EP061 - A primer on machine vision: technologies and use cases - Iain Smith, Managing Director, Fisher Smith
Wednesday, Mar 25, 2020

In this episode, we give an introduction to machine vision technology, use case adoption trends, and key success factors for an high accuracy solution. We also review decision making process to determine the right tech stack, and cost structure for low and high complexity systems.  This is part 1 of 2 with Iain Smith on machine vision. Iain Smith is Managing Director and Co-Founder at Fisher Smith. Fisher Smith designs and supplies machine vision systems for automatic inspection and identification of manufactured parts on industrial production lines. https://fishersmith.co.uk " />
EP060 - How to integrate the IT and OT for better IIoT deployments - Keith Higgins, Vice President of Digital Transformation, Rockwell Automation
Tuesday, Mar 17, 2020

In this episode, we discuss how to improve the ways a company accesses, processes, and leverages data to make better decisions. We open the question of what digital transformation actually means for industry as opposed to non-industrial sectors. Keith also discusses what the PTC partnership means for IT and OT integration.Keith Higgins is Vice President of Digital Transformation at Rockwell Automation. He previously served as Vice President at FogHorn and CMO at RiskVision until its acquisition in 2017. Rockwell Automation, Inc. (NYSE: ROK), is a global leader in industrial automation and digital transformation. We connect the imaginations of people with the potential of technology to expand what is humanly possible, making the world more productive and more sustainable. Headquartered in Milwaukee, Wisconsin, Rockwell Automation employs approximately 23,000 problem solvers dedicated to our customers in more than 100 countries. To learn more about how we are bringing The Connected Enterprise to life across industrial enterprises, visit www.rockwellautomation.com." />
EP059 - Connectivity, processing, and storage trends reshaping the IoT landscape - Ed Kuzemchak, CTO, Software Design Solutions
Tuesday, Mar 10, 2020

In this episode of the IIoT Spotlight Podcast, we discuss the improving connectivity, storage, and processing cost structures, and 5G’s relevance or irrelevance for IIoT. " />
EP058 - A conversation on AI, social responsibility, and future disruptions - Neil Sahota, Master Inventor, IBM
Monday, Mar 02, 2020

In this episode of the IIoT Spotlight Podcast, we discuss the current state of AI in practice, the integration of AI and IoT, and the importance of new data sources designed for machine use. On a more human angle, we highlight the importance of measuring social impact while deploying disruptive technology.Neil Sahota is an IBM Master Inventor, United Nations Artificial Intelligence subject matter expert, and on the Faculty at UC Irvine. With over twenty years’ experience, Neil works with clients and business partners to create next generation products/solutions powered by emerging technologies. He is also the author of Own the A.I. Revolution, a practical guide to leveraging A.I. for social good and business growth. " />
EP057 - Databases designed for the IoT - Syed Hoda, Chief Commercial Officer, Crate.io
Thursday, Feb 13, 2020

In this episode of the IIoT Spotlight Podcast, we discuss the challenges facing companies when they scale industrial use cases that rely on large complex data streams and trends driving the development of systems at scale by traditional industrials that are expanding their business scope.What is different about data in IoT and why is it harder than other projects?Why do traditional databases and dev-ops methods fail in IoT projects?How do consumer application development trends interact with, and drive, the development of industrial applications?Syed Hoda is the Chief Commercial Officer and President North America at Crate.io, which develops IIoT data management solutions for customers around the world. An IoT industry veteran, Hoda most recently served as CMO of digital manufacturing platform company Sight Machine, and previously had been CMO of ParStream, the IoT analytics company acquired by Cisco Systems Inc. He lives in Palo Alto, California. https://crate.io/products/crate-iot-data-platform/________Automated Transcript[Intro]Welcome to the industrial IOT spotlight your number one spot for insight from industrial IOT thought leaders who are transforming businesses today with your host, Erik Walenza.Welcome to the industrial IOT spotlight podcast. I'm your host Erik Walenza, CEO of IOT one and our guest today will be Syed Hoda, chief commercial officer and president of North America at crate.io. crate.io is a data management software company that helps businesses put machine data to work more easily in real time to enable gains in operational efficiency and profitability together, we discuss the challenges facing companies as they scale industrial use cases that rely on large and complex IOT data streams. And we also explore trends driving the development of systems at scale by traditional industrials who are expanding their business scope. If you're building a complex large scale system that requires real time access to IOT data, I highly suggest reaching out to sit or his team. And if you find these conversations valuable, please leave us a comment and a five star review. If you'd like to share your company's story or recommend a speaker, you're also welcome to email us at team@iotone.com[Erik]Thank you for joining us today. So this is an interesting topic for me because we're going to be discussing database technology for IOT. And I think databases are one of those technology domains where everybody pretty much understands what a database is and why it's needed, but once you scratch the surface, most people, myself included know almost nothing about how a database actually works, what differs between one or another. So this is going to be, I think, very enlightening for, for me and the rest of the audience. I really appreciate you taking the time out of your afternoon to speak with us today.[Syed]My pleasure I've been passionate about IOT for a long time, as we talk further, you'll realize.[Erik]Before we jump in there, I want to understand a little bit more on your background. You have quite an interesting background. So you were working with Cisco for a number of years, and then recently you've had a couple of roles with senior roles with startups. What triggered this transition for you from working with a larger corporate towards taking these roles with young scaling companies?[Syed]I spent 13 years at Cisco. It was a great career. I have a such deep respect for the company and the, I learned so much with gigs around the world and in Paris in Bangalore, India, as well as the U S. And in fact, the term, the first time I ever used the phrase IOT, was it just go about 2007 or 8:00 AM. I heard this weird phrase called internet of things, and it didn't even sound like a normal phrase, what tech that. And so I learned a heck of a lot there and was probably the chief storyteller of a lot of our smart city IOT stuff that we were doing back at Cisco for many years and decided eventually that it was as much as I loved my time at Cisco. It was really fun building a business at Cisco. Although I was in a very large company, I ran often new immersion groups that were in the business of monetizing a new idea, scaling it on a large level around the world and creating a business side of it.I kind of ran sort of startups within Cisco as my previous jobs there and got the bug to do startups. I live here in Palo Alto, California. We have a few famous startups that have started around us, a company garage called HP that people have heard of, and then a company called Facebook and Google. And when you live around these big names, you get inspired and you want to be part of that world. And so I decided to leave Cisco and went to my first startup, which less than two years later got bought by Cisco. That was exciting. And then since then I've done a couple others and now eventually here by myself to create as a chief commercial officer.[Erik]Okay, great. Well, was there any coincidence between you joining the company and it being acquired by Cisco or Cisco already had an interest in the, in the company?[Syed]I think it was just a really good fit with what Cisco is trying to do back then. They needed some edge technologies. They were actually with a company called par stream, which was a database for edge analytics and sort of emerging IOT use cases. And Cisco really wanted that gained some software footprints in that area. So it was a good strategic fit and no, it was just a coincidence that I was at park street at the time.[Erik]Okay, great. And now you're at crate.io and it was founded in 2013. That's right. So it's in this phase now of kind of having a fully functional product and really now scaling where, where do you see create now in terms of their development as a company from the ideation stage now moving towards really scaling out the solution.[Syed]Yeah, we started a number of years ago and of course, like any start up, there's a point of, you know, it's just a good idea. When do we have a product that works then eventually moving toward a product that actually works well. It is a large company. And then eventually you get to the point of, we have a product market fit, and now we're going to hopefully increase the amount of repeatable and grow the business. That's where we are now. We're raising our next round of funding and we're growing the company. We've got kind of 50 to 60 people right now. And probably about a year's time. We'll probably double that number. We right now are headquartered in Europe and have most of our employees there, but I think over the next year or so, it'll become more balanced between North America and Europe where most of our businesses. And so now we're, we're yeah, we're growing to go to market side. We have a tremendous engineering team and we have a very, I would say, a light go to market team and the goal is add to that and grow it. And our biggest, I would say challenge looking ahead is getting to the customers on a wide enough basis and kind of help them close the deals.[Erik]I'm curious why, when you left your last company, I'm sure you had a number of different avenues you could have taken what convinced you to land at, at crate and take on this role. What was it about this particular company that was intriguing?[Syed]Yeah, that's a great question. I learned from some of my really most influential mentors throughout my career. When you think about new opportunities, one, you never leave a job because you're sort of unhappy with it. You should leave a job when you're actually happy and content and doing well. And this was opportunities because that's, when they'll make a more prudent decision, you'll actually, you don't want to run away from something you want to run to something. And so I really had deep respect for my last team that was coming to call cite machine manufacturing platform, a great company, great people. And I just felt it was time to stretch myself in perhaps a different direction or a wider direction. I was a CMO over there and now I'm running in crate.io sales and marketing and partnerships, business development.What attracted me here? Wasn't the technology. It wasn't say the market we're serving, but it was the people I met. I think it is such a privilege to work with people that you respect, that you like and you enjoy working with, and you can learn from when you have the dose things coming together, then it makes your life better, frankly. So when I met the people that created crate.io, I was very impressed. These were incredibly smart people, but also very nice people, people that you enjoy working with now that really matters in a startup much more so than in the big company, because unlike a big company where you might have 50,000 people and you can go hide in living number 45 in a startup, you're a family. I mean, this is it. There's no hiding me. We worked together. We spent a lot of time together. We play together, we fight, we, we brainstorm. And so you really have to be a part of a culture that you believe in and you feel like you can learn from and thrive in. So for me, that was why I joined crate. I was very impressed with people. Of course, after that, is there something you can do to help? Is there a, does it align with what you're good at, what you want to do and what their business needs?