Ryan Chacon is joined by the Director of IoT And Emerging Technologies at InfluxData, Brian Gilmore, to discuss time series data. Brian introduces himself, InfluxData, and what time series data is. He then talks about how it compares to other data and its unique value and benefits. Brian then connects it to the real world by telling us how customers engage with InfluxData’s product and use cases with which time series data works well. Ryan and Brian then move into a high-level conversation around challenges in the IoT space and advice for companies trying to recognize where they need to improve.
About Brian
Brian Gilmore is Director of IoT and Emerging Technology at InfluxData, the creators of InfluxDB. He has focused the last decade of his career on working with organizations worldwide to drive the unification of industrial and enterprise IoT with machine learning, cloud, and other truly transformational technology trends.
Interested in connecting with Brian? Reach out on Linkedin!
About InfluxData
InfluxData is the creator of InfluxDB, the leading time series platform. They empower developers and organizations like Cisco, IBM, Siemens, and Tesla to build real-time IoT, analytics, and cloud applications with time-stamped data. Their technology is purpose-built to handle the massive volumes of data produced by sensors, systems, or applications that change over time. Easy to start and scale, InfluxDB gives developers time to focus on the features and functionalities that give their apps a competitive edge.
Key Questions and Topics from this Episode:
(01:26) Introduction to Brian and InfluxData
(04:20) How does time series data compare to other data
(06:32) Value and benefits of time series data
(07:33) Use cases that benefit from this data
(09:45) How does this data fit in with edge computing
(14:26) How customers engage with their product
(17:55) Biggest challenges in the IoT space
(20:38) How can companies identify their needs
(23:04) Current state of integration and interoperability
Transcript:
– [Voice Over] You are listening to the IoT For All Media Network.
– Hello, everyone, and welcome to another episode of the IoT For All Podcast, the number one publication and resource for the internet of things. I’m your host, Ryan Chacon. If you are watching this on YouTube, we truly appreciate if you give this video a thumbs up, as well as subscribe to our channel, if you haven’t already done so. If you’re listening to us on a podcast directory somewhere else, please feel free to subscribe to get the latest episodes as soon as they are out. All right, on today’s episode, we have Brian Gilmore, the Director of IoT and Emerging Technologies at InfluxData. They are the creator of InfluxDB, the leading time series platform, and they help companies, and developers, organizations, and such, build real-time IoT analytics and cloud applications with timestamp data. So we talk a lot about what time series data is, why it’s well suited for IoT, the value and the benefits it provides, kind of how edge and cloud are being developed in the space, the role they’re playing, as well as challenges they’re seeing in the space from their perspective. So all in all, fantastic conversation. Brian is a great guest, and I think you’ll get a lot of value out of it. But before we get into this. If any of you out there are looking to enter the fast-growing and profitable IoT market, but don’t know where to start, check out our sponsor, Leverege. Leverege’s IoT solutions development platform provides everything you need to create turnkey IoT products that you can white label and resell under your own brand. To learn more, go to iotchangeseverything.com. That’s iotchangeseverything.com, and without further ado, please enjoy this episode of the IoT For All Podcast. Welcome Brian to the IoT For All Podcast. Thanks for being here this week.
– [Brian} Thanks, Ryan, glad to be here.
– [Ryan] Absolutely. Lots of good things we’re gonna talk about today. I’m very excited about this conversation, but I wanted to kick it off by having you give an introduction about yourself and the company. Maybe start with yourself, just background information about how you got into the space, anything you think would be relevant for our audience.
