The World’s First Battery-Free Sensor with Richard Sawyer, Vice President of Engineering at Everactive

(Intro Narrator at 00:00:01) Today, we're talking to Richard from Everactive all about the world's first battery-free IoT sensors. You're listening to Joel Beasley, Modern CTO.

(Richard at 00:00:14) Hey, buddy. How are you? Hi, Joel.

(Joel Beasley at 00:00:16) Well, it's pretty cool what you're doing, and you got a recommendation through, I think it was Mike at Timescale.

(Richard at 00:00:22) Yeah, that's right.

(Joel Beasley at 00:00:23) Is that a product you guys use of theirs, or what?

(Richard at 00:00:26) Yeah, exactly. So we use Timescale in our managed data service. They've been a great product for us.

(Joel Beasley at 00:00:32) Why do you need them?

(Richard at 00:00:33) Sure, yeah. So Everactive develops IoT sensors. Our niche is sensors that don't need batteries. And one of the things that that unlocks is the ability to collect lots of time series data. So Timescale is built on Postgres, but it specializes in time series storage and retrieval. So we actually use that as our main data store in our system. We use that for the access pattern where users want to zero in on a particular part of their data that we brought their attention to, and so Timescale is great at that kind of query. We also use it in our area analytics processes. We'll look back over months or years worth of data for any given time series dataset. So it's really core to where we're storing data in our system.

(Joel Beasley at 00:01:25) Yeah. I love Postgres, magical elephant that stores information for you.

(Richard at 00:01:30) That's right. Yeah, that's what I call it. I love it.

(Joel Beasley at 00:01:33) The first time I saw it, I saw it operating on one of my friends, Max, and it had the little bouncing elephant going. I was like, what is that? He's like, oh, it's this database. And then from there, it worked really well with Rails, and I kind of fell in love with it. But I am curious. How do you do this? How do you make it without batteries, without external power? Are you doing piezoelectric stuff or what?

(Richard at 00:01:55) Yeah. So the sensors are self-powered. We remove the battery. And so instead of having a battery, we're harvesting energy from the surroundings. So typically, that's light using a photovoltaic cell or it's heat using a thermal energy generator. And what makes that possible is that we're building and designing our own chips. And so on that chip, we've integrated an energy harvesting power management unit, which we just refer to as the PMU. And that can operate from lower and lower amounts of energy with each generation. Then on the same chip, we've combined that with an always-on radio receiver and we also have an ARM core, which is pretty typical for a microcontroller or an embedded system, but we've redesigned sort of the logic cells and the libraries that the ARM is built from. So we license ARM, but we've replaced the building blocks that that is made from. And so by bringing all of that into a single system on a chip, we can optimize really for power for an embedded IoT device so that it can be a self-powered energy harvesting sensor. So a lot of what we do is chase microwatts. We're concerned with every microwatt of energy that's being used in the system.

(Joel Beasley at 00:03:12) Because that's the big thing that's holding it up, right? Because some would say, why don't you just buy a—if we're talking about vibration, right? Piezoelectric thing, buy a chip, stick them together. But it's a power consumption problem ultimately, right?

(Richard at 00:03:25) Yeah, exactly. So the your average power, the peak power, and the amount of energy that's available to you for your application. Light and heat for a lot of our customers are great sources of energy because they're, lights are either already on in the room or there's just waste heat. So what makes our system unique is, if you touch the surface of the thing that you're measuring, if there's just a couple degrees of temperature delta between—so it's just kind of warm or cool to the touch—we can harvest that level of energy.

(Joel Beasley at 00:03:59) How? Magic?

(Richard at 00:04:01) So a thermal energy generator as a harvesting source, the voltage varies. And so what we've optimized for is at that very small end of the energy harvesting spectrum, that's a very small voltage. And so we've tuned our circuit to be able to operate at those really low voltages. And that's something that no one else is doing. And then because the rest of the system is optimized, we also don't need as much energy once it's harvested. And so it's really combining those two into one system.

(Joel Beasley at 00:04:33) Is this something that you just do for companies like projects, or can I buy it myself? How does that work?