[Erik]Yeah. Great. Yeah. Well, I can certainly buy into that where we're 12 people sitting here and it's certainly a family. That's the game when you're young. Let's learn a little bit more about what crate actually does. So, you know, again, I think if you say great does databases for IOT, people generally have some vague concept of what that means, but what is really creates value proposition?[Syed]Step back a bit as the chief commercial officer of a database company, here's my challenge. Nobody gets up in the morning and says, Hey, I want to buy a database. Nobody says that ever. And in fact, nobody actually wants to buy a database. They already have databases. There are plenty of them more than they wish they had. And so the last thing we want is just another database. So it's not easy. Our jobs, it's a hard job because it's, we're selling in a space there where someone goes, Oh, I already got one. Actually I already have five. Okay. So the question is, why should somebody care? Why should somebody listen? Well, let me, before we go down to the database level, it's going to go higher up a bit. And in the space of IOT, when you look at the research done by a number of firms like Cisco, like McKinsey, Bain, BCG, et cetera, then these firms have found that roughly about 70% of IOT projects have failed.And what does failed mean? It means it didn't reach the expectations or aspirations of the original charter of the project. Now, maybe there were unrealistic expectations, maybe not, but roughly it has a 70% amount of failure. Okay. Now, why does that happen? Why did it happen and what can be done about it? We look at the top three reasons of why projects have failed in my OT. There are three consistently, again, same kind of companies, including others like IDC and Microsoft have studied this and found that there are really three reasons that are common. Number one, the lack of necessary skills at the T number two, sort of a siloed and resistant to corporate cultures. So the first two reasons are about people and organizations and culture. The number one technology reason is the data and it infrastructure. It's just not what it needs to be to support these applications and initiatives.In other words, IOT projects fail technologically because of data. That's why. And the question is why it is so hard that industrial space it's harder than any other space. Morgan Stanley study data in my factory was that in our classmate verticals, there were a couple of unique attributes. One, there was more data and factories and manufacturing, supply chains than anywhere else in the world. And number one, vertical number two was government. Number one was in the industrial space, manufacturing factories and supply chains. And it was growing faster and the data variety was higher than anywhere else. So that's what makes it so hard is that the shape and scale of industrial data is very different than anywhere else. The problem is that the tools, the legacy tools that are being used to manage this and bring it to life are designed for this. So we have a challenge in our hands, traditional databases, traditional infrastructure technologies weren't designed for this.They were designed that many of them for kind of the web scale world. Let me give you a little, maybe a little brief history about kind of how we got here a long time ago. There was something called ERP and client server computing and mainframes that all sort of made up the landscape. And back then we had relational databases that started to emerge and companies like Teradata and Oracle and IBM and Microsoft, all sort of flourished with relational databases. Then something happened a few years later called the internet business e-commerce e-whatever. And what happened there, that was the beginning of web scale and web scale meant that now more than ever availability and redundancy mattered the ability to support a massive number of users, concurrent users. And so all of a sudden the companies like Mongo DB and Splunk and Hadoop and other sort of came out, were born with flourished and then something else happened the world of digital transformation and IOT, et cetera.And what this meant is that the landscape changed in a number of ways. It meant that now machines were users. We had connected things I OT and it began to convert. And what's expected from databases and technologies that support data were very different than before what's happening. Now, is that the volume of data while it's very high and big, data's become bigger and faster. The variety is way different than it used to be. That's the biggest difference. Variety of data has been massively increased compared to yesterday. And that's, what's made it very hard for infrastructure to catch up and support IOT projects. We expect data to be adjusted fast, to be analyzed quickly and scaled, and it's not easy to do.[Erik]Okay. So you have a, yeah, you have a kind of a nest of issues here, right? You have the variety of data. You have maybe different real time requirements around that data. So some of them might need to be stored and processed on the edge or on the device or on facility, or, you know, some can be moved to the cloud. You have a lot of messy data that doesn't have good metadata attached to it. So data that's maybe high in volume, but hard to, hard to actually interpret or analyze what is possible for a database to address. And, and what are, what are the pain points here that are somewhat outside of your scope and maybe need to be addressed at the protocol level or by, by some other solutions?[Syed]What a good database does in this world is it handles the data. It's quite simply, it's able to take the data you got and make sense of, but give it a context. Number one. So it's able to, one is just handling it, handling the kind of data that exists out there, handling the speed of the business. So ingestion speed and, and the real time nature of the use cases require databases to be fast but fast at scale. And that's, what's been very hard. A lot of databases like acknowledges can be really fast with small amounts of data or very simple queries, but what if the queries are complex and what if they're different every single time? What if the data is in terabytes and petabytes and beyond? What if the data is mixed and even mix of structured and unstructured data? If it, what if it's not just time series that's, what's hard. And that's why no company frankly, has been purpose built for this world right now. If you look at most IOT platforms, if you look at most applications, people buy by industrial companies, they're using web scale technologies to try to do IOT scale and we're failing because you can't scale it. It's so hard with databases we have to do is handle kind of the scale with ease, kind of the speed with ease and the data types with ease.[Erik]And are you doing this at great deal? Are you doing this across from the device through the cloud or where can create be deployed?[Syed]One of the advantages we have is we're actually exactly, you said their device to the cloud, we're an edge. We can be at the edge and we could be at the cloud. And what's unique about us is it's the exact same functionality, whether it's in the edge or the cloud and in between. So if you look at most use cases and especially factoring, that's what people want. They want to be able to do certain things in the cloud for things on the edge and kind of go back and forth as necessary as they wish. And so we have no limitations and I'm helping them do that. And with massive amounts of data, and we can run on something as small as a raspberry pie, which isn't recommended for most use cases. It sounds nice, and it looks good on a POC, but the reality is it's probably a little light for most of these cases that we get involved with more of a gateway up to the cloud. And you know, we, we have a very strong partnership with Microsoft Azure where we also, one of our biggest customers runs on AWS. And so it's important to give people a little bit of flexibility from that realm, but we are friendly, but both edge and cloud.[Erik]Okay. Yeah, this is, it's kind of an interesting topic. We had a panel a few weeks ago and we had somebody from Microsoft and somebody from Nvidia on the panel and, and it came up that they were partnering with each other. And so the, you know, we asked, well, don't, you, you kind of have competing stories, right? Microsoft is saying, move everything to the cloud. And the video saying, you should compute on the edge. Cause that's, you know, that you need to be faster and more agile. And it seems to be emerging that we really need this more nuanced approach. And, and actually they both bought into this. And then yeah, we're partnering because companies really require the right deployment at the right place. And this, I think has been one of the challenges in the IOT space of companies figuring out where data needs to be deployed. I think it's still somewhat of an open question. How involved do you get when you're working with your customers and advising them? Is it, do you have this kind of advisory function? Because I think this is a big open question for a lot of operators where, where they should be managing their data.[Syed]Yeah. Yeah. This is a great question. When we hear this all the time, should I do this or should I do this cloud? And the answer is super simple. The answer is yes. Where is the use case happening? What do you need to do? Don't get, so I tell my customers here at previous companies don't get lost in the buzzwords. Don't look for an edge project. Don't look for a cloud. Do what makes logical and business sense for the project, for the, for your projects and the use case. And I can tell you that whatever you want to do with the use case, there is a technology that can help you make it happen, but don't let the technology lead you, but the use case lead you. And so the answer is it's going to be a blend. It's the same thing people talked about with public clouds, private clouds, hybrid clouds, yes. All the above, but don't let the technology dictate how you do what you're going to do, figure out what you want to do. First technology will follow. We tell them, look, we're agnostic. We can do either one. What makes more sense? What's going to be more effective and that's what you should do.[Erik]Yeah. And this brings us back a little bit to your earlier point without out of these three pain points. Two of these are, are human related. And one of the big challenges here is that, you know, the it department would, this would typically be in the function, you know, the purview of the it department, determining which technology to use, but especially in the IOT space, the use case, the application, the form factor for the application, these are, these are very critical and these to help to determine the technology. So you really need a very strong role or a strong input from the business. Whoever is going to be operating this application. And, and there's a lot of friction there, right? Between decision maker, intuitively I would think you're selling to the CIO and the CIO is team, but the application is so critical here. How do you communicate to organizations between the different stakeholders?