– [Brian] Sure, yeah, well, it was A pretty long strange trip for me to get to the tech business, believe it or not. I actually, I started, well, I guess I started sort of in tech business, I worked for a hospital, I was a data analyst, that was pretty cool. But then I spent a little bit of time in public aquariums, believe it or not, I worked on some jellyfish stuff, and some sea horses. Yeah, yeah, that was fun. Those systems are actually quite technical, and sort of the integration between them and the computers was the part that I found most interesting, and so I built a system. I worked at the Georgia Aquarium in Atlanta, which was the largest aquarium in the world. I built a system down there that just integrated data from the digital life support systems and made that information and data available to people in the lab, people in life support, people on the husbandry team, and I, and the mechanical contractor who built it, thought that, wow, that could be actually a business that other people would use, and so I joined the mechanical contractor. We built an application that was used to analyze smart buildings. We did a really successful project in the US Air Force, and I platformed the final version of that on a piece of technology from a company called Splunk. And I was lucky to join Splunk, and work for them for eight years, and sort of stepped away from Splunk as the company had grown and was doing a lot of the stuff that I always knew they could, and decided to join InfluxData, and, you know, I’m super happy I have. It’s awesome technology. It does exactly what our customers need it to do so.
– [Ryan] Yeah, so let’s expand on that for a second. Tell us a little bit more about what InfluxData does and also the role you all play in the IoT space.
– [Brian] Sure, yeah, I mean, I think people may not have heard of InfluxData, but they probably have heard of InfluxDB. That’s sort of our flagship product. It’s an open source time series database. We also have converted it into an enterprise application that, you know, a organization can purchase with support and scalability and all the enterprise integrations they would need. And we just, in the last two years, have released it as a SaaS. So you can actually go to InfluxData, sign up and, and you get a time series database in the cloud. You know, time series’s data is anything that’s timestamped. It’s pretty easy to understand, I think, you know. People think of sensor data, they think of application instrumentation, all of that. Our business is to make it consumable at massive scale, and then to allow stakeholders in the organization to actually do something with that data, whether it’s just basic monitoring, or if it’s some sort of predictive or AI-driven thing.
– [Ryan] So, you know, when we’re talking about time series data, it’s actually a topic we haven’t covered that often. I mean, from a simple sense, like you just said, it’s anything that’s time stamped. How does that compare to other types of data that is collected through, let’s say smart sensors in the IoT space, just to kind of give our audience an understanding of time series data versus, you know, non-time series data, what that is, and how they both kind of fit or potentially play a role in the space.
– [Brian] Yeah, sure, I mean, I think when you’re looking specifically at sensors, all IoT data is time series data, but only after it is sort of passed through that real time domain, right? So you have a sensor that’s spitting out a value and you know, what is it now? What is it now? What is it now? If you just apply a timestamp to each of those what is it now? And you put those in a database like InfluxDB, you then have that sort of recording of what the values have been over time. And this isn’t brand new. I mean, this is something that process historians have been doing for decades. I think the thing that we’ve done differently is that we have broadened greatly the sort of type of data that can be instrumented and converted into time series data, everything from the IT space and the IoT space. You know, I think, time series data, as compared to say like relational data, you know, charting relational database tables, you know, as a backup mechanism or as a way to understand changes over time can be timestamp data. So I think the easiest way to think of time series data is that it is the sort of historical record of the situation or state of other information, other data sources, right? So it can be documents. It can be, you know, a movie is basically time series pictures, if you think of that way, yeah.
– [Ryan] Absolutely. Yeah. I appreciate you kind of breaking that down. When people hear of it, and they’re not necessarily technical, it seems more complicated than the general concept actually is, just being able to time stamp data for you to be able to understand what is happening, when it’s happening, and how that fits into pretty much every IoT application and solution, like you’re saying, makes a ton of sense.
– [Brian] Exactly.
– [Ryan] And are there any, like, I guess, can you break down a little bit further the value and the benefits time series data has for use cases in IoT?
