The IoT Battery Problem, Everactive’s Solution
The Industrial IoT (Internet of Things) has a major problem: batteries. In this episode of Empowering Pumps & Equipment’s podcast, Peter Woodman discusses Everactive’s Machine Health Monitoring solution: a batteryless, always-on condition monitor for all rotating equipment that reduces downtime, improves efficiency, and extends asset life.
BEKAH: Hey everybody, this is Becca from Empowering Pumps and Equipment, we are the online resource for the Pump and Related Equipment Industry. Today, I have PETER WOODMAN from Everactive and Peter, why don’t you introduce yourself?
PETER WOODMAN: Hi Bekah, I am the Principal Sales Engineer here at Everactive, uh so I help our customers figure out if our sensors are a good fit for their environment, and answer any questions they might have about our technology.
BEKAH: Awesome! Well great to have you here Peter, and I’m excited to learn with you.
PETER: Yeah, I think it’s going to be fun to get people in front of this technology. First, I just wanted to say to the folks here on the webinar, thanks for joining us um you know. Many of you might have been thinking about digital transformations for a very long time, others this might be new to them, but in the current state we’re in uh with Covid, I think it’s given people a lot of a chance to kind of reevaluate you know what happens when you’re on-site versus off-site, you know what it means to be able to get access to your data from anywhere so that’s been a driver for us. Thank you for joining me this way kind of over the wire.
Machine Health Monitoring
So I wanted to first talk about kind of what it is we’re going through today and that’s our Machine Health Monitoring Solution. That’s for monitoring things like pumps, fans, any rotating equipment.
So this little green cube here it’s the first kind of screening tool for vibration analysts that’s always on, continuously taking tri-axial accelerometer data and sending it up to the cloud. I think in the right hands that it could change your career, so we’ll take talk a little bit more about that in our later slides; but for now, just know that this sits out on the edge and feeds back new data streams from all your machines, giving you continuous insight into how your plant’s operating, including alarming. A few things I wanted to highlight here we have a great range around industrial environments, continuously on and operating so it doesn’t have to shut itself off like a battery-powered sensor, and we think of it as um you know truly being self-powered so you don’t have to go back and swap out a battery or maintain it in that way. So that’s what interests most people in us is that we’re batteryless so I want to talk a little bit about that.
What is Batteryless?
There are four sources shown here that we can harvest energy from we’re working on a fifth as well. The focus for a machine health monitor is these two on the left so temperature differential uh something that’s warmer than the ambient air temperature or colder –in this case, a warm spot on a machine, or a pump, or the equipment around it –or the presence of light and that could be indoor light like the LEDs or CFLs or incandescent bulbs you’re sitting under now, or it could be outdoor light like the Sun. Now, this is not a lot of electricity that you get from harvesting from these sources, it’s a very small amount of electricity but at our core, our technology is the lowest power electronics and radios in the world. And that’s what allows us to use these very humble sources of energy and still power our sensors continuously. The conventional electronics you’re probably used to seeing around your plant require a lot of electricity from the very base of their design so we started over. Our co-founders are researchers at universities: the University of Michigan and the University of Virginia, and they started by creating these very low-power radios and low-power processors. When you bring those together, you can get away with some pretty humble sources of energy and still run continuously. So, let’s talk a little bit more about how this fits into sensing things and the industrial internet of things (IIoT).
The Industrial Internet of Things (IIoT)
The internet of things, in general, has not lived up to its hype back in 2012 the IBM Watson team said that just three years later in 2015 there would be a trillion IoT connected devices in the world — that’s a huge number –and 2015 came and went and we were nowhere near that. Everybody who’s taken a swing at projecting since then comes up with a much smaller number so this is 50 billion, not 500 billion, not half, but 50 billion and then it’s shrunk again analysts say today that we’re somewhere in the high teens maybe around 20 billion iot connected devices and that’s everything from you know smartphones and wearables to your smart toaster refrigerator, uh still to be under 20 billion. IoT connected devices and that’s everything from you know smartphones and wearables to your smart toaster refrigerator, uh still to be under 20 billion.
The industrial IoT is a small fraction of the number of IoT devices that are out there in the world, so there are very few industrial IoT connected devices out there — fewer than we ever expected at this point. It’s not because people don’t want that data , they do want continuous insight into their equipment and how it operates but there are two factors that have kind of artificially lowered the ceiling.