[Syed]So that's a really great question. We've actually modelled the kind of the buying process is pretty complicated. So here are the two points in time. When people say, gee, I want to buy a database. There's only really two points that happens very often. Number one is when a couple of companies building a new application or buying a new application and this application that I'm going to build, run on the databases that I have. And if the answer is yes, great, we just go right in. Good. We do it. If the answer's no, then we figure out, well, how should I write this application on what database the buying center in that case is very often heavily, I would say recommended the head. Recommenders are DevOps, developers, developers have preferences. And frankly, it's like a religion, right? They love a certain kind of database and they don't like another time.And it's always a religious conversation. Normally the CIO doesn't really care, frankly, if I'm going to buy a new database from modern application, dev ops will have a very strong voice in that. And a strategic shy will say fine. You know, whatever is most effective. The business leader just wants to deliver the use case. They want, you know, good data data. They can trust. They want applications that run in a way they're expecting the use case to be delivered. So what I want to buy a new application, I look for database dev ops has a very strong role there, more so than probably anybody else, but maybe they're not their decision maker, but it's hurting your recommender. Now, the other point in time, we called the second control point of database opportunities with this. I have an application I've developed it, let's say at small scale, in a POC or a pilot, and now I want to scale across 150 plants or machines or whatever.I'm rolling it out. Oh, now it all of a sudden comes to a standstill performance level. Do we need needed to be, or it's very expensive because now I've got to store a certain amount of data and I got to buy more instances or larger databases. So the cost goes way up performance. My neck goes Michael way down and all of a sudden, my application's not running the way it should. I got a problem who is the buying center in that case, all the dev ops, person's already doing the application typically, no, it is a CIO to your point earlier. And the business side saying, gee, we just invested all this money and time not working, fix it. And now the buying center has been kind of widened. And now there's more scrutiny on the choices that were made and looking for the choice that can properly enable this use case.[Erik]Gotcha. And I guess then it's the business side. There's going to be primarily the side that's complaining. That's going to the CIO and saying, Hey, listen, this this application is not meeting our requirements, but they're not going to be then making the decisions. They're just going to be driving, driving that a decision needs to be taken.[Syed]Yes. And, and the reason we've had a problem here with OT and it, and dev ops is because lawson the flavors of technology that, you know, that we're the dev ops guys have learned on. Aren't the ones that scale the best. And very often the ones that have scaled the most like legacy, large companies, we all know about the Oracles of the world and IBM's world and et cetera. Aren't the ones that modern developers, frankly like to use like these other tools. And so these worlds have to try to connect. And now you have to have the ease of a database and tools from modern, modern applications, with the scalability and reliability of the legacy applications. And that's where frankly, we play, you know, we come in with as much focus on ease of development as on ease of scale.[Erik]Yeah. That seems to be a big trend. Now it seems to be that industrial applications are learning from consumer to an extent. Whereas I think if you looked 30 or 40 years ago, industrial is really leading the way in consumer tech was somewhat following, but at least that's, that's somewhat my impression, but now it seems like usability is, is something where industrials are or industrial technologies are really trying to play catch up now and are learning somewhat from that consumer landscape.[Syed]And it'll continue to it'll continue exactly. You know, because I'm an age myself, but there was a point in my career. I worked at IBM. I came out of college and you know why you went to work. You went to work to go check your email. When I was a fresh grad, I went to work to go do your expense reports. And so you did like electronic stuff at the office because you didn't have a laptop. No, you don't have to go to work to do any of those digital things. It can all be done from anywhere in the world. You've got to work to meet people, talk to people, your brainstorm life. And even that you can do on video, of course, right. But what's happened is, is the next generation and the next generation, whether we're at the office now, often we're behind a generation technology than at home.When I was very young, fresh grad, I went to work and I was ahead in technology at home. I was behind now, it's flipped. Now the things we have in our home are smart. You know, Google and Amazon devices, et cetera, can be often more advanced in our work environment, which is why, what we expect in our daily lives being delivered in the home. We now also expect at work and work is catching up. So what used to be considered to be a good interface in industrial application now looks terrible because companies like Apple have redefined our expectations of what good interface means. And certainly we're learning from that. So it's not just pretty charts and pretty slides it's even the ease. So in my world, the database world, it means when you install something, it should be zero configuration. So just sort of automatically get set up when you expand it and scale, it should just, you know, quickly sort of drag and drop things. It shouldn't be a PhD and, and databases to be able to run this thing should be anybody[Erik]Interesting trend. Let's talk a little bit about your customers because on the one hand, this is a solution that could be applied pretty much by any industry for any use case. But I imagine that in reality, there are some industries and some use cases that you tend to be serving more. Is there kind of an 80 20 rule? Is there a cluster of industries and use cases that you found for whatever reason have a, a greater need right now for a really IOT design databases?[Syed]Yeah. And, and I'll say, you know, we use the word IOT as a moniker, but I, I think it's really any industry with a lot of data. And in particular, a lot, a large quantity of machine data where we'll time matters. And so our customers you know, they, we have a lot of industrial company or companies that are our customers, but we also have, for example, one of our biggest customers is McAfee software Qualtrics. The recent acquisition MSAP is a customer as well as big industrial companies. So it's a, it's a wide range of array of industries, but certainly have you got any industrial sites? The common theme amongst these companies are that they are enabling use lots of data, terabytes, petabytes, et cetera, time matters. The shelf life of data can often be very short, you know, in a company like a Mac, either they're doing security well, security time matters.It'd be able to find out what's going on and do something about it is very time sensitive or in a factory. One of our customers, a company called elk love, you probably haven't heard of them, but they aren't a multibillion dollar packaging company that makes packaging for companies. You have heard of like Coca Cola and P and G and others. And what matters there is that when I'm running a production line at very high speeds, with high numbers of products, if there's any issue I can know about it immediately and act on it fast. So lots of data, real time use cases that rely on data being delivered to where it needs to go quickly with insights to help me act on it immediately.[Erik]Yeah. It makes sense. Right. So you, in this, I guess, is this a cluster that's expanding or the number of industries that are dealing with this, but right now, yeah, you, you have your kind of core niche of industrials and then other companies that are serving these use cases, which are maybe technology providers like McAfee, the system integrators, are they heavily involved? Because I mean, I guess on the one hand is, you know, needs appear when there's an application being built and, and the system integrators or the other application providers that are going to be the ones that know when an application is being built or maybe when an application is struggling, probably more than your sales team. Are you typically going to market together with other companies who are helping to build these applications or selling, and in your kind of saying, you know, when you encounter this problem, then we have the solution that you can bundle in, or are you going directly and in managing relationships, how does that?[Syed]Yeah, both, both the channel is very important to us and it'll grow over the next year. And beyond that, 10 will become bigger and bigger. And I would say a very big part of our actual revenue will come from channels over the next a year or two. We're growing that out right now. We're directly, we're mostly direct selling, but then this is 2020 the shift beginning to happen in a dramatic way. A lot of our customers, I would say our IOT platform type of companies. And I mean that in multiple ways, for example, some large industrial companies, we have two right now that we're very deeply involved with publicly, but there are two different companies that are both industrial companies that are building their own IOT platforms for their own manufacturing of products, as well as selling those platforms to their customers. So these are companies that are not software companies, they're industrial companies that happened to be selling a platform.And so that's a very typical, I would say, customer similarly, there are software companies that we're talking to right now who that's all they do. They don't actually make a physical product, but they build software with one of them will have a press release in the next 30 days or so coming out with, we're gonna announce a partnership. We're going to be the engine inside their platform. Their platform is application layer delivers business value based on certain use cases, we're going to be the engine and side that helps take all that data and make it useful. It will give it a context to make the application work better, frankly, and faster. It's a combination of, I would say companies that are building their own applications and some that are building applications to sell the companies that that's right now, where we at and channel become an extremely important part of our business. Yes,[Erik]That's an interesting trend. This first set of examples that you gave of industrial companies who are building platforms for both our internal operations and also as a new revenue model. This is something that we've seen very actively in China and where you have the Petro China and the state grid and these very construction equipment OEMs. So he's very traditional industrial companies who are building platforms and they use them first for their internal operations. And then they start to sell them up and down their supply chain to partners. And then they start to aspire to moving into new verticals. But it's created kind of a new dynamic where these companies don't have very much it competency. Even if they hire a couple of hundred developers still, it's not really in the DNA of their organization. So they're much more in the role. There's a concept that Alibaba is kind of pushing here, which is one plus end, which is, you know, Alibaba provides the high level of functionality.And then you build your vertical platform on top of that. You provide the domain expertise, but of course, it's, it's like one plus N plus X. You still need people. You need crate. You need other companies to fill in a specific competencies because really what the, the industrials and I, I think this is going to be a big trend. I think that we need more, the application is so critical that we need the industrials to really take an active role in building these applications. But the tech stack is so complicated. There's no way that they're going to manage entire tech stack, which means that they, it needs to be really a partner approach.[Syed]Great observation. It's so interesting. You say what you just did because we had in three different companies, we had a couple of different philosophies in our technology strategy, and I think one of them was wrong and a couple of them are right. Here's the difference. You have to figure out what you're good at. What is your real value property you really good at and weigh it? Secondarily. So the smart platform companies are smart. Industrials realized their strength is what the shop floor, the process, the industrial side, the automation side, that's their strength. And while they have a lot of software, people have some very good ones. They should probably use them off the tool, off the shelf tools where they can, because database company puts all our blood, sweat, and tears in the database. You're not going to design a database better than us or any other database company, because that's all we do for living.So you got to focus on what you're good at and pull in products where, you know, you don't have that level of depth. And so the smart ones are doing just that. The smart ones are also trying to minimize what I'll say, the grunt work of their own teams and focus on the value layer. What does that mean? Forrester studied the world of data analytics and data science. A couple of years ago, I had a very well written document about how I'm going to be rough here. About 70% of the time that data scientist works in a company is spent on grunt work, cleaning data, sorting data, organizing data, et cetera. 