IoT Thought Leaders – Interview with Ido Sarig of Wind River: B2B Market Research podcast

Modified Transcript:

[Sean]:

Welcome to another episode of the Competitive Intel Podcast. In this episode, I have with me Ido Sarig from Wind River, and we’ll be talking to him about the Internet of Things (IoT).

Before we get into that though a few brief programming notes. First, Cascade Insights is a competitive intelligence firm focused squarely and exclusively on the B2B technology sector..

If you’d like to find past episodes of this podcast or a pretty big array of free competitive intelligence resources, you can do that at www.cascadeinsights.com/resources and for the podcast in particular, you can also find it on places like iTunes and Stitcher Radio. With that, let’s go ahead and let Ido introduce himself.

[Ido]:

Sure, Sean. I’m currently the general manager of Wind River’s Internet of Things Solutions Group. Prior to that I was a venture investor as a partner with Thomas Weisel Venture Partners where I invested primarily in enterprise software and before that I spent the bulk of my professional career with a company called Mercury Interactive that practically invented automated software testing and was eventually acquired by HP.

[Sean]:

One of the reasons I reached out to you is that obviously in this day and age many companies and individuals are starting to attach themselves to the Internet of Things. But, you look like as if you’ve had actual practical experience with it, which is one of the reasons I wanted to talk with you about it.

We are going to stay away from some of the broader things that we could all just go find in Google – like how big is the Internet of Things going to be.

I really want to dig into things that I think you can provide some insight on and are which are maybe some of the questions that don’t get talked about quite as much. The first one is, there is so much emphasis when you look at your average Internet of Things story on wearables and B2C. You could almost say they’re synonymous in 80% of the articles, but B2B is where IoT has been already existed in many ways with machine to machine (M2M) and things like that.

M2M is not what the IoT is going to be, but I imagine you can talk a little bit about the history of what we maybe used to call M2M initiatives in B2B and what will in the future call the IoT and where you think that’s all going to meet up in a B2B landscape.

[Ido]:

Sure. First of all, I want to thank you for not delving into the various things that you can Google on. It’s been a pet peeve of mine that just about every IoT presentation that you go to these days opens up with the obligatory three slides that say there’s going to 20 billion devices connected by 2015, and there’s going to be 50 petabytes of data produced every hour, and all that is, a: repetitive, and, b: not really interesting because it lacks the business context.

The second thing is that you’re right to carve out the difference between what most people think about when they’re thinking Internet of Things which tend to be consumer-type products like wearables or maybe a Nest thermometer in the home and the commercial or industrial internet of things which has been around for many years.

In fact, when I talk to some of our existing customers and our prospects, some of them are actually a little upset that the Internet of Things is being positioned as something new because they’ve been doing machine to machine for many, many years. They’ve just been calling it something else.

The reality is that what was called machine to machine and is now rapidly being transformed into IoT is not exactly the Internet of Things because those machine to machine interactions in many cases were using proprietary protocols, and the machines to the extent that they were connected, were connected to some proprietary network, and really …

[Sean]:

Well, let’s stop there just for a sec because that’s one of the key things to understand. On one hand there’s the hype cycle around the IoT to your point, and it’s funny because you and I ended up interjecting some of the same statistics, but at least we didn’t spend a lot of time on it. There’s that but there’s also this issue of, like you said, it’s IoT is not exactly M2M like I was saying earlier. It’s not exactly the same thing.

What you’re getting at is that there were these proprietary silos, things were not standardized, and not all these devices were connected to what a normal person would think of as the Internet. The issue then becomes how that’s going to evolve, so talk a little bit about that because that’s exactly where you were going.

[Ido]:

Sure so here’s one of the interesting ways in which it’s going to evolve. When you were doing M2M with proprietary protocols on your proprietary network, there tended to be much less of an emphasis on security because after all, you’re talking about your machines that only you have physical access to. They are inter-gapped from the outside world. I’m not suggesting that there were no security capabilities in place, but to the extent that they existed, they tended to be very rudimentary.

Now, when you’re connecting these things to the Internet, you’re exposing them to all the network-born attacks that we all, I don’t want to say know and love, but certainly know. All of a sudden this CT scanner that used to be nearly immune from attacks originating from outside the hospital, can now just get any old malware, any virus that happens to target the operating system that it’s running.

