Erik: 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, Erik Walenza, CEO of IoT ONE. And our guest today will be Bob Nunn, CEO of Everactive, Everactive produces data intelligence for the physical world based on wireless sensors that operate without batteries to deliver cloud based analytics in situations not possible with battery powered sensors.

Together, we discussed innovations in ultralow power circuit design, and wireless communications that allow sensors to operate exclusively from energy harvested from vibrations, indoor light and other environmental sources. And we explored how new generations of batteryless sensors can enable a trillion sensor world.

I hope you found our conversation valuable. And I look forward to your thoughts and comments. Bob, thank you so much for taking the time to speak with us today.

Bob: Erik, it's my pleasure. I'm happy to be part of the podcast.

Erik: So Bob, you've got actually really fascinating history. So before we get into Everactive and your technology in your business, I'd like to learn a little bit more about you personally.

Bob: I think if you look at my career, it's been 30+ years of growing startups. And there's kind of a theme to what I've done during that period of time, which I would say is applying breakthrough technologies to real world problems. And Everactive is a great example of that situation.

Actually, my first startup, when I joined, they were probably calling themselves the gas company, which you could imagine the maybe you were distributing gas or something like that, but they actually had developed a really novel gallium arsenide semiconductor process. And of course, gallium arsenide is abbreviated GAS. And the idea was that the market would just come to them because they had this novel technology. So from that very first startup, I realized that no cares that we’re the gas company, they care of what we can do with these gallium arsenide process technologies that's important to them.

And what we ultimately ended up with that company was making it a company that helps people serialize and deserialize data. This was back in the 80s and early 90s when the internet was becoming a big deal and this idea of being able to move data very quickly was just starting to grow. So we really became a data transport company, had nothing to do with the material except the properties that the gallium arsenide material offered to the components that we put out. But that's a good example of where a technology may start as a great idea. And the entrepreneurs think, oh, well, I've got something really interesting. But it really doesn't matter until you find a real world problem and you apply the benefits of that technology to that problem.

Erik: And I want to get into a little bit later in our conversation some of the problems that you're focusing on at Everactive because you've identified some very specific use cases, whereas in particular, a lot of startups that I look at, often have very generic domains that they're addressing. But I can see that you've maybe brought those learnings forward and really highlighted the use cases that make sense. Maybe you can first brief us on the story behind Everactive because this is a technology that I think actually many people maybe people listening are not aware actually exists in the world today.

Bob: And actually, I think I've Everactive is a collection of breakthrough technologies. So it's actually a number of technologies coming together with a very specific purpose to remove batteries out of the electronics devices, in particular the IoT devices that we're starting to use more and more of. And actually, my favorite story about the company was actually started by two professors, one at the University of Michigan and the other at the University of Virginia. But their history together goes back to graduate school at MIT, where they both were studying low power electronics.

But as they were going through their careers as professors and they were collaborating on different things, they started hearing and maybe six, seven years ago, they started sharing forecasts that the IoT was going to have a trillion connected things within the next five years or something crazy along that order, and they just simply did the math and they said battery life, where it is today, if we assume that it increases dramatically, let’s say we can get 10 year battery life on average for all devices, then if there's a trillion things connected, we would need to change batteries 275 million times a day to keep up with all the trillion tanks.

And they said, the IoT is not going to take off until someone figures out how to remove the battery. And that was really the genesis for them wanting to start a company was let's take what we've done in low power electronics, and let's go apply it to what we see as the thing that's going to hold back the Internet of Things, and really focus on what's it going to take to remove the battery from the devices completely.

Erik: And then you joined in May of 2017, is that around the founding of the company, or had they had maybe a research phase of a couple years before bringing you into to upskill?

Bob: The professors spent a couple of years just taking what they had done at the University and transferring it out into a company, they actually did their first call institutional funding in 2014 and they started to build a team at that point. But just a few people early on, and like you said, it was really technology development and proving out their ideas. They wanted to create a semiconductor device that had all the components needed to support an IoT sensor that was 1/1000 of the power of traditional semiconductor technology. So, as this crazy goal that they started the company with 1/1000 of power, and that first few years was really around proving that they could do that, that the innovations that they had developed, both as graduate students and then further as professors, could lead to that kind of breakthrough device.

And so when I joined, two years ago, they had proven the technology. The technology was already proven, and they had actually honed in on the industrial market as the first place to apply the technology. But they hadn't really developed a product yet and they haven't really developed a business model. They just had a concept that in industrial market, people are already using wired or battery powered IoT devices so they don't have to teach people that it's important. But there are a lot of things, a lot of assets within the industrial environment they're going unmonitored because people are unwilling to put batteries. It's just not economically feasible.