– [Brian] Yeah, for sure. I mean, I think like if you’re just thinking about simple monitoring and diagnostics, right? I mean, real time visibility of course is huge in IoT, but like, you always need to put the perspective of what you’re seeing the asset or the application, the sensor doing right now, in terms of what was it doing 15 minutes ago? What was it doing at exactly this time five days ago or at exactly this time, the last time, you know, it was 70 degrees outside or whatever. So, you know, I think of the time stamp and then sort of all of those ancillary derivatives of the timestamp, you know, the hour of the day, the day of the week, the week of the month, the weekday, all that kind of stuff as the primary key, you know, in terms of the way people would wanna recall or query or aggregate information.
– [Ryan] Right. And are there any use cases that, I guess, have seen even more of a benefit from the time series data, and maybe even your product on the market as well, that you kind of have a focus on, and I guess one other question I have, kind of loosely related to that is, is how you kind of fit into the IoT solution an end-to-end IoT solution, and kind of how you partner and work with other organizations to bring something to market, not just the product itself on the SaaS side, but also like playing a role in the full scale or a full scale solution out there trying to solve a problem?
– [Brian] Sure, sure. Yeah, so in terms of use cases, I think, like the most, like I would say the holy grail of use cases, especially around IoT, is like predictive or autonomous maintenance. So you know when you’re thinking about managing an asset in the field, you know, sort of collecting the data about the sort of historical activities behavior of that asset is important. But if you then also, of course, have, you know, a time-stamped occurrence of things like anomalies or actual failures where you know this is where it broke a belt or a bearing or whatever the sort of break might be. You can use that time series information to keep that happening again, you know, keep it from happening again in the future. I think the other thing that people are really starting to do now is they’re starting to take that time series information in large volume, and not just from like one sensor, but from a large number of sensors that represent the application of the asset, or, you know, even some of the actual digital data, the process steps, things like that. And if you know it failed, you can feed in, say, the last hour, or the last day before into a machine learning algorithm that will actually take that data and sort of come up with a pattern where you can then continuously process new information from that asset, match it to that pattern, and then when it does match to a certain confidence, you can alert your maintenance teams that, hey, you know, based on the time series information we’re seeing from this asset, we believe that it may break in the same way its peer broke, you know, last week or whatever it was.
– [Ryan] Yeah. And how does this all fit in? And I guess, maybe, so I get a little more technical, but how does the time series data fit in with the kind of growth of edge computing in IoT and the cloud side as well? Like where, with predictive maintenance stuff, obviously, and I guess with predictive maintenance, you’re looking at trying to make decisions relatively quickly before things break, where some other use cases just having the historical data timestamped is valuable to be able to go back, but you’re not necessarily having to make decisions in real time based on that data. So how does the edge and the cloud kind of play a role in making time series data more valuable, maybe tied to a particular use case, or just in a general sense? Because I feel like those two pieces have been being developed pretty heavily, as it relates to IoT, and are becoming more and more valuable with every new use case and solution that’s kind of put out on the market.
– [Brian] Yeah, that’s a great question. I think I can kind of maybe pull in a response to the second half of your question as well as that, like you know, an IoT application, it’s an ecosystem, it’s a snap, right? So you’ve got, you know, your cloud service providers who are providing that like real horsepower of, you know, the sort of parallelized compute, if you need that, or the, you know, elastically scalable microservices, if you need that, the serverless computing, if you need that in the cloud. Now all of those pieces are, of course, very important to a use case like predictive maintenance or to, you know, a remote monitoring use case or whatever it might be. On the other hand, you know, part of the digitization or the digitalization of assets and, you know, products, devices, in the scope of IoT is the fact that they also themselves have horsepower. They have compute, they have storage locally, right? And I think what people are trying and starting to figure out is that it’s not just doing at the edge what needs to happen at the edge, and doing in the cloud what needs to happen in the cloud, and doing what needs to happen in between in between, it’s how do you stack all of that capability and all of that horsepower from end to end into something that serves the solution, not only from a capabilities perspective, but also an architecture