IoT GUIDE

This is an episode of the "Ventures in Industrial IoT" series brought to you by GE Ventures. In the series we explore success factors and challenges in Industrial IoT markets with CEOs, investors and experts. 

IoT Spotlight Podcast introduction:

The IIoT Spotlight Podcast shines a light on Industrial IoT solutions that are impacting businesses today. Every episode we interview an expert about a specific IoT use case. Our goal is to provide insight into the planning and implementation of IIoT systems, from new business models to technology architecture selection to data ownership and security. The IIoT Spotlight is produced by IoT ONE, an information platform that provides market insight, partner development, and go-to-market support for technology providers, end users, and investors. Don't forget to follow us on twitter. You can contact me directly at erik.walenza@iotone.com.

Summary:

The Industrial IoT ecosystem is a truly dynamic space and a large part of that dynamism can be attributed to the investment and venture capital space that provides the financial impetus to drive innovation in niche fields like edge computing, cybersecurity and blockchain. 

In the second episode of our "Ventures in Industrial IoT" series with GE Ventures, we are pleased to welcome Nicholas Pappageorge, IoT market analyst at CB Insights to the show. Brace yourself for a highly engaging and data-driven conversation as Nicholas shares with us findings about the latest trends impacting the venture capital space surrounding the Industrial IoT market.

Learn more about CB Insights: https://www.cbinsights.com/

Podcast Links: 

Libsyn: http://directory.libsyn.com/episode/index/id/6016713

iTunes:  https://itunes.apple.com/us/podcast/industrial-iot-spotlight/id1228185407?mt=2​ 

Speaker bio: 

Nicholas Pappageorge is the IoT market analyst at CB Insights where he analyzes startups and venture capital. His academic expertise is in the analysis of economic, political, and commercial issues in the Eurozone and Latin America. He has even worked on the staff for the "Obama for America" 2012 election campaign.

Company Overview:

CB Insights has built a tech market intelligence platform that analyzes millions of data points on venture capital, startups, patents, partnerships and news media to predict technology trends. They use machine learning to crawl hundreds of thousands of sites and in order to gain an incisive view on private markets across the board. As such they are able to provide an unparallaled standard of analysis and consulting services to their clients about acquisitions, investments and market trends.  

Blurbs:

 "There's a lot of noise about what's going on and I think still indicative that IoT hasn't quite arrived yet. Especially in the industrial world."

"Improving a production line by one percent efficiency is just an enormously valuable business problem to solve given that there's really low digitisation in a lot of industries." 

"I'm also still looking for a more "uber-fication", more on-demand companies that I think are really interesting. Lately I've started to see more financings going on to Robotics-as-a-Service."

"In the short term the two big trends I've been seeing are new cyber security layers developed and edge intelligence being done at the device level." 

Show Transcription: 

 

 Welcome back to the Industrial IoT Spotlight.

I'm joined today by Nicholas Pappageorge from CB Insights. Nicholas is the technology industry analyst focusing on IoT among other fields at CB Insights. Nicholas, before we dive into our conversation today give a little bit of background to where you know where you're coming from and introduce the work that you're doing at CB and say it's I'm a big fan but some of our listeners might not be familiar.

Sure. Wonderful. Well I'm really happy to be here. First thank you for having us on. And I come from my first job out of college was at MakerBot industries in the 3-D printing world. The last two plus years I've been at CB Insights for those who don't know what that is, CB insights is a tech market intelligence platform. And so we use machine learning to crawl hundreds of thousands of sites and basically see what's going on in private markets across the board. So everything from financings to trends broadly going on. If you're a tech investor in the VC space to  corporate innovation group, you'd use CB Insights as a platform to basically unlock what's going on in private markets. So what that means is you know we track a lot of financings going on into everything from auto tech to industrial IoT where I've been focusing on is the latter. Taking a look at advanced manufacturing and basically just connected hardware in general in IoT.

Gotcha so you're focused on connected hardware. Are there other analysts that are focused on the software side?

How did you come to focus on the space?

Sure. Yes we have. We have analysts focus on everything from fintech, healthcare. Logistics, e-commerce, auto tech and in pretty much, Every category you could think of including some of the horizontal ones like artificial intelligence, blockchain and cyber security, things that are kind of more horizontal and touch all verticals. And we're actually scaling up the research team you know every every day we do this. But how I came around to doing it is you know my background in hardware just seeing the 3-D printing world up close it seemed like a natural fit, you know actually when I started in consumer IoT especially in the wearables world was super hot. And then since then I think. As you know we saw some innocent rough news coming out of the jawbone and a lot of these wrist-worn wearables and the focus, as far as investing and investment trends goes has been more into the industrial side.