(Richard at 00:04:39) Yeah. So that's interesting. Our journey, we started actually implementing a couple of products end-to-end ourselves. And we just refer to those internally as our full stack products. And those were focused on industrial environments, mostly industrial manufacturing factories, big facilities. Both of them primarily harvest from heat. The first one is actually measuring steam systems and steam gets used in all sorts of industries—chocolate factories, beer factories, pharmaceuticals. And so we're measuring a component in the steam system that when it fails, you lose a lot of water and you're just wasting energy and water. And so by just putting a simple sensor on there that can measure temperature, we perform analytics in the cloud that notify the end users that the product has failed and that, hey, there's this opportunity to go save money on your energy bills. And so the sell there is that it pays for itself. So that's a product that we're actively selling and delivering today. It's actually just recently, in the last six months, that we've shifted our focus towards opening the technology up so that folks—anyone—can take our technology and build their own self-powered sensor using our the same technology. So we're in the process of that, and we just released our first eval kit just a few weeks ago. So I think this is where I put in my shameless plug that your listeners can go to our website today and buy our first eval kit. You can also actually buy our—what we call—ready-to-deploy sensors. So they're industrial hardened. It can measure vibration, temperature, humidity. You can buy that. If the sensor that we've already created solves your problem, then you can buy that one off the shelf. If you need a different type of sensing or a different type of harvester, you can buy the core technology and develop your own products with that.

(Joel Beasley at 00:06:32) So if I want to buy one, like an adapter and stick it on my truck, right, as I drive around town and things like that, will it just output a bunch of information? What would it be able to do?

(Richard at 00:06:42) So we're not building any of the sensors themselves. We're using something that's off the shelf. So it's like a microcontroller development kit where you connect the things that you want to sense. So you find an off-the-shelf sensor that can measure temperature or you hook up a microphone or an ultrasonic sensor. Really anything that you can connect to a standard electric bus like SPI or I²C and analog output. And for our application, I guess if you put it on a truck, that wouldn't be the ideal application for us. You could still make use of our power management unit. You could still do the energy harvesting, but you wouldn't be able to take advantage of our radio receiver today.

(Joel Beasley at 00:07:26) I wouldn't? Why not?

(Richard at 00:07:27) Because the truck's moving around town, you would need a network to follow you.

(Joel Beasley at 00:07:31) But couldn't—could you talk to my cell phone and look in the car?

(Richard at 00:07:35) Not out of the box. You could add just the same way that you add a sensor, you could add an LTE modem that could talk to the cell towers or to the radio in your truck. But the system that we have, so I talked about the chip. We actually develop our own Linux appliances, which are the gateway to the internet for those sensors. So in an industrial environment, which was our first target customer, you might have hundreds or thousands of sensors all in the same factory building. And so the Linux appliance, which we just refer to as the gateway, aggregates all of that information. It has an LTE connection for backhaul and we actually, that's part of the solution that we're providing. So we actually have thousands of those Linux appliances installed across the country as well as in some other parts of the world like South America, Europe, Canada, mostly in the United States. And there's actually a dedicated engineer on the team who manages that fleet of appliances.

(Joel Beasley at 00:08:41) So how do the chips—how do they talk to the appliances? Is it Wi-Fi? What type of technology are they using to talk and send data back when they're in the industrial environment?

(Richard at 00:08:49) We're using the 915 megahertz spectrum. So 902 to 928 in the US. It's actually a protocol that we call Evernet. There's sort of two reasons for that. One is that at 915, you get better RF propagation for your energy budget. That's just the laws of physics. So at that frequency, you get better penetration through walls. You get—even if there was just air—you get a longer propagation in free space as well. So that scales with frequency. The other reason is we have an ultra-low-power radio. That's part of our differentiating technology. And the thing that's interesting about the radio is that it's so low power. It's like most embedded systems. The amount of power we're using would be in the noise. It would be less than the standby current of your embedded system. And because it's so low, that means we can leave it on all the time. Whereas most radios would be a hundred or a thousand times more power. You'd have to turn them on and off. That's the reason for that. But there really isn't anything stopping you from, like I said, adding another kind of radio and using at least two-thirds of the core technology and that's still possible.

(Joel Beasley at 00:10:00) Does that have internal logging? Can you have internal data stored on the chip?

(Richard at 00:10:05) A certain amount, yeah. And it's just a trade-off of RAM and flash, essentially, the storage that you have available.

(Joel Beasley at 00:10:12) So is most of the processing being sent back or is it happening—do you have anything happening at the node?

(Richard at 00:10:17) Yeah. Actually, you can do quite a bit of processing on the nodes or at the edge. And so our vibration sensor, we're actually doing a lot of what I would call digital signal processing. So we're performing FFTs. We'll take a large chunk of samples and then we'll post-process them and only send the interesting metrics that you can derive from that sample. And in some cases, you might identify a sample that's really interesting. And so then instead of just sending the summary information, you could send the entire chunk of data sort of a little bit at a time and we'll reassemble that and you'll have the sort of small file pop out the other end of the system.

(Joel Beasley at 00:10:59) Are you concerned at all that GPT gains consciousness and uses your technology to enslave humanity?