perspective, right? So like, you know, if you’re looking at like a traditional manufacturing floor, you know, the automation and controls of those assets are not something that a company like AWS or, you know, InfluxData is gonna really be too concerned with, right? That’s in the scope and the domain of the PLCs and the RTUs and the automation control systems. Now, that information, of course, is super valuable to the teams managing those assets to do things like predictive maintenance. So what, you know, we can do is we could use, like there’s a lot of gateway software out there, stuff, you know, from the Keplars of this world or the, you know, automation or the HighBytes, and they, you know, sort of specialize in downstream connectivity to those proprietary data sources at the edge, kind of sit at that gateway layer, and then they can pass data to, you know, InfluxDB, or really, you know, a lot of other services as well nearby the edge, kind of you think like if you’re familiar with CloudFront sort of as being like that ingress layer to the cloud there’s a gateway layer at the edge, which if you wanna call it edge front, I don’t know if anybody’s doing that yet, but that’s, you know, where you would sit a piece of our technology because, you know, you want the sort of short round trip analytics. They’re down at the edge where the data is born, because data has a half life. Like the older it gets, especially in IoT, the less valuable it becomes, right? Especially when you’re talking about doing things in any form of real time. So getting a database down close to the edge, processing that data, in, you know, near real time with nanosecond precision is now possible with the translation being done by something like HighByte or if it’s an IT system, we can do it with a open source tool like Telegraph, and we can process that data there. Then the question becomes, how do you strategically take the output of that edge side analysis data processing, and strategically move that data to the cloud where you can apply cloud-specific horsepower and cloud-specific capabilities, like, you know, global visibility through mobile apps, or like integrating, you know, machine learning services for many of the, you know, the big three CSPs and putting it together like that as an architecture and showing customers how they can do all that is really like, that is the key to finally people going, oh, wow, okay, like I can see how this actually would work in my business and then the value it would deliver to my business.
– [Ryan] Gotcha, and can you elaborate a bit on when you’re working, like, I guess, how do customers, or how do your customers kind of engage with your product? Is it something that like a systems integrator or somebody putting together a solution comes in to bring in your tool, your technology, into the solution themselves? Is that something that you’re working hand in hand with them to like, from a partnership standpoint, to help deploy? Like what does that typical engagement look like from your end?
– [Brian] Yeah, I mean, I don’t think there is a typical engagement yet. So I think, I mean, our product, which I think it is a little bit unique in this way is that it’s available through self-service and it’s also available through delivered services. You know, we’re also included, embedded, OEM’ed by some of the largest industrial IoT platforms out there, so we’re always sort of there, either because somebody on the factory floor has downloaded our free open source software, and has just integrated it because it did the trick. But then we’re also starting to see these solutions providers and these solutions builders who are building end-to-end solutions like I just described, whereas they’re bringing our open source or they’re bringing our cloud to those opportunities. And, you know, rather than having to reinvent the wheel of a time series database, or having to bring in like the, you know, the expense and the overhead of working with one of the big industrial process historians, they can just very quickly prototype and use our product and we’re happy to support them there. Now, you know, we have a commercial model. You know, we’re a company. Generally, our goal is to get people using and seeing value from the software first, and then, you know, we can worry about the, you know, the commercials later, but oftentimes, honestly, oftentimes, we have companies who are out there building solutions on our free open source software and, you know, collecting all of the sort of services money themselves, and we’re happy to see. We just wanna share in the outcomes. We wanna be able to tell those stories, you know, success is…
– [Ryan] I totally agree. And that’s kind of a very similar approach that a lot of companies in the space take. It’s not about owning all the different pieces, putting the right pieces together to bring the best possible solution to market to fit all the needs of potential end user, customer, whether it’s cost, or other ROI kind of factors that matter in order to grow and scale the usage of it. And that’s, I think, been the biggest hurdle for companies, or IoT in general, as an industry, which I think is becoming more and more, or is being solved more and more every day as companies are taking more of that partnership approach, as opposed to just trying to do it all themselves.