Yeah that's been my feeling is that wearables let's say the ideas or IoT and consumer in general, high level the idea seems to be very it presents itself well in social media right.

So connected and so forth. But then when you start thinking about the actual value proposition and in making that value proposition to a mass market as opposed to.

Kind of people that are willing to buy technology because you know for technology's sake so to speak. I think a lot of these solutions. It's actually quite difficult to make a really compelling value proposition out for.

The industrial space.

Often it's a lot more straightforward because the businesses are they've already done to an extent the you know the business mapping they know what their costs are and what the value of solving the problem is and then it's around can. Can these new technology solve problems so make sense that would be transferring. Maybe you can put some numbers behind it because it's you know that's one of the areas that CB Insights excels at. So when you say that manufacturing is the biggest and fastest growing area of IoT. How do you measure that? What's the indicator?

Yes so what we look at first is basically deals going into young companies in the space. So you know broadly in IoT, it's at least eight to nine billion dollars invested in the last five years from 2013. And this is across and at least a thousand deals. In the advanced manufacturing space which encompasses everything from like new composites are not necessarily IoT but also you know connected hardware as well that manufacturing world is around seven billion dollars, in investment spent. So a lot of this is still a very nascent world. So I think you know just backing up about IoT in general I think a lot of people know what it feels like and you know they know it when they see it but don't necessarily know what it means. So when I think about it I think about it is digitization of the physical world in a way that wasn't previously possible. So you know writing a lot of the trends in the smartphone world you know cheaper sensors, cheaper cloud computing you were able to start tracking, you know how many steps you took with your wristband and the same concept is happening in the industrial world. And startups have been doing this in several different ways. So some you know crafting and 3D printing way some are doing this by actually just retrofitting actual machines with sensors to derive insights from them for the first time. And there's many different remixes and ways that's happening but at the early stage though you know numbers wise this is still a nascent category. You know at least 40 to 50 percent of the deals going on in this space are at the early stage. So usually when we see startup categories start to mature the middle stage is the greater chunk of deals happening. But still you know I think this is in many ways you know the lack of security measures and standards in place. As well as that there's a lot of different private silos, there are tons of different cloud vendors. There's a lot of noise about what's going on and I think still indicative that IoT hasn't quite arrived yet. Especially in the industrial world.

Absolutely I was talking to Tim Chou who's a professor over at Stanford and then also you know on the board of Terradata and a few other companies and his perspective was that we're in you know really the early let's say 2,3 fourth year of a 20 year cycle in IoT. 

And so we're very much in the educational. So it makes sense that the investment side as well we'd be looking at at Cedar for A-series. Maybe a bit more of a tactical question. How do you gather this information? I think getting reliable information around seed early you know early investments. Super challenging just from a practical perspective. How do you gather, how do you validate and make sure that the analysis that's coming out is credible?

Yes. So what we do in since inception we've been using machine learning to structure a lot of data from hundreds of thousands of these data sources some of these are actually just startup financing some of these are in reference to that. We structure a lot of the data, it sees some human eyes before being logged in our database. And as we've grown to prominence. CB insights has had some data network effects going on. And by that I mean investors and startups themselves have wanted to edit our database, submit data to us. And you know it's really in everybody's interest to have the transparency essentially as more and more investors and corporate venture arms and innovation groups are looking to CB insights for the right data. So in many ways it's been a little bit of both.

Gotcha. And then you know your data in addition to tracking kind of trends around investment you're also looking at who the investors are. So I think the other side of the equation. In the IoT space what are you seeing? Are we looking more at the kind of C-stage venture capitals or the early stage venture capitals or is this a lot of corporate investment going into early stage more maybe as exploratory investments?

So it's a mix of both but definitely more corporate participation than in a lot of other categories. So most active in industrial IoT broadly usually GE Ventures tops the list and a lot of these industrial categories but you know they are followed behind and not far behind by. Large and active corporate venture arms like Intel Capital, Cisco, Google Ventures. Basically a lot of these are industrials or you know large corporates that stand to gain a lot by strategic investment in the area but also they're joined in by a lot of name brands VCs like. Kleiner Perkins, Andreessen Horowitz, Lux capital, Enterprise Associates and Eclipse ventures. But I would say in the corporate participation world it's pretty surprising. So maybe just five, six years ago in 2012 about 17 percent of the deals in this area involving a corporate venture arm in the deal. But now it's nearly double a little less than doubled to about 31 percent just last year involve a corporate. So. In general this is just seeing more of corporate interests than just five or six years ago.