But with steam traps, there's just too many of them; currently, they're manually inspected. And if you put batteries on all of them, you'd have that $275 million or a million battery a day problem, you have to go around changing batteries as much as you expect them. So, you're trading off one maintenance problem for another and it doesn't make sense. So removing the battery opens up this whole new category of things that you can go monitor in the industrial environment. But my job really was to come in and say okay, with this technology, how do we go apply it to the market? What's the right first product? What's the roadmap? And then how do we commercialize that? What's the business model?

Erik: So who are you addressing primarily today, we're looking at your customer segments?

Bob: Actually, we chose to go after an asset as the approach that we're going to take. And one of the important things to know about Everactive is we quickly came to the conclusion that removing the battery requires more than just a low power chip, it's today at least to remove the battery means you have to have a full vertically integrated solution, and what we like to call a full stack solution. The site goes from the chip to the sensor itself to the network on how the sensors are connected to the way data is managed and where compute is done within this either at the sensor or in the cloud, and as well as the analytics that are applied and how the data you collect and how you apply the analytics to that.

So given that they had determined that to solve the battery problem today you have to have a full stack solution, we decided really what we should focus on is the data itself. And what we like to say is that we're generating a new class of data streams that were previously unavailable because people were unwilling to put batteries or to wire up sensors. And the real key is for us to create actionable insights from those physical world data streams that are valuable to the customers. So because of that we actually came to the conclusion that the right business model for our company was to actually sell those insights, take that data, turn them into insights and sell them back to the customer.

And so we picked one asset, in this case the first asset that we decided to go after was Steam Traps, and we decided to apply that full Stack solution to that problem and then go to a number of customers across industries and say, hey, let's go address your Steam Trap issue. So, examples of industries that were already involved in today include food and beverage, consumer products, pulp and paper, pharmaceuticals, chemical, oil and gas, as well as we like to call district energy. But it's really the heating and cooling using steam in those applications as well.

Erik: Is that mean that you then have an optics pricing model? So is this that they pay for a monthly or an annual fee based on the insights that you're providing, and that you would then deploy the sensors and own the asset? Or how does this model look like?

Bob: The most important thing about working with industrial customers is that they do want to own the assets. In most cases, if it's in their factory, they want to own it. And they definitely want to own their data. So there's no question about whose data it is. You just have to start with the assumption that it's the customer's data. And if you're going to put an asset in their factory, that it's going to be theirs, whether you're charging for it or not. So we started with that premise and we said, okay, let's do annual subscriptions. And so the idea is if we can go in for the first year, provide the equipment as part of the service and you're one demonstrate value, then there will be a number of years of follow on revenue stream for us from that engagement.

So it is what we like to call insights as a service sold on an annual subscription. We've only been in the market and so we’re less than a year about nine months. And we have 15 customers or so today that has signed up to that model.

Erik: So you begin generate revenue around a year ago, are you able to share any data either around your revenue, your gross adoption? I guess, 15 customers, you're early enough that it's probably still a little bit hard to see what the long term trend is, but any anything that publicly are able to share around your growth in since January?

Bob: Actually, so I can share that our original goal was we would have 10 customers by the end of the year. One of the things about the industrial market is the pace of adoption, and the risk averse nature of the customers has to be considered. And so you have to go into the market knowing it's going to take patience and perseverance to first get in the door, and then demonstrate your value before they really start to scale. And patience is not something you normally associate in particularly with startup investors, but it really is a requirement in going after the industrial market.

Big payoff if you have the patience and perseverance to make it through and build the relationships with the customer. But it does take time. So our model was based on that, it's like let's go find a few great customers, go build relationships with those, and then we can sell them more of the steam trap monitoring systems either within the same factory, or in most cases, these Fortune 100 companies have multiple factories that do similar type things. So we can move from one factory to another.

And then more importantly, our plan is to continue to innovate and generate new products going after new assets and then we start adding new assets within the same factory. So we actually expand across multiple vectors with these great customers. So the original plan was 10 customers, we're about 15 now, hopefully, we'll double our initial projection for the year. Most of the deployments are small. We have just within the last 30 days deployed at a, say the customer’s name, but I can say that it's a pet food factory. And it's the first time that we've deployed Steam Trap monitors across the entire factory.

And that's really our goal, it’s pervasive sensing, you don't just do a few of the high volume, high value assets, you do all the assets. And we really have tried to create a pricing model that is compelling to say, yeah, it makes sense if I'm going to do this for one, I'm going to do it for all. And so we're really proud of that one that it was our first full factory deployment. Otherwise what we're finding is most people are deploying a few 100 sensors trying them out for a while, and then again building this relationship and trust, making sure we don't break anything Mistakes are very expensive in these manufacturing plants, so making sure that there's nothing that goes wrong before they start to expand out and deploy more.

