Alexis
Hi everyone. Welcome to Talk Tech with Data Dave. I’m Alexis, your host of this podcast.
During our last episode, Dave and I started talking about the Internet of Things and how the Internet of Things impacts data management. Just like I said before, we had a long conversation about it. The conversation went so long that we had to break it into two podcasts. So, this is the second part of that conversation.
Where we left off, Dave and I were still talking about what things make up the Internet of Things. We were just getting into how it can impact data management.
So we’ll start here.
Data Dave
Realistically, the diabetic sensors for blood sugar that people wear on their arms and so on. Now that’s Internet of Things, right? Yeah. It’s a sensor connected to somebody that they can read themselves on their mobile phone or it transmits directly to the hospital, et cetera, to say my blood sugar level is this. Tracking the location of your partner on his mobile phone is the Internet of Things. You’re not really tracking the person, you’re tracking the phone and the phone is telling you where it’s at.
Alexis
Yeah. And then I can make decisions based on that, but I don’t need to keep all of that data overall time, right?
Data Dave
Exactly.
Alexis
I can make decisions based on whether or not I see my partner at the gas station or at Domino’s. “Hey, pick me up a pizza too.” But I don’t need to know where he is at all times.
Data Dave
No.
Alexis
And that’s such a lackadaisical example there. But the big picture here is that you really can have a huge impact on your data management if you’re in that world.
Data Dave
If you’re in the manufacturing world, if you’re dealing with a lot of equipment, a lot of things, then you have to pay attention to it. It’s a different kind of data than what we normally deal with and talk about, because we’re normally talking about business events rather than instrumentation events. But they do correlate. If you’re in that world, then they absolutely correlate to say, maintenance of an oil rig, maintenance of an aircraft, knowing where the aircraft is. We’re starting to talk about the same kind of thing with how do we track and how do we understand things and how do we collect the data from things? So it can affect your data management tremendously because you want to feed the answers for that into your decision-making processes as an individual, as an organization, and that should be informative to your decision-making processes.
It can be a huge amount of data, so you have to make particular provision for that kind of data. And you’ll find that a lot of the cloud vendors, a lot of other vendors have what they call streaming tools where they take a stream of events and you know, a lot of this is events that can be dropped. If I don’t get every millisecond, do I really care? As long as I got three out of the last second, that’s probably enough.
The sensor might be pushing out one every millisecond. It can say, “My temperature is this, my temperature is this, my temperature is that, my temperature is the other.” And I’ve got to know where that piece of equipment is. If I drop one of those messages, doesn’t really matter. I can still get the trend.
Alexis
Yeah.
Data Dave
So you manage it a little differently and you manage it in a particular style, particular approach, and then you’re looking for the patterns that establish what becomes a broader event that you want to actually pay attention to.
Alexis
It’s interesting how you said a lot of the cloud providers have streaming plans, whatever you called them. There would be no other way to manage that besides in the cloud, because it’s so much data. I just started thinking about that. I was like, man, I mean…
Data Dave
We have managed it outside of cloud, but you’ve got a lot of infrastructure put in place, especially when it’s a true Internet of Things. If you think of a factory, and I’ve got the control systems on the factory that are just dealing with that factory or that plant.
Alexis
Okay.
Data Dave
That’s one structure where I’m pretty local, pretty private. I can do that myself.
Alexis
Okay. Yeah.
Data Dave
If I’m managing a fleet of cars.
Alexis
Oh, right.
Data Dave
But then I need a national or worldwide backbone. And so you find that a lot of this data nowadays is often being trafficked by the telco providers, the cloud providers, etc. And they become aggregate points. Because you’re feeding onto this Internet backbone, you need the communications backbone behind it.
We might have had, back in the day, something might dial up once a day and tell you how it did that day.
Now it can call back every millisecond, every second, every minute, every whatever.
Alexis
That’s crazy.
Data Dave
Broadcast its state all the time.
Alexis
Right.
Data Dave
And it can be a crazy amount of data.
Alexis
Just because we just kind of got off this data architecture path in our last few episodes. I just want to ask one more question to you about this.
Data Dave
Okay.
Alexis
From an architecture standpoint.
Data Dave
Yes.