Interesting this seems like a lot of corporates have made the strategic choice to let's say outsource some of their innovation to startups. I have to say that startups with no particular types of innovation are more efficient and it's better to participate and learn what they're doing follow them invest in them and potentially buy them out. When it when it gets to the commercialization phase. 

Yeah absolutely I think a lot of these industrial conglomerates are really sensing that there is disruption happening but not you know that's like a common startup narrative that everything's getting disrupted I think a lot of these corporates are also seeing value in having new partnerships and conquering new markets either by acquiring or working with.

Startups. It's not necessarily one. Always eating the other either.

If we look at let's say the advanced analytics of you know of this. So if we were looking at a basketball game we could say you could look at the simple analytics kind of the score in the box and so forth and then the advanced analytics may be going a bit a bit deeper trying to explain why.

So if we think a bit harder around what differentiates corporate VC from the traditional VCs What do you see in behavior in terms of how sticky they are to the investment in terms of you know buying out startups are they are they basically testing the waters and then and then buying startups out once they reach a certain stage of commercialization. What kind of insight do you have into kind of the deeper behavior of corporate investors here.

I don't know. Can you try to rephrase the question. I'm still kind of unclear about exactly what you're looking for there.

Sure. Well I guess high level we could say we we've seen a lot more corporate investment in in IoT startups and maybe more corporate investment in general in the past couple of years. Are there other factors that differentiate corporates from institutional investors for example or corporates more likely to stick with the investment over time or maybe less likely or corporates do we see many corporates doing a full acquisition of a startup that they've invested in a let's say a seed or in a round that maybe when it gets up to B.C. the corporate then you know buys them out after they've properly vetted the technology.

I don't have a good answer prepared on that.

OK no problem. Yeah I don't have any any particularly inside either. I imagine there must be some differences between corporates and institutional But in terms of you know long term behavior but not so much a great. High momentum startup So what are we seeing here in terms of the value are We've seen that it's going toward you know in some sectors in let's say shared bike for example. We've seen a couple of companies come out in China and eat up a billion dollars in financing quite quickly.

Are we seeing a similar trend here of having a few high riders that are taking the majority or a significant share of the investment or is it kind of spread thinly over a large large series of startups. 

Well certainly a couple of names that you know keep being able to raise large rounds but there's stacks so to speak in industrial hierarchy. Is.

Pretty vast. So a mapped out the industry before. People want to check out my ideas on this. There is going to see be inside sitcoms slash research and top startups IoT. There's. Just Industrial IoT market map that we have going. There's. Several different layers going on that actually have all been pretty attractive best in. Areas. So you know I think like that top of the Stack you know the the bottom depending how you think about it. You know their sensors and connectivity and the basic things that need like the electricity of IoT. You know there is Sigfox and others that are doing the actual activity basically being telecoms. There's. Sensors there's actual satellite connectivity going on. Companies like. Fleets and Keppler communications are actually enabling constellation satellites to help. Guide industrial IoT in really remote areas. I think above that there's edge devices connected objects. There's robotics and inspection drones 3-D printing things that are. Very much physical worlds that are starting to get. Eaten by software a little bit more. I think above that one further there is in universal platforms. We're seeing companies like C-3 or Ray and places where in the private world that you. Park a lot of your industrial data. Above that you're starting to see more actual applied sensor networks that are specific and tailored to industries and maybe like eye sight machines with manufacturing. And kind of at the very bottom at the very end you know the most software abstracted away. Area you start to see a lot more AI and predictive analytics and cybersecurity for. The OT for industrial IoT. And I would say AI the machine learning the most you know software driven area and this. Seems to be pretty hot lately but it's not necessarily the case some of the most well-funded highest momentum companies are probably desktop metal and carbon three-D. Which are very much. 3D printing company is still very much in the edge device world. And it's hard to say that one is necessarily mean more attractive than the others even connectivity [companies] Sigfox are raising like crazy. From. Great name brand VCs.

No interesting. You know in a conversation last week the person I was speaking with pointed out that.