The Battery Problem
The first here being batteries. So batteries from a logistical standpoint we’ve met a lot of folks working around these plants, women with titles like reliability engineer maintenance planner that sort of thing, but I have never met a woman with the title battery changer. So when you ask somebody to change a battery you’re stealing cycles from something else they could be doing that’s more important, or is more difficult to train on or more intricate, so that’s the first piece.
But it’s not just the person who changes the battery, the maintenance planner has to schedule those repairs, make sure you have the right batteries in stock, and do that all before the batteries expire or you lose data. They also have to do something with the batteries that come out of these sensors once they’ve expired. That’s the environmental piece here, so when you think about batteries inside industrial environments they package them up really heavily to make them intrinsically safe and they usually make them out of heavy metals to begin with, both of those things make them very difficult to recycle. As a result, many industrial sensor batteries end up in a landfill. We’ve even seen entire sensors that aren’t made to have the batteries replaced the battery is soldered into the sensor and the sensor itself is designed to be disposable, so they go straight into a landfill which is really unfortunate. The environmental impacts of that will never scale to hit a trillion IoT connected devices. Even if we had this kind of fantasy battery life of 10 years, most industrial sensors get around a two-year battery life on average. But let’s say we could stretch that out five times to what it is today and hit a 10-year battery life. As those sensors aged out, there’d still be 274 million battery replacements every day — so you’d need a lot of those fictional battery changers running around plants.
The Trade Offs
But there are other trade-offs too. So we talked about the battery piece and what that does in terms of a human cost, but there are trade-offs in the sensors themselves. As a result, since they’re difficult to deploy and maintain there aren’t as many of them out in the world as there should be, you know people choose not to censor up all of their assets, they’ll put a sensor on the top five or ten percent of their assets uh but be missing out on data from everywhere else. The sensors themselves ration out the amount of data they spend- a send, and they do that in order to milk that battery life as long as possible. So you get reports a couple of times an hour to a couple of times a day instead of continuous insight, and that leaves you having to get by with the minimum amount of data instead of the maximum. So you can’t always make a good decision with the amount of data you’re getting.
Everactive Technology
So since our sensors power themselves, it’s a different model. They’re continuously on, always providing data not rationing out how often they send. They do this using those low levels of harvested energy that’s usually plentiful right nearby whatever it is we’re sensing. And they’re made to be ruggedized as well so that IP66 rating is for water and dust intrusion into the sensors, they can be outdoors in all four seasons you can hit them with a fire hose and they’re dust and sand proof. That class one division two refers to intrinsic safety, so they can be in hazardous locations up to that rating. And we support a wide temperature operating range we’ve been outdoors here in the upper midwest when the polar vortex hit a few winters ago and it was 40 below all the way to steam tunnels in the south in the summer where it can be 180 degrees Fahrenheit, without ever missing a measurement. As that data makes its way up to the cloud we think of it as a new data stream; because if you’re used to getting quarterly reports back from your vibration analysts as to how your motors are behaving, in the first hour that our sensors are on you’ll get 15 years worth of that data. Now that’s way more than you could keep up with by reading through it like you do with an external report today. So we have notifications in our platform where we can call your attention to the assets that are in most need of attention and then you can schedule your repairs should make you much more efficient. Most of those notifications today are sent over email, but we do have an API as well so if you have another business system like a work order system or some other data lake where you want to warehouse this data, we can hand it off to you also.
Everything on this slide comes from us, one company. We’re not bringing together disparate solutions from different companies, so you’ll never get stuck between you know two different companies saying ‘hey that hardware is their responsibility or the software is the other guy,’ it comes from one single support organization. So if you have a feature request or a support issue, we can solve it really easily, once again without ever having to change a battery.
What We Provide?
So by the time that science fiction technology we saw in the previous slide these harvesters makes it to you, we’ve thought through everything you need in order to instrument your equipment. The sensors themselves and how they mount to the equipment is thought through the gateways that take that sensory data and aggregate it and send it up to the cloud are included as well. Each one has a cellular modem in it with a sim card that’s already active for 4g LTE so you don’t have to worry about connecting to get your data flowing. Within a few minutes of opening the box, data begins flowing to the cloud. We can provide installation of these sensors as a turnkey service, or we can instruct you on how to do it yourself. Once that data hits the cloud, you can access it from any modern web browser; so that could be a phone, or a tablet, or a pc, um you just log in with a username and password and you’re ready to go— you don’t need any server or specialized app locally.