Alexis
When you’re building data warehouses and data lakes and data swamps and whatever else you’re building, data ponds, as you call them, for Internet of Things versus the just kind of consumer data that we are often talking about, is the data architecture different or is the idea the same?
Data Dave
So, it depends what layer you’re coming in at. At the base layer, it’s very different because you’re talking about streaming data and time series type data more than anything else. This is where I’ve got data that I can throw away. I’m more interested in the trends than I am the actual data itself. And I’m interested in the results of the analysis, not necessarily the data itself.
So, a discrete sale. I sold a car. I’m very interested in that transaction. I must have that transaction. That transaction has to be very good. If I’m interested in the temper of a pipeline, then I’m really interested in how that is happening all the time, just from a general point of view. But I’m really interested if it exceeds a certain threshold.
Alexis
Right.
Data Dave
So, you’ve got threshold based analysis and controls on that. And I’ve got how is it operating in conjunction with the other equipment in its system.
So, often there’s the data management side of it, the data architecture side of it. You’re dealing with, as I said, the streaming data and the time series data, and you’re trying to roll that up and get summaries of that in really quick times for a period of time.
Alexis
Got it.
Data Dave
And then you’re trying to see the trends and the behaviors, that to a point where you want to alert or the point where you think it should be a decision so people will monitor it. But more and more we’re getting to automated monitoring alerting rather than people doing the monitoring.
So, at the base level it becomes very different because it’s a different structure of data as that feeds up. Then it becomes very similar because you’re starting to say cost maintenance, maintenance visit, a service visit, a thing. It leads to something and some event, something happening.
Alexis
Got it. So, at the base level we’re talking about that data management that we’ve kind of been talking about throughout the entire episode. But as we head up the chain with the roll-up data, if you will, it starts to look a lot more similar to the customer and consumer data that we often talk about on the Podcast.
Data Dave
Exactly.
Alexis
Got it.
Data Dave
You might be interested in any one point in time. It depends on the use case. Another example is dynamic speed limits. Have you heard of dynamic speed limits?
Alexis
Ummm, tell me about it.
Data Dave
Based off congestion levels, they can reset the speed limits.
Alexis
Oh, okay. No, I live in a very small town, Dave. Those things don’t happen here.
Data Dave
If you’ve got high congestion. Traffic moves faster if you slow it down.
Based off of the sensors on traffic lights and in the roads, they can tell you how much traffic is passing through and what the average movement of the traffic is. And they will actually adjust the speed limits to get more people through. That’s an automated system that people don’t necessarily monitor it. It now automatically changes the speed limit based off the level of congestion that it detects. That’s again, sensors operating in decision-making capability to return a speed limit. Yes, that’s the idea behind that. But how long is that data relevant for?
Alexis
I would imagine not very. Because the traffic changes.
Data Dave
Yeah, exactly.
Alexis
That’s the purpose of slowing down the limit speed.
Data Dave
From a point of history point of view, I might want to know trending that over the last 10 years the level of traffic has increased in this area by X. But I don’t need to know that to the second.
I do to make the speed limit decision. But I don’t need to know that from a historical perspective.
Alexis
Right.
Data Dave
So, I can throw a lot of the data away, that is the raw data, but I keep the roll up and the decision that says I set the speed limit to this on that day in history.
Alexis
I like this.
This was a very, very informative episode for me. Again, I really didn’t know what Internet of Things was before we joined this episode. So I’m really happy that I learned that. And to talk about it in comparison to consumer data like we often talk about was really helpful for me. That was a good way for me to kind of picture it and put it into perspective. So that was helpful.
Data Dave
Good, I’m glad I helped.
Alexis
All right, well, I like that, Dave. Anything else before we go then?
Data Dave
No, just in summary, It’s really the interconnectedness of things. It’s things that can tell you stuff. Equipment, machines, cars, phones, telling you things that pertains to modern life. And it’s growing. Right? More and more stuff tells us things every day.
Alexis
Yeah.
Data Dave
Every day you get something else that is prepared to tell you something that you didn’t want to know.
Alexis
Well, Dave, this has been great.
Listeners out there. If you have a question for Data Dave, always reach out to us at talktech@d3clarity.com or you can message Dave or myself right on LinkedIn.
Alexis
And Dave, it’s been wonderful. I hope you have a fabulous weekend.
Data Dave
Yes. And you. Thank you very much. Have a great weekend.