Erik: I suppose with Steam Traps, it sounds like that's a use case where current sensors are not a very viable solution and so the solution are people. And so you have providing a fairly radical solution for that particular use case. If you're addressing the whole factory, I suppose now you're competing head on head with more traditional, either wired or battery powered sensors. Can you point to any ROI if you're comparing a batteryless to a battery, or wired sensor, and then what would be the components there? Because I suppose the cost of the battery itself is probably not critical, it's more the labor associated with changing the battery, maybe that would be the critical factor here.

Bob: And actually, it's both. So the biggest competition for us is the old way of doing things, the way it's always been done, which is mostly manual. I'd say 98-99% of all Steam Traps are monitored manually. There are automated battery powered Steam Trap monitors available in the market today and they had that for some time. So it's not a new concept that it would be great if we could monitor these in an automated fashion.

The issue comes in that there are so many of them. You have this battery problem that you're not willing to trade off putting so many batteries that you still have to send out a maintenance guide to change batteries on a regular cadence, the same way you had to do to just go inspect them. So that's probably the biggest thing we run into, is that a refinery can have tens of thousands, so I'm not going to put 20,000 batteries out in my refinery, and then have to send guys out going to change batteries either proactively.

It looks like the smoke detectors in our house, there's only a small handful of them. But when they start beeping, it drives us crazy and we have to find a ladder and go change them. Or we can be very productive and start changing our batteries every six months or a year. So that's what the factory managers look at when they look at battery powered sensors is that there's a maintenance schedule that goes along with that.

But to give you some economics on current solutions out there, so again, we are selling as a service and the hardware is included, so the upfront cost to the customer is really the cost to install the sensors themselves, which of course we are working diligently to make as low as possible, as easy as possible to deploy the solution. With some of the battery powered solutions that are out there, there are hundreds if not $1,000 in capital costs upfront to buy the sensor.

And then the batteries because in many cases, they're going into very difficult environments, some case hazardous locations for batteries as an explosive elements have to be contained, and there's very elaborate enclosures that go with them. The batteries in some cases are two $300 a piece, and with battery life's in the order of a couple of years, two to three years. So the economics of that solution has been looked at and discarded and said that doesn't solve my problem, I'll just keep manually inspecting them.

But when we come in with this new business model with a compelling annual subscription fees, and hey, we'll monitor all your Steam Traps. Do you think it would be useful for the listeners to actually talk about what steam traps are? Do you think we're all up to speed on Steam Traps?

Erik: No, I think that would be useful. So Seems Traps and then Flare Systems are another use case that you have prioritized on your website. So maybe you can explain a little bit about what those two systems are and then also why they are the right asset class for a first entry point?

Bob: We get it often from investors and others of why do Steam Traps?  What was the reason that you pick that? I do think it's a great example of how removing the battery unlocks customer value. So just as we talked about, you're not trading off one maintenance problem for another. You don't have this ongoing battery cost. You really are able to address a known operating efficiency and then something has been standard operating procedure for decades. So it's kind of a radical new approach to a really old long term problem.

So what it is actually a steam is a major part of our industrial world. It's actually been used for over 100 years and really powered the industrial revolution. But even today, there's the Department of Energy reports that 30% of the total energy used in industrial after cations is used in generating steam.

We use a centralized boiler system and then distribute the same steam in pipes. And one of the problems that was discovered on decades ago is that as the steam is used are cools, you get condensate in the lines, and that can make the steam system less efficient, it can damage equipment, and even in worst case scenarios, it can cause explosion through something called water hammer. So many, many years ago, they developed a valve to remove the condensation out of the steam distribution system. So instead of calling it a steam, or a condensate relief valve, they call it a Steam Trap.

The problem with Steam Traps is that they are in a toxic environment, and there are mechanical device of some sort. There are different types of steam traps. But generally, there is lots of data to show that Steam Traps fail on a regular cadence about once every five years or so. About 20%, I think another DoE figure is 20% of all Steam Traps fail each year. They can either fail in a state where they're open and they’re called Blow Through for the Steam Trap is effectively doing its job, is removing the condensate from the steam lines. It's like having your air conditioner on with the windows open. So you're just wasting energy through this open window. Or they can feel closed in a case where they're not doing their job and you have all those risks just associated with before they invented the Steam Trap.

So we actually based our economic case on just the open. And the reason for that is as we talk to people about, it becomes very controversial on how do you value a piece of equipment or a downline or even the injury associated with an explosion. So because it wasn't generally [inaudibke 21:56] so right, let's just go look at this blow through this open case.