70%. 30% is spent on deep data science. So grow work, and the stuff they hate to do is a majority of their job. Alright, now let's translate this into the world of applications and IOT platforms.You should not be focusing your time as the big industrial company on getting the data ready. The data gets collected, and if you have good tools, good databases, for example, that should be done as part of the database. If you're spending time, we just spent yesterday with another company or they're using three different databases because not one of them can do the job. They use database A for this part of the task, I database B for that part of the task and see when they have a time series, only database and they have one for this kind of stuff. And one is expensive. There's you a cold storage, all of the junk. So what happens is they have all these people whose jobs are just to administer the database. That's a waste, focus your attention on the application, the value, add and get product, get the infrastructure products that can take away the crop work for you. And that's kind of the value we add is that we allow them to not have to manage five databases to do the job of what Craig can do. Go focus your attention on the value you bring to your customers, not the database.[Erik]Yeah. I think this is a very positive trend right now because it means, you know, as companies start to adopt this approach, that we're, we're going to be freeing up a lot of effort to focus on solving real problems, as opposed to, you know, building replicas of, you know, of existing systems, right. This is, this is the way we need to be moving. So we have we, we can lower that that failure rate from 70% and start start really growing the market. Can you talk to us a couple about, about a couple of cases and, you know, maybe just walk through us if you're able to, from when the conversation first began through, you know, the decision process into deployment so that we can understand the cycle of a great deployment,[Syed]Several examples of where a company was trying to run an application on a database or some tools, and as the application grew and got bigger, it started to fall apart. And so you know, one example, I'll tell you a company that essentially built an application that was running fine, but our data grew. And all of a sudden we got into hundreds and hundreds of nodes and billions of rows, et cetera. It just started to fall apart and they needed to sort of do this differently. And we went in there and replaced kind of their SQL elastic search database with, with us. We were by 20 times, faster by 75% lower footprint costs and allowed them to grow, grow very easily. What would happen here was they built a really good application on a, I would say, legacy underpinnings. How did it start?Well, you have often you have a like said earlier, you have a business leader of a function or a sponsor, but use case that has a plan to deploy has a plan to then scale. And when all of a sudden scaling doesn't happen, we can come in and help. Hopefully save the day by putting in an IOT skill database to an IOT project. Sometimes it works like that. Sometimes you have a company like the case of, who I mentioned earlier, a manufacturer of packaging and products, where they have an aspiration, they want to do much more data driven, smart manufacturing in their plants. They have 180 plants around 45 countries around the world. They have world-class, OAE, I've never heard numbers this high in my career. I can't disclose them, but I'll just tell you, I'm just amazed how good he is, but they wanted to do better.And they realized the only way I can get better at this is by sort of the last analyzing to the next level of digits, it's kind of the next or last mile of the processes. And what they found was, you know, at a very generic level is that if the people on the plant floor just knew certain information sooner, or if we could filter out junk information more effectively, they're more effective. So in other words, I'm telling you 10 things before those things don't really matter. And three of them, you can wait on, but three, you got to act on right now. I just did that better. I could even improve even well. So this requires being able to get a bunch of data, the hard data that's not easy to necessarily analyze and make sense out of it very quickly in a real time basis. So this case, we aligned the aspiration of what I'll call very visionary leadership with thinking, gee, how can I do this better? And we helped bring it to life by doing it in a couple of plants and then growing it from there, what would be very soon a hundred plants, let's say real good use of data on the shop floor[Erik]For the first case. I'm curious, when you moving into a situation where an application has been built on more of a legacy database, that doesn't scale, well, would you deploy great. Does the application architecture have to change at all? Or is there, I mean, is there a zero change? Are there some, some, some moderate changes?[Syed]Yeah, yeah, of course the classic answer. It depends, but I can tell you that in most cases that we've seen, these are modern applications, I would say generally more modern databases. But what you'll see is that there are databases, cult databases from modern applications that are really good for developing applications on, but not really good at scaling. So the good news is the migration from those databases to ours is very easy. Migration was never end up in much of an issue that has been probably somewhat easy because we can handle a wide variety of data quite easily. That's real space. I would say we've had very little problem with the migration because often even built on data, but I would say more modern databases, but just databases that weren't designed to scale very well.[Erik]Yeah, that's a good point that there's a whole set of technologies that are really designed for helping companies to take it from R and D through a pilot phase. Right? And that's a, that's a necessary, you need tools that are designed for cost effectiveness and, and, you know, quick deployment. And that those tools don't necessarily scale, but they're, they're not bad tools. They're just suited for that particular, that particular set of problems of taking a product through thorough validation.[Syed]They scale, well, just not IOT scale. So, you know, I I've heard this before I put the customer first. They said, well, wait a minute. So, and so company scales a lot, it's scaled to this side. Sure. They scale that size as long as all you're doing is time series data and no other kind of data. And you have extremely simple queries and they're known queries that get repeated a lot, but the reality in the industrial space, the queries are change. And not always super simple. It's not just time series data, as we all know. So they, they do scale, just not in the world that we're talking about today.[Erik]Okay. So we have to stop thinking about scale in terms of just the number of queries or data points, but in terms of also the, the complexity of the, of the system here, you'd mentioned pricing earlier. We don't need to get into the specific pricing of crate, but what are the variables that impact, this is this, what would be kind of a typical equation? And cause I think it's quite challenging often for companies to understand what might be the lifetime cost of a solution, you know, because you know, you'll go in and at first the cost is very minimal, but then of course it can scale very quickly when you're talking about large data sets. If the algorithm is based on volume of queries or something, what is a typical algorithm here?[Syed]Yeah. It's a run to around the number of nodes is that you use you know, the amount of data you have or the cost. I can tell you the street or your sort of rough numbers. If you look at kind of most, I would say traditional modern databases, but what our competitors were often three to five times cheaper, it's the way they were architected. It's the way we scale is the amount of resources that we use. So the question becomes, why are some of these more expensive? Well, it's the kind of hardware utilization. If the cloud footprint that they have, that makes a lot of these companies more expensive, we're actually really efficient in terms of how much hardware we need. We also run on commodity hardware, really cheap instances. And so these are the kinds of things that really make this cost a lot lower.So those are, there's a pure running operational cost. And the pricing model is really bad. Number of nodes and amount of data, et cetera, which is very typical, most databases. The other piece that I'll tell you that we have a big cost advantage is the actual human part Eamon of people. You need to run a system to manage the system to scale it, do these sort of flavors of technology, or are great technologies that help a lot of people, really smart PhDs to be able to do really simple things. And that's part of the challenge too. You want to have databases technologies that are easy to operate, easy to scale and don't need super deep high expenses, costs, resources to do basic things that adds to the cost.[Erik]Yeah. I've seen a big trend there of, of looking at how to improve the efficiency of teams that are developing also on the application development side. You know, there's, there's a lot of different areas where labor ends up being the very significant cost here, right? And this is this is a need, right? Especially when you're dealing now selling with industrial companies who don't naturally have these people on, you know, so this means new hires and hires, which maybe they're not used to hiring. Right? So it's, it's quite different when you're, you're hiring your first set of data, scientists and PhD is where you haven't really managed those people before. It's not just the cost, but it's also the complexity of managing kind of a new group of, of individuals in your organization.[Syed]We'll talk to our customers and we have like reference calls, they'll ask them. So who manages the crate database? And when you scale it, when you grow it, who does that? Our prospects are shocked, just how simple it is to use. They're shocked by the level of person they're running it. They're almost putting that. Well, wait a minute. How about the rest of the team? And there were surprised of how easy it is. And that goes back to your earlier point. Ease of use is something that is expected and the standard for ease of use has greatly increased over the years and that you should expect that from your database. It should be, it should be effortless for the easy shouldn't require massive amounts of people or training in order to run it. And that's one of the value propositions that we talk about. So yes, hardware costs are low, but also the people cost of managing and complexity is also something that really should look at.[Erik]Talk to me, see, what is going on in 2020 for you? What are the, I know that you're, you're moving more from you know, product development towards ramping up sales, right?[Syed]The milestones for the coming year. Yeah. So for us, it will be, we have a big push toward expanding our cows. As you mentioned earlier, we're going to have some very interesting partnerships that we'll be announcing and growing with. You're going to see some of our customers who have let's say started off with a smaller deployment of a few plants growing to a hundred and more comps. So this is a year of scale for us, massive scale in terms of the growth within some of our existing customers to some new, big names that will well from Percy's is about as well as partnerships we're at that stage now where, you know, you can kind of in the life of a startup, you have, it's never a straight line it's lines that go up really hard or fast, and then flattened out and go up again. And this is, I would say, 2020 for a year for us as a year of scale and a year of some pretty good growth. We've already had that growth. I mean, we've grown it three X this past year, and we expect 2020 to be also at three next year to this year. So it's wearing a very good trajectory and I'm pretty, pretty excited about what's coming up.