IoT you know you said 7 billion in investment in the past five years or so. A lot of a lot of interest a lot of invest you know internal investment. Also you know from corporates and internal initiatives or indeed so forth. But when we look at revenue generated even let's say Thingworx from PTC which is kind of a market leading cloud platform they're generating about $100 million which is really not that big of a business.

You know maybe GE Digital's doing more business but a lot of that is is maybe kind of traditional business that they might have been doing for ten years now. Are we seeing some of these startups reach a stable revenue point where they're approaching IPO where they can actually kind of compete in the in the more formal capital markets based on revenue and profitability. Or are we seeing most of these companies still remain venture funded and maybe not the indicator yet that we're kind of hitting that hockey stick turn where where companies really start turning this invested capital into revenue and profit.

 It can be a symptom of just the area of the the acquisition route does seem much higher than going public. And that's probably just because there is a lack of a better term synergy to working with.

Or being acquired by an industrial company. I mean in 2016 and IIoT you saw 32 M&A deals and just one IPO. And you know a lot of these companies for example you know GE bought Norigo. It's two systems basically to shore up its Predix platform. And that's what I mean by then you could probably see more value being acquired than necessarily going public but. That could be indicative of the areas still hasn't you know a long way to come. I imagine you know there will be no opening that platform today that could possibly be a standalone company tomorrow it's just hard to see that given where we are today.

No make sense. I think commercialization in the industrial space is particularly difficult because of all the components to come together so it makes sense that there would be synergies with with corporates that have the existing commercial infrastructure.

Yeah well there's there's also you know there's tons of platforms in the space is incredibly crowded with industrial players. GE has Predix there's PTC Thingworx, Microsoft Azure. There's IBM Watson Siemens Cisco SAP AWS IoT. And that's just like the public tech giant offerings. You know there's tons of universal platforms and I think that's you know that's also part of. It a sentiment least of being young is that they have a lot of. Siloed proprietary platforms going on. And not a lot of connectivity.

Who would be some of the leading corporates here. I think you did mentioned earlier Caterpillar Foxconn who are the company's CEO let's say beyond the corporate VCs space and the companies that are really driving technology development.

Sure. So you know like I said before by a long shot GE ventures is the most active player here but Google Ventures Intel Capital Foxconn Saudi Aramco Robert Bosch. Bosch Venture capital Verizon ventures EVB salesforce autodesk's their venture arms or the corporate world goes you know are very active especially in advanced manufacturing. But you know you're also seeing some new activity. If you compare for example. Caterpillar's activity from 2012 to today you know it's night and day or so and you get more active. Honeywell just just got active as. An kickstarted of venture arm recently. There's just a lot more corporate interests. And in the last couple of years than there's ever been. And where you're seeing a lot of the overlap here you know of these. Industrials. Will co-invest seems to be around 3D printing. So. You know in a couple of these other corporates you know went together on metal and carbon. A couple interesting companies also like MAANA which has been described to me as most like a talented for industry. So. Any analytics question you'd have you know you'd turn and then. Another one would be Xometry. Which I think. I'd like to talk a little bit more about later. It's. Basically On-Demand CNC and 3D printing services kind of like an uber forindustry. And that's kind of like an interesting. First step I've seen in IIoT.

Yeah very interesting and it's interesting that some of the companies you mentioned and Caterpillar and Foxconn both you know what most people would consider to be pretty traditional. You know.

Let's say heavy metal companies so you know on the equipment side in Foxconn manufacturing but we don't really think of them as as driving innovation forward. And then on the other hand we have sales force and Google who you know really we think of as more.

Let's say in cloud you know kind of enterprise cloud computing. Google more on the consumer side and now we have these two very different categories of companies that are competing and to an extent for for the industrial innovation space. This is an interesting and of this is something that you've thought on or have insight but. Interesting dynamic where you have kind of the Googles and the sales forces who are now in a position where they're trying to develop the industrial or vertical expertise the. Use Case expertise to be competitive here and they have the technical expertise. And then you have these other set of companies who have the. They certainly have the end user insight. They have the industrial expertise but they need to develop the competence of I.T.. How do you see this dynamic playing out. Do you see one of these sides making let's say making smarter investments be more aggressive in terms of let's say the traditional corporates versus the I.T. as they are they both you know compete for the IoT market.