(Richard at 00:11:06) Yes. Yeah. Of course. Yeah. We actually have a group that focuses on ethics. And so, yeah, we've discussed that topic. Things like surveillance or monitoring people who might not even know that this technology exists and, you know, just sort of underrepresented groups that might be exploited because of that. And so that was a big topic when we were switching to opening up the technology because at that point, we're no longer in full control of what's possible. And so that's been discussions about ethics within the team. We've had our key board members in on some of those discussions of sort of all levels of the company. And we have an architect who leads up what we call Platform Governance for Everactive. Yeah. It's a little bit early. We don't have quite a platform yet. That's our aspiration. But we realize you can't kind of wait until you're done to say, okay, now it's time to think about the ethics and the governance of the technology. So, yeah, that's a real concern, a real topic of discussion here.

(Joel Beasley at 00:12:12) I like that you open it up though, because if you don't, someone will. This is definitely a technology we're going to see more and more and more of, right? And you've got all these firms out there, all these engineering teams, that they're not experts in low radio frequency. All the little things you needed to get this to work, but they have the need for it. So by opening it up, I think that's going to quickly become the largest part of your business.

(Richard at 00:12:35) Yeah. I agree. Yeah. And that's the dream. That's the hope of making it a platform.

(Joel Beasley at 00:12:40) Nice. So the chocolate company, do you treat them as your favorite client and visit them often because they're a chocolate factory?

(Richard at 00:12:47) Yeah. There are definitely fringe benefits to visiting. You get free chocolates. I also love the names that they have for their facilities. Like, this is the toffee stretching room or the peanut butter cup facility. Yeah. It's something out of a movie. Yeah. I love that.

(Joel Beasley at 00:13:06) That's pretty cool. And are you one of the founders? When did you join the company? Explain that.

(Richard at 00:13:12) I am not a founder. I joined Everactive about almost five years ago and we do have two founders and they actually have the title of CTO. So they're co-CTOs. They're both professors, one at UVA, one at the University of Michigan. Those two founders, they've been researching the technology that we're taking to market since they were both students at MIT. So a third of the team, I would say, is from a related research group either at MIT, UVA, or Michigan. And so that's part of the fun really of being on this team is we're taking what they did as students in grad school and we're making it reality. We're taking that to market. My background is actually wireless communications, and this is the second startup that I've worked at. And so what I've done is bring my knowledge and skills with building IoT and wireless products to the team here at Everactive.

(Joel Beasley at 00:14:11) What did you learn from a leadership career perspective at the previous company?

(Richard at 00:14:16) Yeah. The previous company was the first time I'd worked at a startup. So I was—it was drinking from a fire hose. I'd never done due diligence before. I'd never worked with VCs or—and at that company I started at the very beginning. So it was just me and the CEO and we were building the team while we were still at our day jobs. And so it was just a huge learning curve. But I think there were lots of valuable takeaways like learning to build a team, learning to start things from scratch and we had some initial customers. So, just the amount of effort you put into each customer and not all of them work out. One might be the most time-consuming customer. You're constantly supporting them and that ends up not becoming a reality whereas you have a second customer who generally knows what they're doing and are doing a great job, and they take off, you know.

(Joel Beasley at 00:15:14) That's how it goes, though.

(Richard at 00:15:16) Exactly.

(Joel Beasley at 00:15:16) The worst customers are going to want the most and pay the least, and your best customers are going to pay the most and just let you do your thing.

(Richard at 00:15:23) Yeah. Exactly.

(Joel Beasley at 00:15:25) Where did you learn building a team? What's the one takeaway from building a team?

(Richard at 00:15:29) Well, I feel like I'm constantly learning new things, and I think a big part of it is, I think it's a bit cliche, but just how important hiring is and how important it is to bring the right people on. But then, you know, once you've hired someone, I think identifying their strengths and where they will succeed is equally important. And I think when I'm able to do that, I've been very successful. You know, it's amazing to have that person just crushing it. You know, I'm learning from them as much as they're learning from me.

(Richard at 00:16:04) Yeah. I think at that point, just empowering them, just continuing to set them up for success and realizing that what motivates most engineers or most people on the team is success. And that might be a little bit different for each person. There could be some concern that it's not aligned with the mission of the company. But I find actually in most cases, their definition of success is generally aligned with the success of the overall team.

(Richard at 00:16:31) So just identifying that and helping them to achieve it.

(Joel Beasley at 00:16:35) What currently is your definition of success or the thing that you're going after at the company?

(Richard at 00:16:41) Yeah. Well, right now, I would love to walk into the next time I'm doing an interview like this and just show all the examples of

(Joel Beasley at 00:16:48) Right.