– [Brian] Yeah, I mean, I couldn’t agree more. I think if you look back 15 or 20 years, isolation was the primary method of competitive protection. And now it’s exactly the opposite.
– [Ryan] 100%.
– [Brian] Openness is the, you know, I mean, if you are open and you can play into ecosystems and you build the relationships with other complimentary companies and products, and customers see that and they appreciate it, and they know, you know, it’s sort of an abundance mentality these days in that. There’s so much to be done on these fronts that like why anybody would try to, you know, sort of wall off customers in a way that didn’t work for them. It doesn’t make sense to me, and I know a lot of other people in the field who feel very similarly on that.
– [Ryan] Absolutely. This is a good opportunity to kind of transition to a question I had from your perspective, which I think is a pretty unique perspective to come in and kind of view the market and how you work with other companies, where you see challenges. What are some of the biggest challenges that you all have seen in the IoT space as you’ve interacted with more end users, more partners, more, you know, buyers, things like that, just out of curiosity, kind of where do you see the biggest challenges that the market’s facing right now?
– [Brian] Yeah, I mean, I think it’s still, it’s less technology problems and more people problems, you know? And I think like, to sort of the last point, you know, we have to overcome quite a bit the sort of like ill doing of the organizations of the past, right? Because everybody sort of assumes when a salesperson enters the room that, you know, oh God, here we go. And it really isn’t that way anymore. So once people realize that, you know, folks from companies like InfluxData are really there to like share in value and help solve problems, you know, then we oftentimes see, okay, well we have budget for IoT. What do we do? And it’s like, okay, well, that’s not where you start, first of all. Like there’s an operational sort of strategy, I think, that’s missing in a lot of IoT products, or projects in that there really is no problem to solve. It’s like, hey, we’ve got a tool and we gotta do something with it. And those types of projects, in my experience, are almost always gonna end up sort of in this, like try, fail, try, you know, re-loop of, you know, pilot purgatory, as some people call it. You know, and I think getting away from that, and you know, what I oftentimes will encourage customers to do is to like get key stakeholders around the table and ask them like what their biggest challenge is? What keeps them up at night? And then work with, you know, vendor teams, systems integrators, consultants, whoever it needs to be to sort of take a look at those problems and identify components of those problems that can be solved with technology. And those solutions are gonna include IoT. They’re gonna include time series databases. They’re gonna sometimes include AI or ML or augmented reality or whatever they are, but then you can show that, okay, to solve this problem, here is the set of means to that end. And I think far too many people just consider the technology itself is the end. Well, you know, we signed the PO, we installed the software, we’re done. And that doesn’t work. I can’t believe it worked anywhere else, but it certainly doesn’t work in IoT so.
– [Ryan] And how do you advise companies who are trying to really identify real pain points that need to be solved, which then they can kind of go down the path of figuring out what technology, what solutions are right for them? How can companies kind of get started identifying that and really be thinking about that part of the process?
– [Brian] Yeah, I mean, I think it’s a strategy. I think, you know, I’m a sort of, I hate the term, but like bottoms-up type believer in that, like, I think if you were to walk the factory floor, you were to go out and walk the oil rig, or, you know, pull in a large number of doers in your organization, they are extremely technically savvy. They know exactly what is available for like information instrumentation, and they know all of those obstacles that are keeping them from getting the insights they need there. So they’re informed and they’re motivated, and those are the best people to have at the table. Now I understand, like if you’re an organization with 100,000 employees, that’s not practical. You’re gonna have to like have folks representing those people. But I think like actually getting the people who do the doing around the table is totally key. And then just asking them the questions. Don’t ask them how would they use Technology A or Technology B, right? Again, like get to the bottom of like what their biggest issue is right now. And if it’s an operational issue, like I said, it’s easy to sort of translate technology into that. If it’s a technology problem, which oftentimes it’ll be with those doers, you know, it’s just something you need to fix. And you’re either fixing it through process or people, or literally, you know, changing technology vendors. And I think that will sort of deliver that payoff, that return on investment that’s gonna, suddenly, people will be coming to you as the CDO, or whoever’s responsible for managing these digital transformation, these IoT projects in a company. People will come to you ’cause they’ll know like, hey, you’re like a fixer. Like, you know, I heard Joe in Ops came over and he said that he really needed to start seeing real time visibility into X, Y, Z, or getting these reports, and you figured it out. And those are the use cases, again, that power the adoption of IoT, or of industrial IoT, you know, and it’s just a lot of like asking questions of the people that need to do the work, I guess, is my most common recommendation.