Yes and no I mean obviously it probably helps if you're a young startup to be associated with you know blue chip name brand industrial conglomerate. But then again I think you're seeing a lot of I.T. and OT trends kind of converging. So. You know the typical enterprise. Will have you know like a connected lock for example. And also you know the typical. Industrial enterprise will have a full I.T. stack in addition to Otis stack and so many ways you know the differences between industry and enterprise. Are starting to blur. And. To be with one over the other is not necessarily. The route it depends entirely on you know where you're working and a lot of these startups are attacking just you know one. Issue one. You know kink in the data world of Industrial IoT and you know being closer to an I.T. company might actually make more sense for a lot of these especially this software driven ones that are further. Down or up the stack depending how you think about it.

That makes sense. I guess if you're more horizontal focused on a particular technology domain maybe it makes sense to partner with a Google or a Salesforce and if you're really addressing a specific vertical then caterpillar.

Or Foxconn might be the right partner because they can. Exactly and they can be a huge flagship customer.

So we are looking. Looking down into the future I think one of the things that CB Insights has done particularly well is has been able to kind of leverage the data that you're analyzing and help companies to understand what might be on the horizon. What are you seeing in terms of the advanced manufacturing road map. Where do you see us going.

Let's say in the next three to five years three to five years. I mean I see a lot more. Going to making advanced manufacturing facilities really high the facility is even more. Efficient so I think the advanced factor is you know we'll see a lot of this innovation first. But you know at some point you know that maybe that could be diminishing returns on productivity that is. But also you know I think I see a lot of startups just bringing digitisation for the first time. To a lot of kind. of undigitized area so I actually went down into the offices of Odin technologies which is pretty promising. And Industrial IoT the company here in NYC. And really that I saw what they were doing they were hooking up raspberry pis to programmable logic controllers. Basically that great digitising for the first time the data streams in plastics manufacturing. And they were hooking up these little small computers to PLCs that might even date to the 1960s so there's some of these some areas and plenty of batteries out there today in the United States all over that are just getting digital for the first time in their history. So I think first you'll see a little bit more innovation happening you know at the high end where. Factories that are actually you know use of this stuff will adopt the first. But I think more and more you know after five years from now. You'll see more widespread adoption of industrial.

Yeah that's what we've noticed as well is that a lot of the initial investment is going into the cost side looking at efficiencies and that makes sense because usually the business case is a little bit simpler. You can you know if I if I invest in this technology I can reduce this cost bucket you know and save a particular amount of money.

Pretty straightforward. But you can only gain so much there right. So you can only gain basically up to a maximum your cost if you become a zero-cost company. So that's the absolute maximum and that's not feasible. So then it really turns to the revenue side how can IoT help Industrial's increase revenue and then that becomes in the long term much more interesting but also much more challenging because then you're not just looking at improving efficiency but you're looking at business model innovation.

What are you seeing on this side. So you know in terms of the shift to let's say small batch size I mean you were talking around interest in 3D printing.

That's certainly a business model innovation that that's going to radically shift how value is created. Do you see a you know trends right now in terms of interest in technologies that are focused on reducing cost as opposed to technologies that are focused on opening up new revenue opportunities.

Or I see I see a mix but I think reducing cost is you know just the low hanging fruit right now there's just just improving a production line by 1 percent efficiency is just an enormously valuable business problem to solve. And given that there's really low digitisation and a lot of industry I mean McKinsey ran this great graph That I've included a couple of my decks before that. You know the very bottom of you know all the industries the least digitized ones. Agriculture and then construction moving upwards. You know you start to see basic goods manufacturing and even advanced manufacturing that's are behind. Finance and insurance for example. Barely ahead of you know oil and gas digitisation methods. And that's even the best factories and the best factories still have you know we an overall operating efficiency here of. About 85 percent is world class. That's still 15 percent. Efficiency could still improve upon you know the typical typical factory is hitting about 60 percent OEE. In the low end about 40. So really you know. Cutting costs while you know it seems like a fretless compared to opening up new business lines. It really is. A huge market opportunity still and a lot of these advanced manufacturing companies you know will find great niches doing just that. But some interesting new new remixes. That I've been seeing two you know big trends this. Cyber security, end-point security for this industrial IoT world. Also just a lot of edge computing going on and some of the you know "uber-fication" things like Xometry elements these will bring. You know more On-Demand or machine to machine payments and communications. I think that's also an interesting futuristic stuff might be a little farther out. But in the short term I think the two big trends it seems this. New cyber security layers developed and edge intelligence more being done as a device.