[Erik]Right. Yeah. Fantastic. And I know you're also raising a round, so I guess there could be a number of people that would be interested in getting in touch with you, whether they're investors or potential partners or customers, what is the best way for people to reach out to you or, or more generally to the create team?[Syed]I mean, certainly I am on LinkedIn, I'm on Twitter. So I welcome people. I say hello and happy to talk about anything that I won't talk about in terms of our business, in terms of our strategy products or fantasy football, you name it, all those things are fair game.[Erik]Cool. So we'll, we'll certainly put your Twitter feed and LinkedIn in the show notes seed really appreciate you taking the time to talk with us. Is there anything else that you want to cover that we missed before we call it a day?[Syed]No, I like to always ask a good host, like you a question as well. So you tell me something as well, looking ahead now it's appropriate. People always have predictions for the next coming year. What do you think in terms of 2020? What are a couple of interesting or exciting trends in the world wire T that you're keeping your eye on?[Erik]Yeah. I mean, there's a couple of points that we already talked on here, but when we set up our company in 2015, you know, really, but when you go to conferences, people were still asking the court, what is industry 4.0, what is industrial IOT? You know? And then we've moved fairly quickly in the past few years from what is it to, you know, what are the use cases? And then what are the challenges? And, and, and then, you know, what, how should I start? And I think now we're really getting to the point where a lot of companies have done their pilots. They've maybe learned some hard lessons, they've wasted some money, but they've learned. And now I think we're really moving towards a point where companies are going to be able to see scale. And part of this is the learning curve of the end user or developing the applications for their business.And part of it is the development of companies like crate that are providing better functionality to build on top of. So this is really what I'm looking forward in the coming year. We see a lot of this in our class based, you know, companies that have maybe in the past and talking to us about what should we be doing, or how do we start and are now moving more towards, okay, now, how do we scale this, this system that we've been playing with for the past 12 months? So that's, I think the big trend, and then this other trend that we've already touched on a bit, but of companies figuring out how do we build solution that are not just an operational improvement, but that are impacting our, our bottom line. So we see a lot of companies that are, you know, they've been elected trickle or mechanical engineering companies or whatever engineering companies, even for the past, you know, 50 or a hundred years, and are now looking, how can this new set of digital tech [inaudible] enable us to grow revenue?So we see chemical companies, for example, that are saying, you know what, maybe we can sell a smart factory solution that includes sensors and platform and applications to our customers so that when they purchase our chemicals, they can utilize those more efficiently. And, and, and, you know, this then makes them more likely to use us as a supplier, but also we can have incremental revenue on this IOT or this, this system that we're selling them in the new, you know, we see construction of companies that are also looking at how can we not now compete on lifting capability, but we can compete on usability or remote control and so forth. So we see a big trend among these companies to look beyond operational improvements. Although those are, those are still critical, but to look really at how they can start to integrate this technology into their, into their offering.I think these two things, I think we're going to see companies really moving from pilot to scale, and we're going to see more companies figuring out how to adopt digital technologies. So I would say non-IT companies that are not digital natives, but are now adopting digital technologies into their core offering. Those are the two things that I'm really looking forward to in the next, you know, and this is, you know, I'd say, I think we're going to see a lot of growth in this in the next year, but this is a decade trend.I think we're just at the start of a longer term trend here. I totally see that as well in the, and those things are very connected as well, by the way, of course, right? You gotta, couldn't pilot to scale, to fundamentally changing your business models that you're suggesting we're seeing the same thing. We're seeing the smart companies, the leaders not getting excited about a 14th pilot. I can excited about another science project, but really saying, how are we really fundamentally improving the business? And if so, let's do this, do it all over the place. And so then your first point of pilot to scale it, we see that often that excites us. And that's why we're here in business, frankly, a great 2020 for both of our organizations. I hope it's been really a pleasure to speak with you.[Outro]Thanks for tuning into another edition of the industrial IOT spotlight. Don't forget to follow us on Twitter at IoTONEHQ and to check out our database of case studies on iotone.com/casestudies. If you have unique insight or a project deployment story to share, we'd love to feature you on a future edition. Write us at team@iotone.com" />
EP056 - Airbnb for telecommunications - Frank Mong, COO, Helium
Monday, Dec 09, 2019

In this episode of the IIoT Spotlight Podcast, we discuss how Helium is providing universal and affordable connectivity for IoT devices, provided by people, and secured through a blockchain. How do you provide connectivity where there is no infrastructure? How do you incentivize people to contribute to an open-source connectivity ecosystem? What are the connectivity challenges facing IoT device manufacturers today? Frank Mong is COO at Helium, where he is leading the global go-to-market strategy for the company. He is responsible for sales, marketing and business development at Helium. Before Helium, Mong spent 20 years in cybersecurity including CMO at Hortonworks, SVP of Marketing at Palo Alto Networks, and VP/GM of security at HP. frank@helium.com Helium is building the world’s first peer-to-peer wireless network designed specifically for IoT devices. Helium's technology will help spark a wave of innovation by enabling a new generation of small, mobile, low-powered IoT devices that include remote wildfire and agricultural sensors, micro-mobility trackers, and more. helium.com Use cases: Helium.com/enterprise Developer resources (software, hardware, firmware, connectivity documentation): helium.com/developers Buy Helium Hotspot: helium.com/store________Automated Transcript[Intro]Welcome to the industrial IOT spotlight your number one spot for insight from industrial IOT thought leaders who are transforming businesses today with your host Erik Walenza.Welcome back to the industrial IOT spotlight podcast. I'm your host, Eric Walenza CEO of IOT ONE. And our guest today will be Frank mung, COO of Helium. prior to joining helium in 2017. Frank gained 20 years of experience in cybersecurity as COO of Hortonworks as VP of marketing at Palo Alto networks and VP of security at HP. Helium was founded in 2013 by Shawn fanning, the founder of Napster and aims to provide affordable and universal connectivity for IOT devices through the use of a decentralized network, owned by people and secured through a blockchain together, we discuss the connectivity challenges facing IOT device manufacturers today, and how Helium's peer-to-peer wireless network could reduce costs and improve coverage and open spaces as an alternative to cellular coverage. We also explored the role of blockchain and crypto tokens and securing the network and providing an incentive model for the individuals who build the network by deploying helium hotspots. I hope you found our conversation valuable, and I look forward to your thoughts and comments.[Erik]So Frank, thank you so much for joining us today.[Frank]Thank you, Erik. It's a pleasure to be on.[Erik]Frank, before we dive into helium and the technology and the business model, I want to just share a little bit of your background because you have really quite a deep background in cyber security. So people know who they're talking to. So you've, you've worked with Hortonworks with Palo Alto networks with HP. Can you just give a quick overview of what your roadmap has been and how you ended up now joining a startup helium?[Frank]Sure. It's always been in cybersecurity. It did start in early days and they started as well. When I first started my working life, I was at a startup called ignite technology is a company focused on managed security services or managed firewalls back in 1999. And when I was 23 or 24 at the time, it was the innovation companies and individuals were first connecting to something called the internet and using something called Netscape browser. If you remember back in those days. So that was definitely revolutionary time for us those years, it's really been focusing and honing in on different aspects of cyber security, running product marketing there, and you don't want to trend micro to run marketing, both of which are antivirus companies. And I steered away and went into more general management roles in cyber security and at Palo Alto networks where I was the senior VP of product marketing for the next gen firewall business. And and then I took a jump and went into more of a COO type of role. And really, I think at that point I realized I, I did not, you know what? I wanted this to be different. I wanted to really challenge myself and get out of my comfort zone. And that's sort of when my search for startups started up again, really thought through about what it would mean to actually be at a startup again. And I found helium or helium found me through a trust live ventures, which is their series, a investment company.And after meeting with his team and I realized, you know, helium is definitely a company that I'm interested in and what sealed the deal for me was my meeting with the cofounder Amir Halane. We realized that he and I have a lot in common growing up. We both love video games. He goes like to tinker with hardware and really the mission and the challenge is what sold me figuring out how to build a ubiquitous networks for IOT devices to connect to that. And that was it. And here we are two years later.[Erik]Yeah. Great. And it's yeah, really quite, quite the team behind helium and quite the ambitious mission as well. So shaping future by rewarding anyone for creating a global decentralized wireless network for billions of machines. So that's just at least your, your mission as it is on, on LinkedIn. Can you just give us the high level, why is this important? Why is, why is the work that Helium's doing important for the future of the internet?[Frank]Yeah, I think the challenge of internet of things, IOT. So for folks that aren't familiar, you know, we're talking little sensors, little devices, potentially autonomous machines that have to do simple work, or maybe it's complex work, but they need to connect not just inside a factory or inside the smart building, but think outside of the building and going outside of traditional where wifi sits. Right? So if you think about extending outside the buildings, the only connectivity option is cellular. It's not easy tooling to create something from nothing using cellular is actually quite difficult. And so we saw an opportunity as a company, we saw an opportunity to change that. And part of that is having an existing wireless network that can compete with the conductivity of cellular, not from a speed perspective, but from a coverage perspective for IOT and just built for IOT.