A lot of emphasis is going to be placed on making these systems much more secure and in fact they have to be secure to a level that today is probably nonexistent because it’s one thing if your personal credit card information is hacked off of a Target data base that, at worst, causes some inconvenience for you in having to get a new credit card number and some headaches of getting your personal finances in order. It’s quite a different thing if what is being hacked is the POC that’s controlling your local power utilities nuclear plant and exposing it to a potential explosion.

[Sean]:

That takes us to a question that people in the B2B side of this problem have questions about is where are we at with standardization. When you think about it that’s always the problem. There was a good article in the Harvard Business Review just recently that was talking about RFID standards versus IoT standards, and how it took forever to get standards on RFID which was driven by big organizations like WalMart and whatever using the technology.

The authors said, “We can’t take 15 years to get there this time.” The implication was these devices are going to quickly become connected regardless, because that’s really where we’re headed, so we better have the standards and the protocols, and the security measures in place. Otherwise, we’re going to create a pretty big problem for ourselves at some point or another.

[Ido]:

That’s true. The reality is today we’re pretty far from being close to standardization. I was speaking just last week at a conference here in San Jose Digital Deception on a panel about the importance of standards and open source to the development of IoT. Unfortunately, the consensus among panel members is that if anything right now, we have too many competing standards there. You’ve got the AllSeen Alliance trying to promote one standard. There’s the OIC Consortium trying to promote a different standard. I don’t think that we’re anywhere near achieving standardization today for IoT.

[Sean]:

That makes complete sense given what I’ve seen as well in terms of the market. A different question would be..what problems do you think the IoT is going to cause. When you think about their being many sensors out there, many actuators, because that’s the other potential challenge.

Not everything that is a sensor is going to actually be able to actively do something on it’s own as an IoT device, but there’s going to be some of that as well. You’ve got this many to many problem in essence. Probably the ex-database developer in me that’s got me thinking that way.

Then on the other side, you have many analytic systems, many different types of storage systems. Many different ways of interpreting the data. It seems like that’s going to be another pretty big problem because even with the complexity of your average kind enterprise IT stack today, it doesn’t really approach the complexity that the IoT is going to give us in terms of the amount of data we have to analyze and how we’re going to basically represent that analysis to an organization, so they can actually do something with it.

[Ido]:

That’s really true and I can only give you our approach, our approach being both Wind River as well as our parent company, Intel, to this problem. On the device side of things, I agree there is huge diversity amongst these devices. Many of them especially the lower end sensors and actuators won’t run a full-fledged operating system, won’t have a complete IT stack, and they just won’t be able to connect to the Internet directly. They will require some sort of intermediary, and towards that end the first product that we came out to market with in the IoT space is an IoT gateway.

One of the components of that IoT gateway is a translation layer that will enable the diverse devices to talk whatever protocol they want within reason and have us standardize the communication from the gateway up into the cloud addressing the diversity and lack of standards that way.

On the cloud side, we’ve taken a somewhat similar approach of not trying to solve all the world’s problem but to bring the data in to a certain repository and make it accessible through a public open RESTful API that will enable different customers with different analytical engines to access that data and do whatever they want with it, we’ll just provide the ease of access via this RESTful API.

[Sean]:

That makes sense and Intel obviously has a pretty big history of working with standards, various groups, and doing a lot of work in that regard. You guys obviously are related organizationally, and that makes a ton of sense that you would have a lot of play there. Let me ask you this as well, where do you think there is going to be the fiercest competition and/or consolidation?

Said a little differently, what are the companies that we might be talking about today when it comes to the IoT, but in a few years we won’t ever remember them enough to bring them up?

[Ido]:

Sure and I want to separate this from our view of who the competition is because that delves a little bit into information that I’m not at liberty to discuss …

[Sean]:

Sure. Completely understand.

[Ido]:

At the high level, traditional industries should be concerned and worried about people they haven’t even thought about. I’m sure that five years ago, certainly ten years ago, Honeywell didn’t think that Google was going to be one of their main competitors. If you look at what Google is doing now with Nest, I’m pretty sure that the folks at Honeywell are scared and are right to be scared just because if Google has come in this reach, and this acquisition of Nest has put them squarely as I would say probably the number one competitor of Honeywell and people like Honeywell if it’s Johnsons Control or Rockwell Automation.