Machine Health Monitor Overview
All right, so let’s take a look at the actual sensor itself. Here’s four of them, instrumented across a machine train. So anywhere you see one of these little green cubes with a fin on the top that’s our sensor. Each placement location you’re getting tri-axial accelerometer data, so there’s actually 12 points of data coming across this machine train. A little bit about our harvesters here, this little puck here with the blue fins on it is our thermoelectric generator, so it’s connected via a cable. We can make those in a bunch of different lengths to connect to the sensor. That’s modular, so you can disconnect and reconnect those cables. To suit, we place this out here find a warm spot nearby set our thermoelectric generator down. And then if there’s uh times where the motor is warm and times where it’s not, we can also use a solar harvester to bolster it so you’ll see here this one sensor has two sources of harvesting nearby. Now on this particular example, this is a boiler feed pump, so the motor itself runs warm at some times, and other times it’s close to ambient which is why we’re using two harvesters. The driven equipment, the actual pump, that’s a hot water pump. So we know anytime it’s running it’s going to be hot, so we use this thermoelectric generator to scavenge that waste heat that emanates from it, and that’s all we need to power the sensor in that application.
The Eversensors
Taking a little look through the parts of the solution itself the sensor is the brains of the outfit, so that’s where those super low power radios I talked about are and the processor is. We don’t use batteries anywhere in our solution, even rechargeable batteries have a limited number of cycles until they wear out and you have to go back and service them. But we can store energy onboard on this sensor by using supercapacitors. They don’t store as much energy as a battery would, but we don’t need very much and they’re rated to last over 20 years. So you can set this sensor out and know that you won’t have to go back to do a battery change you know in 18 months or 24 months. A couple of different options for mounting this, I’ll show you those uh here on the camera in a moment together. This is a smart sensor, so in addition to getting the overall vibration levels from our accelerometer, we can also generate an FFT, a Fast Fourier Transform right here at the edge and send that back as well, so you can use that to analyze the frequency magnitude pairs in the vibration spectrum. As I said before, we’re indoor-outdoor all four seasons so we can support spray down and whatever else you might throw at us to an ip66 rating. and it doesn’t require a very hot motor in order to power this sensor if it’s warm to the touch about 15 degrees Fahrenheit difference from the air temperature, that’s all we need to get by.
Under the hood, as I said before primarily people are looking at the triaxial accelerometer for the vibration analysis data, but we also have a magnetic field sensor and that can take the rate of stator excitation that the VFD is feeding to the motor and report that as well. So we report that in Hertz, you can take that put it alongside your accelerometer data, figure out what one time is, and see if there’s a slip between the VFD and the motor. We also have temperature data at the sensor and also at the skin temp of the motor and humidity readings as well in the room.
This is a smart sensor, it’s capable of receiving software updates firmware updates, and executing code that we send to it over the air so you’ll never have to go back to do a software update or a firmware update, you place this once and we can manage it remotely for its entire life.
The Thermoelectric Generator (TEG)
Close up here on our thermoelectric generator, this is also magnetic but could be epoxy down. if you don’t have a magnetic motor. It’s the warm spot on the motor, the difference between that and the air that generates our power – so these fins here are what allow us to exaggerate that natural temperature differential and generate the most electricity possible in a small spot. This doesn’t necessarily have to be right on the motor body, if there is driven equipment nearby that’s warm or something else nearby that generates heat you can place it there too. Wherever we place this thermoelectric generator, it’ll report its temperature so we can use that as a trending data point as well.
The Solar Harvester
Our solar harvester is the energy source that harvests light and when we say solar, that means the sun, which can be confusing because it also works for indoor light, so they also call this a PV harvester sometimes for photovoltaic, you can place this anywhere as well nearby. The cables we make run 10 feet or longer if we need to so we can use a short one if it’s a short throw so you don’t have a lot of slack hanging around but if we need to get further away to find daylight we can run a chord too. so so I’m going to end the slideshow here for a moment and switch over to my camera in the room.
