part of tech has always been tech for good.
And tech always figures out a way to solve its own problems; although sometimes I think we solve other industries problems before we solve our own. But, certainly having that digital divide to people just having access to those resources is going to be huge. And we saw this in COVID we saw what happened when, you know, churches were having to turn up Wi-Fi in their parking lot; because there was communities that didn’t have Internet, but all these kids were expected to learn on the Internet. And so all of those, I think it was kind of a really good shake-up as far as bringing attention to the fact that not everybody has those pretty basic services.
15:55 John Krzesicki
Thank you for the insight, Carrie. That was a great overview of the digital divide.
So, that’s going to conclude our podcast for today; but I would just want to send out a reminder that we are going to be addressing, again, smart cities, industry 4.0, smart ag, and some other industries as we continue with the podcast series.
Carrie, again, thank you for your time. I really appreciate your insight; and, and have a great rest of your day.
16:21 Carrie Goetz
Yeah, thanks John. Have a good one.
16:22
To read a transcript of this podcast, visit the PSR website at powersys.com and click on the podcast archive.
Thank you for joining the PSR PowerTALK podcast. This podcast will be available on demand. Search for “PSR PowerTALK” in your podcast app or Google “Power Systems Research” to sign up today.