[Erik]Right. And that's one challenge. How do you do that? And then the second piece of that is how do you make it extremely easy and simple for anyone that has a great idea or has an application that they want to bring to reality? How do you make it super easy and simple for them to build on that, on that platform? And usually when you, when you think about super easy, super simple to build in involves open sourcing involves a community of folks with their likeminded that have the same interests, really build a network, using a decentralized method. It involves a lot of, I think a lot of components beyond just wireless technology, but I think the key component is incentive. How do you incentivize individuals, consumers to create a network that's ubiquitous that works for everything. And once you're able to do that, how do you get folks to build the technology and applications using that? That is it, that's the challenge we're in that's the road we're on right now is to create that two sided marketplace of both network operators who are consumers and individuals and network users who may be a simple hobbyists building IOT devices for it could be a large enterprise that wants to deploy millions and billions of machines.And then you're using the helium tokens in order to create that incentive model. Can you give a, just a quick overview of what are the different elements? So you have the helium hotspot, the connectivity device, you have the helium tokens that provide the incentive mechanism. What are the other elements and how do these fit together into[Frank]The comprehensive solution that enables this marketplace? Sure. There's probably three components, three main components to think about one is the healing spot itself. It's a, it's a standard off the shelf hardware we're running on LoRa technology. So it's LoRa hardware and lordships built by syntax or any compatible version of that. And to make that work and to make that run and as low cost as possible, we use off the shelf componentry complementing that LoRa technology. So we have a processor in there. It's a dual core processor. We have, I think is something like, I want to say six gigs of Ram 64 gigs of storage for SSD space on the hotspot itself. And we created a open source protocol running on the Lord hardware. And the reason we had to do that is because our networks are not essentially owned. So we could have used some of the other protocols out there like lower went in because that requires a centralized server for data and key management.We are decentralized because there is a second component in there, which is the blockchain. The blockchain is a homegrown invented by helium from the ground up in our blockchain, similar to big point that most of your audience buys a Bitcoin is where Bitcoin has proof of work for hashing and rehashing data. Our proof is proof of coverage. And so if helium as created a blockchain that can allow the healing hotspot to communicate with other hotspots nearby and create a trusted that word automatically. And that that trusted network involves reliability, health, security of network to prove that they're providing coverage, that the coverage is, is healthy, it's reliable, it's secure. So that's the second component of what we've built and that, and that for the hostile attitude, prove that when they prove it, they mind the cryptocurrency Helion. So this the work they've done, I used to prove that they provided coverage.So as, as a reward for doing that work of proving is providing coverage. They the us let's mine, the cryptocurrency, and then the third component is really the users using the network. How do you get people to build on top of that network? And that's where Helium's open source SDKs and API APIs, and all of our developer kits come into play. It's doing everything possible to make it easy, affordable, and simple, to onboard new applications. And whether it's software or hardware onto the healing network. Folks can take a look at our new designs and our specs and code and build their own hotspots that they want. So those are probably the three main components that build up what we've created.[Erik]Frank, when you look at who's purchasing the helium hotspots, are we talking about small businesses that want to develop their own networks to serve their devices? Are we talking about enthusiasts who are particularly interested in the technology of wireless connectivity and want to be part of this, this trend? Are we talking about well, intentioned individuals like my mother who might not know very much about the technology itself, but wanting to provide connectivity for her neighborhood,[Frank]I call them crypto enthusiasts, people that are into cryptocurrency that probably have a Coinbase account that purchase some Bitcoin or Ethereum. They tend to also be very interested in, in helium. Hot-Spots from a purely from a mining perspective, mining cryptocurrency, and they're, they liked this technology because it's very consumer friendly. We're probably the most, consumer-friendly both IOT wireless gateway hotspot, and a very customer friendly blockchain that kind of takes away all the technicalities of blockchain and cryptocurrency and really makes it a nice user experience. And so those are two early sort of adopters of our technology. The healing definitely, you know, if your mother knows what blockchain is or sorry, knows what big point is. She would be potentially interested in this because the hospital, it's also very low power. It's like five Watts of power. So it's like a led light bulb, right? It's not, it's not going to consume a whole lot of energy that's noticeable.And so from that perspective, it's fairly low cost over time. But it, it allows you, anyone that understands how to use a smartphone and apps to get going. And that's attractive for a lot of folks. That's sort of the profile of our initial adoption. We actually did a survey recently. I think we asked about, I want to say 280 people who've purchased. Hot-Spots like, why? Like, why did you buy hotspots? And it was interesting that overwhelming, roughly 60% of the respondents told us that they purchased a hotspot because they want to build a network. They want to have a sense like that they're an entrepreneur and they can own and operate a network, which, which I think I I'm, I'm surprised, but not surprised because if you think about what we're doing, it's very analogous to Airbnb. We're enabling people to become network operators and everyday folks, average people like Airbnb has enabled average people to become hotel, hotel, hotel management, where they can rent their room or rent their apartment to anyone Helium's doing the same with the network, your home network at home can become something that you can leverage. And the, you know, we're trying to create the other side where others want to use your network and we'll, you know, we'll, we'll reward you for it. So that system, like an Airbnb is what we've done for what we think we've done for telecommunications. Really, really changing that model.[Erik]Yeah. And I, you know, of course for Airbnb people that are not necessarily doing this out of an idealistic, you know, objective, but it's, it's because there's an ROI on, on an asset that's otherwise being under utilized. And certainly that's the case for many people's internet, much of the time, right. Where we're not utilizing it. But I guess this is a little bit early to point to what the ROI is, but is there any indication of what somebody might expect to return on a hotspot that's in a moderately, the medium use case present,[Frank]We don't know, and that's not part of how we sell. It's not, you know, we can't talk about potential future gains and we're not involved with that at all. So it's unknown. And so for someone to purchase a hotspot today, and the ones have really believe in the idea that you're doing this to build a network, you're doing it to become an independent network operator in a very simple and user friendly way. And then the hope is that someday just becomes big someday. The network becomes extremely useful and it's something that's highly valued. And in that world, every everyone in that ecosystem wins. So, you know, it's definitely not for everyone, but you know, there are those that believe that, and we're super happy. They're there with us. Sure.[Erik]Yeah. Well, you don't, you don't need everybody. I guess if you have a couple of hundred people[Frank]That's right.[Erik]Says that a so limited quantity available for a us ordering 77% sold. Are you able to share with us the number that are out in adoption right now?[Frank]Yeah. So if you actually download the app, there is something called a blockchain viewer. So in our app, you can actually see the total number of hotspots that are out there today. And so I think like right now, the exact number of hotspots installed across the United States is 395 hotspots. The predominant cities that are covered are Austin, Texas, San Francisco, Boston, Massachusetts, Chicago, Atlanta parts of Florida. And then we've got I would say another handful, another dozen cities. We only launched in Austin, Texas first. And that was August one. That was our first city that we launched in San Francisco has a nice network because of our employees and friends and family who are doing a lot of the early alpha testing with us. What we've done over since August one is we've started to really test our supply chain and we've shipped hotspots to various customers that ordered really just so we can make sure that our logistics is sound and iron out the wrinkles and correct mistakes, which there are a lot of them it's early days.And so we're learning quite a bit about that, but everything we've done so far is assembled in the United States. We're building everything in San Jose. We're not offshoring this anywhere else. And so we're proud of that for the same time. It's very, very expensive. You're in Shanghai. So, you know, it could be a lot cheaper. We built this in Asia somewhere, but we've elected to do it in the United States for security reasons because our encryption keys and primary injections are happening on site in our factory. And we're going to keep it that way for as long as we can. That's what that means in our next shipment is coming out October 15th. So just in a less than two weeks, we're shipping thousands of units across the United States. And this indicator that you see online today is for a November shipment. So we're probably going to start doing monthly rollouts where if folks order, we should be able to fulfill within 30 days, max probably less. And so right now, because we're shipping in unit volumes of just thousands, not tens of thousands or hundreds of thousands, it's very expensive. And it's almost just in time manufacturing, which is costly for us. And so we try to batch the orders as best we can. And that's what that bar is indicating. So that bar is indicating in November, should change shipment, which is a couple of thousand, a few thousand quantity. And then we have another batch for December, which is a few thousand more.[Erik]Okay. So if you're putting out a few thousand, a few hundred thousand a month, when do you expect to have sufficient? I mean, I guess this is still a city by city, but when do you start to have sufficient coverage where you can go to, let's say a large corporate that maybe they need a solution that can serve them, you know, in a significant portion of the U S and you can go with a viable proposition and say, listen, the network is ready for you. We've got the top, you know, the top 50 cities or 20 cities or whatever. And we have this percent average, 50% average coverage in these cities. Do you already have a kind of a roadmap for when that be?[Frank]So when we ship in October 15th, we will be in 263 cities across the U S and some cities will have more density than others, but I would say 10 to 15 cities are, will be very well covered. And these are the major metropolitan cities, United States like San Francisco Bay area, Austin, Texas, New York city, Bronx Boston Boulder, Colorado, other parts in Texas, as well as Atlanta, Georgia, and beyond. So those are in Chicago as well. So those are kind of the typical major Metro areas in the United States should be very well covered at that point.[Erik]Great. And then if I'm, let's say I'm I'm manufacturing smartwatches, or, or bikes, and I want coverage for my device, then I go to you and what's the conversation? What do I have to install in terms of hardware and software to, to join the network?