Gotcha. What about supply chain or let's say block chain for supply chain so supply chains.

Other area where there is rampant inefficiency and waste and for relatively known you know known causes that just have not been able to be solved with existing technologies. Now block chain some companies I was recently looking at a company called Statwig which you know a small company based in Singapore which is building a solution kind of a tamperproof end to end tracking solution for pharmaceuticals. Obviously that's a very high value problem because a bad batch of pharmaceuticals or a counterfeit that could not just cost money but but really harm people. What are you seeing in this space. Are you saying this is a space which is getting significant real investment. Are people really just dabbling in the concept and is this more of let's say five to 10 year time frame as the technology matures or do you really see this on the rise in terms of companies being able to commercialize solutions and really scale.

So this is you know extremely cutting edge more nascent than nascent. Well you know as a cryptocurrency enthusiast I've been seeing whispers of this you know of. Deeds and an actual supply chain management. I know bitcoin or etherium has been talked about since the early days and I still have yet to see it done. Really well. You know recently there's definitely been an uptick in conversation around this. I think it was you know Tyson or Perdue and some chicken manufacturer is really. Trumpeting that this could be you know really big for. Actual transparency you could see down to the farm of where. You know the chicken that you're eating was raised. And I think that's really interesting I think it could bring you know down the line it could mean that there is in your. You know perfect. Consumption and production you know with proper data complete transparency. It's a really interesting dream could make you know. Really Conscious Capitalism happen if you're able to look up every. Wish. You know input yet you're. That where it came from but it still is so young and there's a lot of different products like hyperledger they're working to attack this. I think what's interesting. Is that you know the first question you should really ask about blockchain in this space does it even need a blockchain. You know a lot of these. Companies already trust each other and do business together and that's kind of like the essential question is. Do these you need to actually trust less computing to make a blockchain make sense. If it can be done with the database today it is just be a more efficient database. So it might not require blockchain.

Yeah I mean the other one I'm considering block chain and you know kind of looking at how the use cases are structured. The other challenge I see in you know I'm no expert here but. A block chain as long as it's fully digital right it's the transactions are all on a computer fairly fairly clear how the data flows. But when you actually get down to say tracking a chicken to some extent you need a device implanted on every chicken and then a reader to you know read that data off of that device at multiple points through the supply chain. I was I was maybe an extreme case I was talking with a. Guy who was running a business in India says MES for for you know steel factories and they're building a solution for scrap yards. Right. Steel scrap just the messiest a supply chain you can imagine dumptruck dumps a piece of you know a big pile of scraps somewhere and then next door they dump another pile of scrap in next door another and we've got a hundred different piles of scrap you know different mixes different qualities all sitting there mixed in together and somehow you have to pick these up and you know repeatedly move them around or eventually put them into production and you have to say this is this you know this scrap has this quality and it's produced in this grade of steel.

But then how do you on the digital side. No problem. But in the physical world how do you actually track the movement of that you know good with amount of reliability. I see this as a problem with supply with with blackchain just that the human or the physical challenge of actually tracking and scanning this. I know you're not an expert you're more of a you know I guess I'm interested in the space so I should put you on the spot I this but I don't know if you have any insight into the let's say apply blockchain to the physical world and the human challenge is actually tracking physical goods.

Yeah absolutely. So you know in supply chain management there's not necessarily one source of truth. So if you have like a trust list network going on. You actually are. Let me let me step back and rephrase here so you know if you're trying to do and apply a blockchain to supply chain management you first need to make sure that there is. No one source of truth. Hold on her. So let me just think about one second. So. I think what happens in supply chain management there isn't. You know one source of truth there's employees that can mishandle a shipment you know can be knocked over and unreported. Bad quality delivery could go a week before discovery. And you know you really would need some verification of a physical and digital Ledger. So really what that would require to me is you know that's the essentially the promise of IoT. you know software eating the atom world or you know allowing you to reliably. Verify what's going on in the world. The thing is though you know will everybody in a supply chain be that diligent about updating the blockchain if something is wrong. It's hard to say and that's why it might not even be for the next few decades. A block chain problem. It might just be a database problem. And you know you can have all the perfect RFID tagging systems in the world and it might make everything more efficient but. It's still unknown even require a block chain to. Extract. You know that improved efficiency. So I'm curious to see how it happens going forward. And I think in theory and unlocks some really amazing things especially when you bring in. You know the machine to machine payments. Or when autonomous vehicles are the ones shipping between the suppliers. And you have a very much an autonomous. Supply chain going on. I think that's where you know you'll really see you know riding on this vehicle wave. You could. Seem to be in that timeline but still. That it even requires a block train has yet to be proven. And there's some really high profile projects like I mentioned hyperledger. That are really trumpeting this. I'm still kind of. A little bit skeptical on just remaining just waiting to see some solid use case for this.