So the Steam Trap, actually, beautiful because it's easy to value the energy lost in that case, and it's simply a measure of looking at the pressure of the steam, the cost to generate the steam, the size of the orifice of the steam trap where the steams blowing through and then how often you inspect your steam traps. With those variables you can go in and you can actually calculate if I was to inspect the steam traps every minute, and I was to replace them when they failed, I would be able to save X number of dollars.

So business case that we like to use is from a chemical refinery one of our customers, and they have 3,800 steam traps. And using their own calculation using their own numbers for the cost of steam and the pressure and so forth, they calculated that they would save about $5 million a year if they address in case failures blow through failures immediately. We created a business model that actually provides them a payback in less than six months. So they go in, can pay for the first annual subscription and all the installation costs within six months from those savings, and 3x ROI over a five year period. So it's really a compelling economic case. And that's why we've gotten so much attention from such great customers.

Erik: Sounds like the status quo was checking them once or twice per year, what did you arrive at as a reasonable cadence? Are you able with your technology? Actually, I'm curious maybe as a separate point because your technology is barely less, and then driving energy from the environment, how frequently you are able in an optimal situation to collect and transmit data, but maybe that's a separate question. But what did you determine as the right cadence here?

Bob: I actually had a development partner on Steam Trap monitor and it was a consumer products good company, one of the top Fortune 100 companies. And their original thought was we check these once a year right now. If you just checked it once a month, that would be great. Why don't we check it once a week? And then pretty soon, we said, we can do it every day, we can do it every hour. We got it down and said why don't we just do this every 10 seconds?

And the beauty is as you generate more data, you start to see new insights. And one of the things that is important to us is to go from individual assets to entire environments. So in this case, it would be going from the Steam Trap to the steam system to that whole distribution system. And already with the data we've collected, we've started seeing things like equipment interaction with the boiler causing it to basically over-boil for a short period of time, but on a regular cadence where you're wearing out the boiler.

So I was at one of our customers where we saw the spike in steam temperature, we didn't know what it was. We went from say, hey, what do you think this is? And they say, oh, I don't know, but that's not. That shouldn't be happening. So they figured out it was a reaction with another piece of equipment. So, that's something you would never find if you were doing it once a month, or even once a day because unless you checked it at the right time.

So, the beauty is with our battery list technology, but our devices are always on, always active, and thus the name Everactive. We have the ability to drive down to very fine grain measurement levels. And actually, our devices aren't simply just collecting energy, and then take measurements when they're ready. We actually have energy storage in our sensors, and we proactively pull the sensors in and ask them to report back. So they're always on, always available, always ready to do whatever the job that we set them to do.

Erik: Let's take a few minutes now and go a bit deeper into your tech stack, maybe starting with the power system itself. So my intuition initially was that you would be able to collect data, maybe once a week, maybe once a day, but you'd basically have to be generating energy for some significant period of time in order to collect data and transmit, it sounds like that's not the case. What is the underlying technology that allows you to have a fairly high frequency of data transfer?

Bob: Actually, what you describe is exactly the way many people are trying to solve the batteryless problem. They're simply turning the device off, letting it collect enough energy so that it can do its function and then wake up and transmit, basically just seeking out the answer, hoping someone's listening and then going back to sleep until they collect enough energy to do it again.

So the key breakthroughs for our technology really are at the chip level, and it's going back to our two professors. And it's really across a number of innovations, most important of which are the receiver. We've actually separated the receiver from the transmitter in our radio, and the receiver, we've lowered the power to such an extent that it is always on. So if you're familiar with watts associated with the light bulb, and lots of watts associated with your computer, we actually are working off of nano watts, a few 100 nano watts for our receiver radio. So it's an amazing innovation. It was a huge breakthrough. This is some of the research that the company got started with.

The other thing that the company has done, and this was, again, in association with this 1,000 times lower power goal, we have actually readdress the way that digital electronics is done. And you'll find other chip companies starting to use this technology as well, is called Sub Threshold Digital Processing. And the basic concept is transistors are like a light switch, they're either on or they're off. When they're on, they're in the one state and they're off in the zero state, we can do useful things with that binary action.

But the reality of the transistor is that when it's off, it's actually leaking current. So it's not completely off. It's like a faucet you turn off that's still dripping water. And the interesting thing is as you go, you turn transistors on and off with voltage, and as you drop the voltage further in the off state, you get a different leakage current. It actually goes down further lower than voltage goes. So we're actually doing useful work when the transistor is traditionally considered off. And it's kind of like washing the dishes with that little drip of water from a leaky faucet.