– [Ryan] Totally agree. Last thing I wanted to ask you before we wrap up here is, as it relates to interoperability and integration, that’s always been a big topic in bringing any type of new technology solution into a business. How have you kind of seen the current state of the market on that side of things, as it relates to the use cases and solutions you’ve been a part of, because obviously that’s a piece, right? The integration of your technology or your offering into an existing potential infrastructure, potentially existing solution already. And I feel like that has always been a thing that’s kind of hamstrung companies in the adoption of IoT. How have you kind of seen that evolve lately and kind of where we are with that?
– [Brian] Yeah, I mean, I think the technology has, you know, improved the situation quite a bit, at least over the last, I would say, two or three years. You’re starting to see the emergence of a role of data operations or data ops. And I think, along with that role, you start to see new technologies pop up that are much more about the pipelining of data in terms of, you know, capturing it from proprietary protocols or, you know, from other third party sources, databases or whatever, and then, you know, transforming that data into a domain-specific model and then sharing that data broadly through what is an ever narrowing set of interfaces. Like I think, you know, back when you had 45 different standards and protocols, it was a little bit harder, but now, you know, all these companies are really focused on like normalizing everything up to HTTP or to MPTT, or even just putting it like in a text file or something like that. And because the data is sort of, you know, it’s pre-formatted enriched for specific uses, then it’s really easy for applications like InfluxDB just to tap into those sort of sources and you know, siphon off what we need to siphon and put it in the hands of, you know, the very capable operators who need to monitor or troubleshoot or build KPIs, do predictive maintenance, those types of things.
– [Ryan] Absolutely. First off, I really wanna thank you for your time today in talking about a topic we do not cover very often. I think it’s gonna be a very informative and valuable piece of information, piece of content that our audience is gonna really love. For our audience, though, who is listening and watching this and wants to follow up with questions, wants to learn more about what you all do, how this may fit into kind of maybe their solution, their technology and what they have going on, what’s the best way they can do that?
– [Brian] Sure, I mean, just to learn more about the technology, I would just go to influxdata.com. That’s sort of our website. If you are interested in the open source projects that power all of our technology, you can find us on GitHub, we’ve got 23,000 stars. It’s not hard to do. I would also recommend folks who wanna actually ask questions of our community and of our experts at InfluxData to go to our community Slack channel. It’s very easily found at InfluxData. It’s the InfluxData community. And then, yeah, just join up, do some reading, do some learning, download our open source, run the Docker container, whatever you wanna do, sign up for a free cloud account, and we look forward to working with you.
– [Ryan] Fantastic, well, Brian, thanks again so much for your time, really appreciate it. And hopefully can have you and other members of the team back for some other video content that we’re building. I think what you have is going on is absolutely fantastic and a very good fit for a lot of the listeners out there to adopt and understand how it really fits into providing more value for their solutions. So thanks again.
– [Brian] Yeah, thank you, Ryan, it was a pleasure.
– [Ryan] All right, everyone, thanks again for watching that episode of the IoT for All Podcast. If you enjoyed the episode, please click the thumbs up button, subscribe to our channel, and be sure to hit the bell notifications so you get the latest episodes as soon as they become available. Other than that, thanks again for watching, and we’ll see you next time.
Industry News