(Richard at 00:16:48) how people are using our technology. You know, I want to walk into each boardroom like, hey, look at company X. They just built this new application with our technology. And the next time it's, you know, there are three or four examples and it's just growing exponentially. So it's really finding the pain that we're helping people solve and creating that solution for them.

(Joel Beasley at 00:17:09) So those first kits just shipped and so you're waiting for those stories, right? When do you think you'll have some of them?

(Richard at 00:17:15) Ideally, in the next few months. I think we'll be there. We'll have the first one or two of them integrating our technology into their products and applications.

(Joel Beasley at 00:17:25) Interesting. I could see there's just quite literally an infinite number of applications across all the different industries. It's basically, here's what you could already do, but now you can do it without a battery.

(Richard at 00:17:38) Right, exactly.

(Joel Beasley at 00:17:39) And that's fascinating because there's so many times you'll cut the install cost down for certain companies that are building IoT solutions, right? Because they won't have to run wires and do other things that they might otherwise need to do. Yeah, that's pretty cool, man. I love what you're doing.

(Joel Beasley at 00:17:57) How did you find out about this opportunity?

(Richard at 00:17:59) Yeah, actually, I was recruited. So I was at the last place I was working, a recruiter who I had never met before, she actually left a voicemail for me at work, which was, I thought, pretty bold. So at the time, I was kind of in between things and I was actually working with a couple of other guys that I'd worked with before and we were brainstorming on starting a new company. So I was kind of in that space of looking for the next challenge. They had other things going on in life that were just way more important than what we were working on. So it was just really great timing. And so she kind of left a brief description in the voicemail. And so I called her back and, yeah, that's how it happened.

(Joel Beasley at 00:18:44) Nice. And so now you're the head of the engineering there?

(Richard at 00:18:47) Yeah, exactly.

(Joel Beasley at 00:18:48) Very cool. How many people are in engineering at the company?

(Richard at 00:18:51) Right now, we have close to 50 engineers on the team.

(Joel Beasley at 00:18:54) Oh, cool. So you're getting to that point where you're figuring out how to communicate to all of these people?

(Richard at 00:18:59) Yeah. One of the maybe fun parts and one of the challenges is that we have so many different disciplines. A third of them are focused on designing chips and SoCs. We have some developers building the electronics and the mechanical designs around them. We do compliance with wireless standards and hazardous locations. It's like its own world. And then, you know, we have firmware engineers and embedded Linux software engineers and cloud developers. And each of them are experts in their own field. You know, some of them, it might have started when they were 11. They've been doing this for 30 years. They're now professionals. So that's the challenge where it's like recognizing, appreciating the expertise of each of those disciplines. But it's part of the fun. Like, I love that about engineering, all the various backgrounds and sort of journeys that each of them might have been on to get to where they are today. Like, some of them, they studied music as a PhD or, you know, they didn't start in engineering at all. And that's one of the things that I find really valuable. You know, maybe they never went to college. You know, they just went straight into engineering. And, you know, once you have a few years of experience, that kind of doesn't matter anymore. And so we've had all, you know, we have kind of all of that. There's that diversity of backgrounds.

(Joel Beasley at 00:20:24) Nice. I definitely think that what you're doing is super cool. I've gotten to experience some of these systems, but not as a practitioner, just being nearby. And I've always thought it was really, really cool, you know, seeing them soldering and all that stuff. I was like, that is so neat. But I personally just never got into it, and then I started to see some of the kits come out on Amazon for your kids and things of that nature. And I have kids now, and so I was like, hurry up and get old, kids, because I think when they're like seven or eight, they'll be into it, but they're like five and three right now, so it's not the right time.

(Richard at 00:20:59) Yeah. I love embedded systems, and, yeah, it's been a big part of my career.

(Joel Beasley at 00:21:05) All right. So, as we wrap up, what's the call to action? Go buy the kit? Go play with the self-serve kit? Where do they do that?

(Richard at 00:21:12) Yeah. So, you can go to everactive.com and you can purchase it directly from us. And, yeah, the call to action is that, you know, this is real. You know, you can go and build a sensor that doesn't need batteries and can communicate wirelessly today. You know, that technology is no longer a future reality. It's something that you can do now.

(Joel Beasley at 00:21:33) That is so cool, man. Thank you so much for listening. And if you found this episode useful, please share it with a friend or colleague who you think would get value from it. And if you have topics that you'd like to hear discussed on the podcast, either add me on LinkedIn or send me an email, [email protected]. Every time I get an email or LinkedIn message, it absolutely makes my day and inspires me to keep going.