And that's so that subthreshold digital design technology we use both for our processing elements as well as for another key element, which is how we interface to the energy harvesting unit, and then how we store that energy and have it available when the device is called upon to do something useful. And so that power management, you'll find those and our phones and many other IoT devices that we've specifically designed ours using a subthreshold design methodology to be focused on using harvested energy and effectively harvesting from multiple sources being able to put that in storage on it and then have it available when needed.

So those are three of the key innovations. Of course, putting it all together on one ship is a big thing, is from any chips company you to talk to the integration, what they call the system on a chip is a big part of the design. So you can have low power components, but you won't have enough with a low power chip unless you put a lot of thought into how to do that. So it's really that the low power receiver, the subthreshold voltage, the energy harvesting PMU, and what we call the Uncor, the term we stole from Intel on how everything gets put together that really makes up. And we did hit that goal of 1,000 times lower power than traditional chips that are up in the world today.

Erik: And then the energy that you're harvesting, you mentioned that you can come from different sources, is this third party or partner technology that you're using to harvest the energy, or was this also designed internally? So what are you harvesting and whether this is maybe commercially available technology or something you built up? And then the second question, as a follow on here would be, how much energy is needed? Are there some environments that are more energy rich where this functions well, and are there other environments that maybe don't have the energy sources, so signals and so forth that background radiation that you would be harvesting here, and are then more difficult to operate it for that reason?

Bob: There are lots of energy sources available, but there's really four main ones that we can draw from. And they have one characteristic in common with the exception of sunlight, which is very different. But once you come in doors, we look at indoor lights, thermal differentials. So like in our case of the Steam Trap monitor, the difference between the pipe that's carrying the steam and the air around it, you can also look at vibration as an energy source, and then RF signals as an energy source.

The one characteristic that they all have in common is very small amounts of energy. And you're talking about tens of microwatts per square millimeter of whatever material you're using to harvest energy, whether that's a PD cell, or thermal electric generators for the temperature differential. And so, that you're dealing with micro watts of power that you're getting from your source and traditional electronics, even traditional low power and we talked about BLE as a low power radio technology; even when it's turned off, it's consuming more than that in power. So once you turn on, it's lower watts of power, so 1,000 times more power.

So this small amounts of energy that you get from indoor sources, that's been around for a long time. So there's nothing a novel that we've done there except look for and partner with the best companies. And there is a lot of innovation going on in the harvesting space, and we are working with them. Unlike once you go outdoors and you get the energy from the sun for us, that's a massive amount of energy. Outdoor life is much better source of energy than anything that we can get indoors.

So actually, it brings up a good point that we have developed the electronics for managing those energy harvesters, we also manage all the sensors. So we do not do our own harvesters, our own sensors. That's not part of the innovation of the company. But we do bring that all together. And we do work closely with those vendors to make sure that the interfaces are low power, and that the same way we're obsessed and our passion is about removing battery from the sensor that they work with us to take out the big power consumers out of their product as well. So we have basically Everactive certified set of energy harvesters certified sensors that we work with. Then we continue to look for new vendors, and new breakthroughs in those fields.

Erik: Well, I'm sure the energy harvesting companies are happy to see you enter the market because other otherwise they have a technology that doesn't by itself provide the solution, right, they have a component?

Bob: They have a lot of devices that are off 99% of the time. The key here is that because of that low power receiver, our device is always listening, always active. And actually, I should mention, the bigger the temperature differential, the easier it is to harvest energy. So the steam pipe has a ton of energy for us. So we actually, for our second generation platform, you mentioned Flare Systems, and it's a derivative product off of the same platform that we're doing the steam product.

But we've actually developed a new generation, a new chip, new capabilities that we'll be releasing the first part of next year. And in that platform, you can actually run off the heat of your thumb. So you can put your thumb on a thermoelectric generator and generate enough temperature differential for us to prop up the device.

Erik: I'm curious about the building or the infrastructure use case, because I've talked to a number of companies that are somehow interested in embedding sensors in infrastructure, whether that's a building a bridge or road to determine the state of that infrastructure, and when it might need to be repaired, or whether it's damaged. And then I suppose you have maybe in a bridge, you do have some significant metal vibration and ability, and probably less, so you probably have some energy constraints there. And also, we haven't really talked about the transceiver yet and the receiver, and then how far you're able to broadcasts. But I assume that in some situations, that can also be a constraint on the use cases.

Bob: The transmitter, I am challenging the team to figure out that they insist that there are physics laws that we cannot break to reduce the power of the transmitter further if we’re working on it. But the power of the transmitter is directly proportional to the distance and the quality of the signal and the amount of data they transferred. So out of our sensor itself, the largest power consumer is the transmitter.