It. No that's a great point. I mean it's so yeah sensors are ubiquitous and an automated.

Probably. That's the first the first barrier and then do we really need a block chain or is there already trust and it's just a matter of data accuracy.

Last last question if you could you kind of shine the spotlight on one you know one company that you've been following and are super interested but might be kind of under the radar for for most investors or just kind of the general industrial audience. Who would that be if you can maybe just share a little bit about what they're doing.

OK let me let me think for a second here. You know. When. You take a look at something form. Well OK. So I think one interesting wave that I mentioned before. Going on with the the edge computing trend. You start to see also more increase cyber security measures being taken. So you know one of the big drawbacks of IoT has been that. It's extremely vulnerable to cyber security hacks. I think if anybody. Who paid attention to them. Mirai botnet could tell you that there's. Tremendous. Possibilities for IoT to be used for the wrong and you know when you're starting to connect. Critical infrastructure especially scanning systems and you know highly sensitive. Areas like utilities you know things that would be the first targets for you know a hacker and especially a nation sponsored one securing you know critical infrastructure so important I think a really interesting company that illustrates you know landed on premise And you know digital security as this company called Niamey and to illustrate what he does you know the kind of they've been operating in wearables for quite some time. But what they do is they use a wristband to read electrocardiograms signals on your wrist. So the example would be you know let's say you're Homer Simpson and your you know your operating your your nuclear facility you step into the bathroom and some you know. Bond James Bond style villain were to come in and attack you. They wouldn't be able to get into the system you know in your absence for example. They would need to have your electrocardiogram on premise in front of the computer to verify that you're you and also able to be in the computer. And so that it's kind of a blend of either. The on premise idea that you that you are you in atoms and then also that you are where you're supposed to be. I think that's kind of an interesting blend. I also think critical infrastructure is one of those things that. You know it just deserves to have a lot of security. And I think you'll see more interesting blends. That will. That will bring biometric security as well as the. Actual.

Digital Security. Ok super interesting use case there. So they require basically biometric proof that the individual that has access to this is sitting in the air in the vicinity basically. 

So talk to me about Edge computing. This is a space I think. What I've been understanding is that there was initially a rush to the cloud because cloud is something that enterprises are comfortable with a lot of traditional software migrated to the cloud. In industrial It doesn't make as much sense because getting the data out of a proprietary environment to factory for example up to the cloud you have latency issues you have security issues.

So edge computing for some use cases makes a lot more sense. We've recently seen more interest there from your perspective as an analyst. What do you see in terms of the maturity and interest level uneducated.

Well I think it's just where the future is going broadly for IO t in general. So right now we live in a very centralized world. When you run a Google search you're hitting the central Google data center and being routed back with your answer. The same thing happens when you're doing a lot of these industrial hoti networks you're sending to a centralized source. But a lot of industry can't afford that ladies like you mentioned. And you're starting to see more compu being done at the device level and a lot of this is riding some machine learning trends going on so you're starting to have. More capabilities you know especially as Nvidia and others are making. Machine learning ever more powerful more machine learning algorithms and apps will be able to kind of catalyze this trend of more computation being done at the device level. In industrial networks too. You don't necessarily need to be sending. In 90 million messages that everything is going fine everything is normal. And you know really it's the anomalies that are worth sending to a centralized cloud. And you're going to see a lot of emerging companies like foghorns the mach. Blade that are you know really annoying af layer and algorithms to help facilitate that. And there's also kind of a subset of security that's also going also as you have more. Device level computation going on. You also need like an identification can I call this identification of things. So you have no Konna for example that does this. I-T and points and and OT and points you know broad industry. Kind of coverage. And then there's also another one Rubicon. That's up that's doing a similar thing. But specifically in industrial Senora's.

Gotcha these vast security companies are these companies that have kind of been doing more traditional I.T. Security previously and now are developing solutions for Edge.

Where are these startups that are 100 percent focused on the edge.