[Frank]Yeah. So first and foremost, we're open sourcing all of our tech and the open source SDKs and KPI's are coming out. I believe that beta is coming out towards the end of October, that will be a set of tools that developers would need. And then we would have a hardware development kit that's scheduled to be released towards the end of the year. And that would include other wing boards, development kits, reference designs for whether it be a farm dog color smartwatch or, or some kind of tracking device for bikes and scooters. All of those things will be available. We have a lot of stuff in there get, get hub, which will release. We have over 200 developers already that's on the waiting list. And some have already started developing with us in our early alpha. And so all the tools that developers need will be there for them very, very soon within the next couple of weeks.[Erik]Okay, great. Do you, do you have any visibility in terms of what the cost of the hardware might might be in comparison to other, let's say the legacy or the protocols that they're using today, or the, the solutions that they're using today?[Frank]Because we're, we're on a Lora hardware, existing Lora devices that a lot of developers in the world of IOT understand or have played with should, should just work, we'll build on what they will need as long by the long fight SDKs and long by API. So that's zero costs really. They can take their existing hardware and modify it to be compatible with the healing network. I think anything beyond that, it's exploring the latest and greatest hardware, which is off the shelf. So nothing should be too pricey, nothing should, everything should be fairly competitive in price based on volume, but certainly to complain around with the test, it's, it should be pretty simple. We're going to package up dead kids for folks. I don't have pricing on that, but we're going to try to make it as competitive as possible and as easy as possible to adopt the healing network and our open source technology,[Erik]And then connecting the device. And so let's say the manufacturer has adopted the necessary technology to, to sync to the network. The device is now on the street, and it will then pair automatically with whichever helium devices is closest, or how does this pairing happen?[Frank]It's really cool because we're on a blockchain. Every hotspot is running a node, an independent Nova and blockchain. And what that means is the blockchain knows how to wrap the traffic. And so the owner of that sensor does not need to worry about the onboard of the device. They just simply need to ship it out there into the world. As soon as it talks, a helium hotspot that understands long flight will hear the sensor and know how to route that traffic back to its owner, fully encrypted. It's using something. We called an UI organizational unique ID. That will be a private key public key pairing between the device and the cloud database for that particular device owner. And doesn't matter where you are in the world, everything is permissionless and from a wireless onboarding perspective, it's ubiquitous. So there's no concept of roaming cause you don't need to roam. And that's a huge advantage of what we're trying to build, which is it's a public network it's decentralized, it's owned by the people. And if your device is talking long thigh and the healing that data get back to you, regardless of what country you're in, regardless of whose network it is, it doesn't matter essentially is built in. So that, that hotspot delivers the data.[Erik]Okay. Interesting. I guess two topics that come up, one is security, and I know this is your background. And the second is, let's say government regulation. So I suppose some governments are going to be more receptive to having a, a new network. That's not under the control of state owned enterprises, for example and, and others less. So maybe we talk first about security. So I imagine that companies always very sensitive about their data. And now the data is moving through devices that are being owned by a lot of individual. And I'm sure that blockchain has a role here, but how do you ensure that somebody is not hacking into a data stream?[Frank]Yeah, so each device that's out in the world will be called an ECC chip where the private key is an injected. This is a private key. This actually sitting in hardware and only the owner of the device has the public key to decrypt that data. As soon as the data's collected on the sensor and it's transmitted over the air, the hotspot, here's an encrypted packet essentially. And the hotspot does not know the contents of the data. It cannot unpack it. Doesn't have the keys to unpack that data. What it does is it looks at the organizational unique ID. It looks in a blockchain to see where, where the destination addresses and it's an encrypted address. And that address then then the data is sent to that address. And once the address received the data, the credit then goes to that hotspot.So the hotspot will get credited for transmitting the data. And that's part of, part of our mining protocol. The other aspect of, so that's, that's sort of one aspect of security is that that data from collection to motion it through a public network is fully encrypted. And it's never unencrypted until it hits the sensor owners database and their private key is the only thing they can unlock that data. So there's the security, at least in transit is fully secured. Once it hits the customers AWS or Azure or Google cloud instance that then us where the processing of the data occurs in processing of your keys. The other aspect of security is on the blockchain itself because every hot-spots a node on the blockchain, the blockchain is immutable. And so it's a collective peer to peer network where there is an attempt to tamper the blockchain.It takes 51% to overrule consensus. If you have a network of thousands or tens of thousands of nodes, it is a lot of work to try to take over with network. It can be done, but it's very these hotspots and these nodes require you to be in a unique location and you have to prove your location. We can't just be a virtual machine. On some instance, you can't create hundreds of thousands of virtual machines to try to fake the blockchain if it won't work. So we have the, you know, we're sensory resistant. We, you know, we're trying to protect yourself against civil attacks and DDoSs and so forth. So all that, all of that is built in to our blockchain logic. It's native to what we do is that hopefully answers your first part of your question. The second part of the question really is I think less about government philosophies or more about what empowering their people to do, right?Especially in a world of IOT, if regardless of your government mission, government philosophy, if every person under your care and your government in your country benefits because they get a chance to own and operate something in this case for devices, I think the incentive is there where everyone will want to do it and they'll want to continue to do it just like Airbnb and other things. It's, it's good for everyone. I don't know why government would be against it, especially in the world of telco. It's so expensive to build infrastructure. If you can offload that to the masses, give them a, give them, give, give them a piece of the action. I don't know why, how it's hurtful anybody. Sure.Yeah. Maybe the point here is a little bit more that there's, there's very strong legacy players that, and it's a highly regulated industry. It has this been an issue for the other players that the Sigfox and roll out have they been coming up again in lawsuits or I don't know about lawsuits, but if you think about, you know, at and T and Verizon, they have their own approach, their, their supporting things like NBC IOG and LTM. Ultimately we, we, we think those are flawed approaches because of a small little sensor has to pay anything close to three, $4 a month on a data plan to provide somebody temperature information, as an example, that's never gonna work. It just doesn't scale, right? It should be fractions of a penny. And of course, traditional telcos, aren't really interested in that business because they don't have to be there and providing great service for traditional cell phones and iPhones and Android phones. And every, every person that has a cell phone plan is a product right? Every year it gets more expensive. It's a great business. Why leave that business? Why not focus on that farm? That as long as you can, and if there's a budding IOT, you know, emerging market that's happening, we'll let that grow. I mean, I think that's the best, best handled by disruptive technology, disruptive entrepreneurs, perhaps the people and let them grow it.[Erik]Yeah. Yeah. Perfect. I think you're, you're probably right there that this is right now, not interesting to them. And by the time it becomes interesting to them with the landscape will look quite different, right? 5G what is going to be the potential impact? I think there's a lot of, there's a lot of hype and probably some of it is well placed and perhaps some is misplaced, but how do you see 5G impact your business in the future?[Frank]I see zero impact. I think your audience in the world of industrial IOT knows so far, I think 5G is a farce. It's just a lie. There's nothing unique about it. It's, you know, to truly get gigabit speeds or fiber, like speeds on your phone, you need a back haul to carry that through. Right. And so how many, how many homes are there in the world that have fiber back haul? It's certainly not the majority. There may be a minority of that happening in new buildings and new Metro areas, which is great. But for you to get the amount of data you need, right? Imagine streaming, you know,  4K on your phone using 5g for 10,000 people at the same time, that pipe massive, but then you have to be super close to the gateway. It may not even go a block.You have to be within half a block of by gateway to get that kind of. And so what I've heard is one, that's incredibly, that's a lot of hotspots. That's a lot of gateways that telcos will have to deploy. It's insane. It's an insane amount of equipment, which is expensive, right? So again, we're getting back to building infrastructure. If you had to put a 5G gateway and on every block to provide 5G service, I don't know how that's, how is that possible? How is that even possible? What kind of costs is going to incur? Right. It's just ridiculous. And so then you think, okay, well, great. Well maybe it's not every block. Maybe, maybe we go every, every five blocks, every 10 blocks, guess what? That bandwidth goes down to like LTE speeds. Oh, that's what we're on. Right? We're on four GLT or are we that's what the world's on? So I don't know how this is any different. I think ultimately the idea of Fiji unfortunately, is the doing of like guys like Qualcomm, the, the, the phone makers that should have makers, they have to iterate, they need everything to refresh their network equipment. Guys. It'd be the Erickson's of the world that were selling telco equipment. If you don't have a new standard new chip, there's no swap, right? There's no infrastructure change. So there's no revenue. So to me, I think the whole thing is just a game they have to upgrade. But, but in any case, you see this more as them upgrading towards a higher bandwidth towards streaming videos, more efficiently, not, not serving[Erik]Sure. I mean, at best, at best. My understanding was that that that was at one of the propositions. Although I know that the standard is still incomplete. So I think there's a lot of but at least that was one of the propositions was that five G would be also better suited towards low bandwidth devices small data packages and so forth. But yeah, I guess we won't, we won't know if that is the reality for some number of years.[Frank]There's 5G has chip technology that uses low power that we can run on a little, a little lithium battery for my three years. Maybe I doubt. I doubt that's the value prop. I haven't seen that. I haven't seen anything actually other than 900 megahertz. Right.[Erik]Okay, great. No, it's funny. I was, I was having a conversation with somebody earlier, let's say last week, who was just convinced that 5G was like flipping the switch and then IOT was going to all of a sudden BA be everywhere. And I'm thinking generally, when we're talking to people about, you know, making IOT use cases work, this is not the, you know, I mean, they have a lot of, they have a lot of problems, but you know, this is not the necessarily the core one here. Let's talk about one or two specific use cases of maybe some of your early adopters. You'd mentioned ag tech earlier. Maybe that'd be a good starting place.