And so a lot of what we've done, we've been using our subthreshold digital design techniques to put processing elements inside the sensor itself so that you don't have to communicate all the raw data out on the receiver. So it's really understanding what can you process locally, and the tradeoff of doing computation that again, also harvested energy but doing it locally, versus sending out all that raw data on a very expensive transmitter. So that's another big part of what we do is being able to adaptively figure out where is the right place to do compute.

So, in the first product on our second generation platform will be focused on motors and monitoring the health of motors. You take a ton of data: vibration data, electromagnetic data, temperature, humidity, all these data has being taken at the motor itself. But we can actually turn that time series data into frequency domain data, and then look for the key parameters and then just communicate those out. And actually through analytics in the cloud, we can get a good picture of what's going on with that motor, and be able to detect when something’s happened that's causing motors not running as efficiently as it was.

So in that case, we can then send a signal to the sensor and say, okay, send us or start collecting raw data, and then transmit that out through the receiver. And of course, we can only do a certain amount of raw data because of the cost of transmitting that to the receiver. But we have the ability to take snapshots in time and say, okay, here's what's going on, and you can then do more advanced analytics in the cloud, based off of that raw data. So you kind of adaptively learn when you can use the KPIs and when you need the raw data itself.

Erik: How many meters would you typically be constrained to right now?

Bob: So that is a key parameter for us. So a couple of key innovation vectors for us are really putting more and more processing on the node, eventually doing inference and machine learning type things at the node level. The other one is to increase the range of the radio.

And the key thing you hear people talk about radio ranges, and there's actually different like all things. You can talk about what happens in free space where there's no obstacles, and life is good, or you can talk about the application itself. So in our case, we're talking about themed jungles of types, and brick walls and all sorts of obstacles. And so when we talk about range, we're actually talking about within that industrial environment.

And so for our Steam Trap monitoring product for generation one, we get about 30 meters or 100 feet of range between the sensor and the gateway that it communicates to. But for Gen two, for the this motor health monitoring product I mentioned, we have increased that tenfold. And again, it's innovation vector, we're focused on to say, hey, we need more range. And for Gen three, we're already working since Gen two is about to come out, the team is already working on Gen three. And the idea is to take it up to a kilometer, so to really take this concept of a self-powered or environmentally powered component and be able to transmit and receive a kilometer away.

Erik: But let's return them to the building. I'm sure this is something you've discussed internally before.

Bob: It leads to a great topic, which is we are startup where just pass through, we'll probably be around 60 people by the end of the year, and we cannot do it all. So the applicability of a batteryless data collection system is huge. It goes everywhere. And so we get a lot of major companies that are approaching us saying, hey, could you put this in the construction space? Or could you go after this asset? Or could you solve this metro problem or many, many different things? And the answer almost invariably is yes, we can. The technology will support that. And in particular, as we continue, as I mentioned, with the increases in range and compute power we can do more and more.

So, construction and infrastructure is one that we considered ourselves. But more recently, we've started working in partnership with one of the major construction companies and really looking at some very innovative job site sensors that would make a huge difference. We don't yet have a partner in actually embedding the sensors into construction. You're talking about into buildings and things like that. It's definitely something that could be done. Again, you'd have to work through all the radio and harvesting issues, but it's quite possible.

And I would say, today, again, we currently support so many partnerships, but there's probably 15-20 companies that have approached us and said, wow, I really like what you're doing. Could you apply this technology to our problem?

Erik: So it sounds like you guys have done a good job of staying focused so far, no doubt, you'll have the continued challenge of deciding where to focus going forward. Let's take some time now then to deep dive into one of these case studies. If you can, and whether it's the pet food manufacturer that you mentioned earlier, or another case that's already been deployed, and walk us through from initial discussions through deployment through operations, and just give us a clearer picture on what this looks like, and then in particular, maybe what are some of the concerns that companies have that you discuss? And maybe what are some of the challenges that you've faced and addressed as you're figuring out the right deployment?

Bob: Our strategy has been to prove that it works and to show people what's possible. And the Steam Trap monitoring device is a great example. Well, we can find steam experts, go talk to them, explain what we're doing, and get in the door and have a meeting with customers. And I should mention that the first 15 customers have all been direct sales from our team. We actually just hired our VP of Sales about a month ago. So it's really been the business development myself and the founders that are out there trying to get the attention of customers.

But now that we have gotten their attention, we are getting the attention of some of the asset manufacturers, so like Steam Trap manufacturers are coming to us and say, hey, what you're doing is interesting. Is there something we could do together? They may have a software tool that they would like to integrate with our data acquisition capability.