I don't know the answer to that right now. I think. So. I think Makana has an actually you know actually I'm not going to speak to either that I don't know exactly their history. But I do know that are their focus increasingly especially the Rubicon has been touting. The industrial side of it more recently. That because of I've noticed a couple of companies that are. You know recent startups focused exclusively on edge a wider wider group of companies are kind of coming from the cloud side but they do seem to be migrating their focus towards.

Edge computing whether it's on the analytics side or the security side. So I think this is an interesting trend to be on the world. It is interesting and you know I think Peter Levine from from A16 Zee has an excellent video on just this broad trend but.

You know writing these these waves of edge computing happening no more machine learning more computation being done at the device level. You know what happens in the future is an industrial drone essentially as you know a data center with wings. A single robot could be armed with the same capabilities and machine learning that you know a an autonomous vehicle you know. Google is today. So you know the future going forward is that just incredible smarts. Without being you know without the latency of a central server. And this will definitely change what he means going forward.

Right it's kind of synonymous to that.

You know the supercomputer of the 80s is now fitting in your iPhone right. So as we as we continue to improve the.

Power of computing and size then we will be moving towards a world where a lot of work can be done in a you know small devices.

I'm also still looking for a more uber fixation more on demand companies that I think are really interesting I think. Lately I've started to see. More financings going on to robotics as a service. And On-Demand ones I mentioned Zamata before which is really. Kind of like something a semblance of an uber for for. Industrial. Production. So the idea being that there is a de-centralized you know you want an order done that a decentralized network of the season. Machinery to help fill your order. Do your parts and I think this is also. Really interesting. Just with the trend of just in time manufacturing. I recently read that the Honda facility in the UK. Only holds an hour's worth in parts and any given time so you know that the exit is actually affecting them directly because now they are able to have. The parts they need clear the border in time to only be able to hold one hour's worth of parts at a given time sounds like a great use case for commercial 3D printing. And I think youll see more of these you know continue I think here in New York is an interesting one called Voodoo manufacturing same idea as an axe maker bot employees for this company that is. You know a production facility employing robotics. To actually you know take the unfinished prints off of the tray. Of a maker by a 3D printer and I think youll see more at scale printing I see and seeing and more decentralized that as well. But you know whos going to be the uber this situation still you know extremely early innings for something like that.

Interesting one of the topics that comes up a lot in you know in China or let's say in Asia is we you know if you look at China India a lot of the infrastructure is not you know it's not built up so there's actually a stronger value proposition for Oubre business models because if you don't have the mature supply chains and if you don't have the assets properly distributed in the you know the distributors in place to make sure people are able to access assets there is a need for you know kind of new low cap ex models.

So in India Mahindra Mahindra they've launched an internal spinoff or internal startup called Tingo which is basically a Oubre for agriculture equipment so if you need to rent a tractor for 2 hours. And if you're a small team and armer right you've got you know you've got to hector a couple of hectares of land. You certainly can't afford your own tractor but you certainly need one. You know Duraid harvests for a few hours maybe a few hours a week to bring the crop in. So really strong business model and there's just no existing infrastructure to meet that need. Are you. Are you following primarily the U.S. or U.S. and European markets.

To what extent are you also in a survey in let's say the developing nations.

I would say most of it of industrial city finances and startup activity is in the United States. And so for that reason it's been most of limited to that. But we track things all over and I think some interesting. Agriculture trends that you were talking about is DJI starting to offer an agricultural industrial sized drones for crop dusting and other things. I think the remixes on Io t you know. Are not necessarily limited to innovation in that United States by any means. Especially as drones and other you know the physical work aspects of the industry is able to be done on some of these newer technologies 3D printing or drones. I think you'll see more interesting things coming out of the developing world actually.

Yeah absolutely. I think now what is it 80 90 percent of drones are manufactured out of out of China. That doesn't mean all the values coming out a lot of the software might be produced otherwise but it certainly is strength there.

Absolutely. So last last question how can people get in touch with you how can they learn more about your work at CB Insights. 

First I'd highly recommend people who are interested in CB Insights check out our newsletter. We're known online for having a very fun and data-driven newsletter at cbinsights.com/newsletter and you'll see a lot of our data visualizations our research. If you want to get in touch with me on twitter, I'm @NpappaG so Nic Pappageorge. Anybody can look me up send me a message, happy to respond. I'm generally interested in connected hardware broadly so if you have any insights or questions feel free to reach out.