[Frank]Yeah, certainly. So we have a customer located in South Carolina and they're a company that provides telemetry for greenhouse farming, things like vegetables and fruits year round. And what they've done is help their farmers leverage technology and healing. And the healing network by building out sensors can measure soil conditions and air conditions, temperature across all the greenhouses from acres on the land, using a one, one helium HOSPA as the collection point for those sensors, and then routing that data back to their analytics platform to help the farmers better manage their crop yield over time. And so that's, that's a great example of leveraging the technology for, for farming purposes to help improve yield and manage that, manage that land abuse over years. The, the great thing about that is that they don't have to deploy wifi across their property. They don't have to use, you know, cellular, which could be very expensive and there are certain areas on their land that doesn't get cell coverage either.So that's problematic as well. So just really use them a lot of the problems. The sensor typically, depending on use the sensor is very low power. So you can plug a sensor into, into the soil and you kind of forget it for a couple of years and not have to worry. Those are all cool things. In addition to that, they've, they've also been testing a lot of our tracking sensors that send GPX GPS coordinates back to the farm owners. So heavy equipment machinery that's on the farm can be tracked, whether it's in use or not in use. And apparently staffed on farm equipment is decently high, and this there's at least a problem there. And so having an independent sensor that's operating on the battery that can last for awhile is very useful for them from a tracking equipment perspective. And that, you know, that kind of ties into a lot of what we've seen around bike sharing and scooter sharing and cities. I don't know how things are in Shanghai for you, but in the cities across the U S there are tons of scooters. One of the companies that we've been working with is line that is that's servicing 200 plus cities around the world, and they've been testing our, you know, our technology in San Francisco and Austin for some time now on basically keeping their scooters available on the streets for customers to use. And if is because they're highly mobile, they're easy to steal, but they can track them down with our healing them.[Erik]Were they using for connectivity before? I was curious about that, cause they all have some sort of cellular for transactions. The problem with the cellular implementation is they're mostly all, all these guys, all these scooter companies, sharing companies are using the cheapest possible cellular available. So that's likely to G in a lot of, to G in the U S certainly is getting decommissioned. So cartridge is really poor and it's not great. And so having helium as a backup, sort of as the, as the aide or the LoJack, let's say for scooters has been really helpful for them.[Frank]Yeah. Well, I know in China, some of these bikes were being manufactured for something like 40, $40, you know, CapEx. So, so I imagine even if there's a $1 a month subscription on, on the, you know, I don't know what that would actually look like, but that's already a significant cost for them to bear. So I'm sure they're also happy, happy to find an alternative, how many devices can one helium hotspot, potentially service.We don't have real world numbers because we have fairly good density across cities, but I imagine it'd be probably like 300 to 500 a burst. So remember these are not sustained. They're not sustained connections. So there are bursts of data that are sent. And so depending on application and use, if it's simultaneously, I'm guessing three to 500 sensors talking exactly at the same time, exactly the same nanosecond is supported for hotspot. But if you have a couple of hundred hotspots in the city, you can sort of multiply that out, decent, decent coverage for the number of devices on it on a moment in time.[Erik]And if we had a device like a, a smartwatch, which was not just sending kind of simple bursts of data, but was, you know, maybe streaming audio for a conversation or something, would that be feasible or, or is that required too much bandwidth for the device to manage?[Frank]I would say likely not a good idea only because for us to achieve, you know, a hundred square miles of range, the data packet has to be small. So we're talking 24 bytes to 20 kilobytes at the high high end. So I would argue voice is probably not a good application. However, text messaging or email limited email could be fine. And so if communications is, is the application of need, certainly you can create some, you know, old form pager system or old form, like the old blackberries text, each other simple things. Gotcha.[Erik]Gotcha. Okay. But so really designed for IOT machine to machine[Frank]That's right. That's right.[Erik]That we're missing here. That's a, that's critical for people to understand it.[Frank]I think probably questions I get the most when talking to potential folks who are interested in using the network is network coverage. And so I think maybe just reiterating how fast we're creating coverage is important. August 1st we shipped Austin, Texas, and within, I would say four or five days, the entire city of Austin was covered October 15th, we're shipping to 263 or 64 cities. That means within four or five days, there will be 264 cities with healing coverage. And that's the days, right. Days, worst case couple of weeks. I mean, that is amazing. So if you think about it, that is bringing a network to market in lightning speeds, that's never been seen before. I think, I think that's the power of the helium incentive and leveraging the people that create this is that that incentive model creates a hyper speed for deploying infrastructure that's useful for enterprises and industrial IOT essentially. And so that's something to think about. That's a hard idea to, to understand, because it's hard to believe that there are people out there, thousands of people willing to pay $500 for the hotspot for the sake of mining crypto, but you know, it's a whole new world out there times are changing[Erik]Well, Hey, I, I believe it. I mean I think finding 10,000 people to pay you know, a bit of money for, for a new technology and, and ownership of a new business model, I think that's that's something that is very feasible today. The challenge for you then is, is going to be to engage in with the the device manufacturers device operators and getting them on board.What would be, if somebody is interested in exploring how their device use the network, what's the best way for them to get in touch with helium or, or to inform themselves around what the steps would be?[Frank]Sure. Yeah. Check out our website, helium.com. There's a number of, I think, great resources on there are helium.com/enterprise talks about the use cases that we've already share. Some of those also, if they're a developer and they want to develop devices to operate on the healing network, check out helium.com/developers. There are great resources. Certainly. Is there a focus on listening to your show that are interested in buying the hotspot.com/store page? Take a look. And if anyone's interested, just send me an email@helium.com or sales@helium.com. And we're happy to talk more about how they can onboard devices to network, or you want a discount on the hotspot, contact me. I'm happy to share a promo code for them to use.[Erik]Okay. Awesome. Well, I really enjoyed the conversation. Thanks for taking the time to brief us on what your ability to hear. I think this is, incredible. I mean, we really do need new solutions to make connectivity more affordable and simpler. So I wish you all the success and really appreciate your time today.[Frank]Thank you, Erik. Appreciate it.[Outro]Thanks for tuning in to another edition of the industrial IOT spotlight. Don't forget to follow us on Twitter at IoTONEHQ and to check out our database of case studies on iotone.com/casestudies. If you have unique insight or a project deployment story to share, we'd love to feature you on a future edition. Write us at team@iotone.com." />
EP055 - How to cut 20 minutes off an emergency 911 response with data - Michael Martin, CEO, RapidSOS
Monday, Dec 02, 2019

In this episode of the IIoT Spotlight Podcast, we discuss the challenges facing 911 IT systems, the impact of data on emergency outcomes, and the role of IoT devices in enabling first time responders and predictive alerts.  How much data does a 911 center have?How do we know which data is useful to a 911 first responder for deciding actions to take?How can software and device companies use their data to increase the safety of their products and services?" />
EP 054 - How removing the battery from sensors can enable a trillion sensor world - Bob Nunn, CEO, Everactive
Wednesday, Nov 06, 2019

In this episode of the IIoT Spotlight Podcast, we discuss the battery problems in IoT that hinder deployments and value creation today, and how innovations in ultra low power circuit design and wireless communications that allow sensors to operate solely on energy harvested from the surrounding environment.Are we trading one problem for another with IoT sensors?How do we create a sensor the size of a stamp with larger range and compute ability than the average sensor today?How accurate are consultancy forecasts about the size of the IoT market? Bob Nunn is the CEO of Everactive. Everactive produces data intelligence for the physical world. Operating without batteries, the company’s always-on wireless sensor networks deliver continuous cloud-based analytics at a scale not possible with battery-powered devices.  Everactive’s end-to-end solutions are built upon groundbreaking advances in ultra-low-power circuit design and wireless communication that allow the company to power its Eversensors exclusively from harvested energy. For additional information, visit www.everactive.com." />
EP053 - The first iteration of cybernetic workers - Gabe Grifoni, CEO, Rufus Labs
Monday, Oct 28, 2019

In this episode of the IIoT Spotlight podcast, we discuss the evolution of warehouse for commerce, and modern solutions and flexible service based business models to support the new business and competitive challenges in logistics.  What is driving the development of industrial wearables?How do you sell Hardware-as-a-Service?What are the disruptive technologies to come in the IIoT?Gabe Grifoni is the CEO of Rufus Labs. Rufus Labs’ WorkHero is the most advanced connected operator platform for enterprise. Comprised of machine learning software, rugged wearable technology, and superhuman support, WorkHero is a complete Productivity as a Service solution for the evolving workplace (Industry 4.0). Compatible with existing WMS/ERP systems, it connects workers to each other and to automation, allowing for optimal coordination between humans & machines. WorkHero provides our logistics, e-commerce, and fulfillment customers superhuman efficiency gains, safety enhancements, and zero downtime, while providing management with robust, actionable, visibility & analytics into their teams and facilities.For info and demo: getrufus.com or enterprise@rufuslabs.com " />
EP052 - Empathy is the key to successful industrial design - Gordon Stannis, Director of Design and Strategy, Twisthink
Tuesday, Sep 17, 2019

In this episode of the IIoT Spotlight Podcast, we discuss human-centric design and change management in high uncertainty innovation processes. What is human-centric design?What is empathy in an industrial environment?How to manage and communicate through the process of trials and failure?Gordon Stannis is the Director of Design and Strategy at Twisthink, A business growth accelerator / innovation firm that partners with clients to imagine, develop and launch exciting new user experiences and products that “their” customers will love! For nearly 2 decades the Twisthink team has consistently unlocked game changing opportunities for clients in a broad range of markets including home, transportation, workplace, healthcare, sports and many things in between. Because at Twisthink, they’re not tethered to industry paradigms or mired by the day to day constraints of those industries. They flow between industries and can spot consumer, design and technology trends that affect them all. They’re entirely devoted to creating the ideal user experience which allows their customers to become market leaders and grow.twisthink.com/iotspotlight" />
© 2020 IoT ONE