And that's also led to partners in the sales channel. So that's been a big part of our growth over this last six months is adding sales partners, so whether that be sales representatives, or value added system integrators, even service organization. So I mentioned the installation, one of the things we like to offer to the customer is the ability to use an outsourced service organization that they already know and is qualified in their factory to do that installation for them.

So that's been a big focus that's part of our growth now, building a real sales team bringing in a sales channel, and that will help accelerate finding customers. And these are complex, large organizations. So part of it is finding the right people, and then navigating through the complexity of making the sale. So you have to kind of walk through what it's like.

So in the initial days, we would be going to conferences, or we would be using email campaigns to find somebody at one of our target customers in one of our target industries to start the conversation. It's a compelling conversation because people know, the Steam Trap problem it's not new, they know that it's been an accepted inefficiency for years and years. And then we can say, hey, you could address that and you can save a lot of money with our solution. So we get in the door and we can have a meeting, may or may not be with the right people, most times not.

And it's something where you go, okay, who in the organization are the right people for us to talk to? And invariably, for a solution of this type, you really need to talk to the people in the facility. So a lot of times that's the maintenance director, or even the facility manager. And a lot of times the business unit executives themselves will get involved because they're for the cost of steam in terms of going into making their product and the benefit of saving that really hits their P&L.

So there's complexity just in the sales process just because they're large complex organization. So we navigate through that, we get to the right people, and they say, okay, great, let's try it out. A lot of times, try it out is I want to do a pilot. And I think there's been a lot of talk in the startup world about pilots being the death of startups. There's actually some clever terminology people had created to say it’s death pipe pilot.

We try to avoid that and we actually say, let's go do a full deployment. Even if it's a limited size, let's go do a full deployment, let’s sign up for a year and let's start collecting data. And that's probably our biggest sales tool, is that once we start collecting the data, and we start showing the savings, our user interface actually will show how many Steam Traps are and Blow Through, how much money you're wasting every day by not fixing them, how much money you save by fixing them.

No, it really is a compelling business case to say, yeah, you're getting value from the solution. So that process is probably a shortest three months, but up to six or nine months of going through, getting to know the solution getting to feel comfortable with it, before we start talking about, alright, let's add more, let's go do the entire plant, like we talked about with the pet food factory.

Or in some cases now, we just announced the funding round recently. And we were fortunate to have one of our customers in that round. So Colgate Palmolive was quoted for us. And so this is one of the few customers I can talk about at this point. But they were already in talks from the first plant being able to start to talk about let's look at other plants throughout their manufacturing infrastructure, and plants that do similar to continuous manufacturing operations.

So that's probably the biggest part. And that's what we're focused on today is okay, we've got a great set of customers across the broad scope of industries, and we need to go build that relationship, go get further deployments. And then start telling them about Flare system, we skipped over that, but we should come back and talk about what that product is about. But all the people that have Steam Traps also have motors. And so in every case, and it's a really hot topic today in terms of there are a number of companies that are doing battery powered motor, or machine health monitoring solutions. And we're the only one coming out and saying, hey, we can do this maintenance-free. We can put a sensor in place that will last for decades without ever being having to send a maintenance guy to take care of it.

Erik: The deployment, are we talking a one day something that's your team is able to handle internally, or is this a larger effort that requires a system integrator to get involved?

Bob: No, it's actually the initial deployments was handled by our team. And we have a master pipe fitter steam trap guy on our staff, and he's been able to lead deployments. More recently, we've been training customers to do that. And as I mentioned, we're starting to build up the service organizations, and we're training them to do that as well. So, ultimately, we're not going to scale to be a service organization to install the solution for the customer; we really need to enable others.

And the goal is to make the sensor itself, you're able to install it without tools within less than five minutes. I think the record for one of our customers is they installed it in about 90 seconds. And then we automated the ability to pair and provision the sensor, kind of like what you do in your home network when you add a security device or an IoT device; but it's really easy to go through a set of online steps and then boom, the sensor is up and running. So, it really has focused on trying to make that as quick and easy process as possible.

Erik: Machine condition monitoring, this is obviously a huge use case. But then the other the Flare system monitoring, that's again, a more specific one, how did you arrive at that as maybe your second to the more focused use cases that you’re prioritizing?

Bob: I mentioned that there's a generation of technology difference between the steam and machine product for us. But within the first generation, we have had a number of customers request specific products. So it's like can we put the same product on our heat exchanger? Can we put it on the ceiling coil or different things?