They should all be looking not just at Google but at who might be the next Google. Who might be somebody with enough of a footprint in the Internet, although they might not have a single thing to do with their industrial controls business, but just in terms of market reach and the ability to make an acquisition like Google did with Nest. In other domains, we see it. There are many traditional industries that if they’re not careful, they’re going to be disrupted in a very similar way. Agriculture comes to mind and the various manufacturers of agricultural machinery, one of the things that we’ve been seeing going on there.

Agriculture, by the way, is one of the most active areas for Internet of Things on the industrial side with sensors going in everywhere into fields to measure things like the level of moisture in the soil, the level of fertilizer in the soil combining that with historical weather data patterns to predict what the crop yields will be next year and how to fertilize or irrigate in preparation for that accordingly.

One of the interesting things that’s happening over there is that companies that traditionally have had access to a lot of data, and those would tend to be insurance companies, are becoming prime targets for acquisition by people operating in this space. Think of Monsanto as a manufacturer or distributor of seeds. The most important thing to them is not probably the specific sensor or actuator that goes into an agricultural field or into a tractor, but in being able to reach and grab the wealth of historical data that insurance companies have about weather patterns to make use of that in their products.

[Sean]:

Yep, that makes a ton of sense particularly in areas like precision agriculture and things like that, that dynamic you’re talking about where you have an interesting menagerie of players working together. I even had a conversation not that long ago with somebody in a particular space that was more manufacturing oriented, as they were talking about how they’re looking toward the big data analytics players and trying to understand how they can relate to them. And this is an industry segment that you probably wouldn’t have expected them to say “big data” in a sentence until recently.

So when you look at it from that angle, there’s a lot of merging and cohabitating between players that you wouldn’t have normally expected which is really what you’re getting at, and that’s going to lead to some interesting disruptions as two players who you wouldn’t have expected to work together now are going to. That maybe leaves the third one a little bit out in the cold.

[Ido]:

Absolutely. Would you have expected a couple of years ago to see maybe a John Deere as a natural partner for somebody like SAS with their statistical analysis capabilities? Probably not, but like you point out, big data is a big part of the analytics that are putting to play because machines just generate so much machine data at rates that humans have never been able to produce in the past.

All the back end IT-type systems that were very good at analyzing data now become very interesting for players in the OT space producing machine data and wanting access to those analytical capabilities to make sense and put the data in the right business context. This is part of a convergence that we’re seeing between IT and what was more of operational technologies in the past.

[Sean]:

Let me dive into that a little bit, because my next question was oddly enough going to be how do you see the IoT impacting enterprise IT? When you think about storage, analytics, archival, you already brushed over the tops of a few of those stacks, so maybe dive into it a little bit in terms of the areas that when you think about enterprise IT, you’re like, “In the next X number of months or years, this is what’s going to be most impacted by the rise of IoT.”

[Ido]:

What’s going to happen is that IT will need to be a lot more flexible towards working with their OT counterparts. Many of the organizations that I’m talking to right now about IoT projects, IT is viewed as an obstacle, as a hurdle that they have to overcome that they have to convince these IT guys to allow their systems to connect into the production facilities and vice versa, and it’s not happening fast enough. If IT doesn’t get it quickly, they will soon find their role being overtaken by OT because OT needs these kind of capabilities. If they’re viewed as obstacles, they’ll just go and create their own.

[Sean]:

Yeah, that totally makes sense. Let me close with this question. As you’re obviously studying this space a lot. I don’t ask this question of everybody because it’s more germane if the space is new and evolving. Where do you go to get information about it? Whether that’s particular bloggers or even books that you’ve read on it or websites or portals that seem to have articles that are more of interest rather than less. Where do you go to get educated about it all?

[Ido]:

That’s an interesting question. The primary source is just my prospects and existing customers. They, I find, are very forward looking in their thinking coming up with ideas of how they’re going to leverage IoT far better than what I could imagine or could Google.  On the specific aspect of security, a good book that I read relatively recently was When Gadgets Betray Us by Robert Vamosi. He was formerly with Mocana. He talks about all the potential security pitfalls of connecting devices that will otherwise stand alone to the Internet. It’s a very eye-opening book.

[Sean]:

Cool. Well, that sounds like a great resource.

With that, I just want to thank you for coming on the podcast. To the folks that were listening, again, if you want to find more past episodes, you can do so on iTunes on Stitcher Radio and on our site. Just want to thank y’all for being along with us and hope to have you along in the next podcast.