The one that appealed to us and again, you mentioned that the focus is key. So you go do so many things startup in parallel. And the one that caught our attention was really in the oil and gas space. This concept in a refinery you have flare stacks, and when they go off, that's a failure, something went wrong. And you would really like to avoid that completely. And the way you do that either to avoid it or to be able to rectify it is you need to know for that overpressure within your complicated refinery, where that overpressure event came from.

And it was described to us by one of our customers that we were talking to about Steam Traps, he said the problem is if I have a service, I send everyone out to go find it, what happened was, where's this coming from. I really need an automated way to help direct that investigation. And so we actually came up with the concept of really putting out, again, a large array of sensors on pressure relief valves that can tell you specifically for the pressure buildups are coming from. And so it's a pretty simple extension of what we've been doing with Steam Traps, again, looking at about and measuring temperature differentials, but applied to a whole new problem.

And in this case, it's a system wide on that. It doesn't do any good. Just have it on one, press release, you really need to look at all the headers within the steam system, and be able to detect. And in some cases when the pressures building so that you can take action, either your recovery system or be able to send people out and take action before you have a flare...

And so that products really focused specifically at the large oil and gas and chemical refinery space where they have flare stacks. And part of the idea behind that product is that it is a higher value solution than just talking about the asset that is the Steam trap. And the fun part is that as you go through the square system solutions, there are thousands of Steam traps sitting around, it's like okay, well, since we're here, why don’t we hook these up, and then you can save money there too?

Erik: I see that you've raised a Series C recently, so congratulations. And I suppose this is going to be the round that really funds launch of the business and the scale. Can you provide a little bit of detail on that? I don't know if you're able to share how much you raise but where you intend to invest this in growth. I think you've already mentioned building out a service team. I see that ABB was one of the strategic investors there. So maybe there's also investment in partners, with companies such as this. Where's your priority going to be over the next two years, is more around building new product, scaling up sales team, entering new geographic regions? Where do you see this funding?

Bob: I am happy to share with you we raised $30 million in this last round. The lead investor was actually Future Fund, which is the sovereign wealth fund of Australia, so it’s a very large venture investor. And then our other key supporter of the fund that has been supporting the company since the beginning is new enterprise associates, which is one of the oldest, most prestigious and largest venture funds here in the Silicon Valley. We also added some great, smaller but focused investors around the oil and gas space or the energy space, and then as you mentioned, ABB, as a strategic investor.

And so without giving away too much, but we were talking about partnerships and motors and obviously ABB which fall into the category of someone, that would be great to partner with around monitoring technology. What do we do with the money, as you said, we are at the point of the business starts to expand. So part of it was getting the money in place so that we could go out and hire that VP of sales and he could go to start to build out the sales channel and really build our presence.

On the development side, one of the great things about this company is that we actually have two threads going on at the same time. One is we are building a business in the industrial space. We have plans for new products. The first the motor health monitor is the first of what I anticipate to be multiple products based on that Gen two platform, hopefully, two or three more in 2020 coming out based on that platform. So really having the ability to do some very interesting things, gas sensing, some acoustic solutions, different really compelling new sensors based off of that.

And then in parallel we are working on the next generation of technology. We're really driving that technology vector to the next level. And it will apply to industrial. But it really also will empower us to be able to expand beyond industrial into other areas as well. And some of the really interesting things there would be around camera technology or location and things like that, but traditionally want to think of to be able to do in a batteryless manner. So that's from the development side is both product development, getting new products out, as well as this never ending drive on the technology side to not only as we mentioned increase the range and the compute capability but decrease the form factor, reduce the cost.

Our goal for this company is eventually create a sensor that's the size of a stamp that has all the components you would need to harvest energy, do compute, have all the radios and the sensors built into that gold stamp form factor that you could put everywhere.

Erik: Is there an early stage startup that's on your radar? I assume that you're in the position in your career where you're probably also doing some personal investments here and there, but it doesn't have to be a company you've invested in. Just is there a company that you think is doing something particularly interesting today?

Bob: But Nyansa is a company that I've admired for quite some time. And they just recently created a security product that I think is just incredible. So they started out as a network management platform, operations management, application proficiency or improvement, but it included the entire network and went out and looked at the WiFi and all the point devices associated with that network, not just looking at the core network itself.

So with that as a basis, they've actually created an IoT security platform that haven't talked to him in a while. But I imagine that's got to be just a rocketship for them. That's where they can go using similar to Google's web crawler to have a crawler that goes out and figures out what's on your network with all the different IoT devices out there. And then because of the massive amount of data they've taken, they can start saying, okay, well, here are your security vulnerabilities based on what's happened with other people or what we've seen with those devices. So they're one that I would call out that I think it's really interesting and compelling.

Erik: 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. 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 erik.walenza@IoTone.com.