Live Demonstration
So I’ll stop sharing here and move over. So I have a small motor here in the room with me, it’s a seven and a half horsepower marathon a little pulley on it just to give you a sense of scale I’ve put a can of coke here so you know what size we’re talking about. I’m going to walk you through installing our sensor and what the individual bits and pieces are so here’s the sensor itself right there next to the can of coke and here up close. So you can see this port on the front that’s where we plug in our harvesters so that’s a USB-C pin out like you might see on some cell phones or laptops, but what makes ours a little different is in order to make it waterproof we add a gasket to it; so there’s a rubber gasket material and that safety screw to make sure it doesn’t get yanked out kind of unceremoniously. So that’s the connector port this little shield on the front here this black button is made out of gore-tex, which allows us to get temperature and humidity readings without compromising our waterproofness, so that’s there on the front. We have a little wake up light on the top that gives you a heartbeat when the sensor is on so you know that it’s running. I’m going to take our base off here and show you the options for mounting this so we have a screw there to keep it set and then this mount comes off, so if you had to epoxy this down to the motor you could do that and leave it behind and swap the sensor out if you’re doing service on it. It’s also magnetic so most of our customers just set it on their motor and it’s strong enough to keep the sensor on in any orientation, whether you’re vertical or horizontal mounted. We can also do a stud mount, so if you’ve got a tap and die into your motor you know to mount another sensor accelerometer already we can just screw into that and that’ll keep it steady as well and get that connected.
Alright, so I’m going to set that there top dead center and now I’m going to grab our harvester. so this is the thermoelectric generator I talked about, you can see those fins on the bottom we just place this on a warm spot along with the motor. It’s magnetic but we also have some little stabilizing feet we can fold out to keep it still. So this is our power source you can see it’s pretty small it’s about two inches square plus the connector cable, and it doesn’t generate very much electricity but we don’t need a lot. I want to give you a point of reference; I’m wearing an apple watch here this is considered the most power-efficient consumer electronics in the world has a very small battery and it lasts all day. The four big consumers of electricity on this watch are LTE, Wi-Fi, the screen, and Bluetooth low energy –so three of the top four are radios which is kind of telling, right?
The lowest of that four Bluetooth low-energy (BLE) has a power budget of 50 microwatts and the way it achieves that number is it powers itself off 99 percent of the time and fires up in little blips to talk to my phone or my laptop. The radio we have in here our Evernet radio, that’s always on and always listening, has a power budget of 200 nanowatts a thousand times lower than BLE which is the number four consumer here. So to charge this watch you would need a thermoelectric generator like the size of this table on a very hot spot, which you’re not going to find in most industrial environments, there’s a lot of places you can find a little one-inch square in the center of this that’s warm. So that’s our advantage there: we can use very small harvesters and still get by. So I’m going to connect this up here, now if this motor is hot sometimes and cool other times I talked about our capacitor banks but we also can bolster with a solar cell.
So here’s the solar harvester we use, or PV harvester. It’s about the size of a playing card, maybe half the thickness of a deck of playing cards. Then there’s holes in the corners we can use to zip tie or screw it in but it’s also magnetic so it’s got the same pin out as our USB-C pin out we use elsewhere, and you can actually daisy chain another cable to this to have both of them together at once. So I’ll set that there so that’s the physical aspects of the solution kind of what it looks like to install one they go on really quickly, you know we’ve put dozens of these out in the day with one person hundreds if we have teams and they can be pretty small teams. And within you know just a few minutes in the amount of time I’ve been on this camera view if we had that thermoelectric generator hooked up on a warm motor we’d already be streaming data to the cloud.
So I’m going to put our sharing back on here and show you what happens to that data once it leaves the sensor. It goes through our pretty standard-looking IoT gateway and then hits the cloud where we can visualize it. So you can see here in this view I’ve got every individual measurement so, each of these color bars is a chart with the overall vibration level and IPS peak of that tri-axial accelerometer. So here’s an example of a motor with some peaks on it that’s running rough and you can see we can hover over each individual one down to the minute to see what’s happening. If you were spot-checking this motor you’d have to be there at you know pretty specific times with your accelerometer to figure out you know what’s going on here when. Likewise, if you had a battery-powered sensor that only phones in a couple of times an hour you wouldn’t have nearly as complete a picture of you know when this is peeking out and how rough it’s running. We use thresholds for vibration levels on these machines so I’m actually going to turn off two of these axes and the one that remains that pink bar is when we’re in alarm, so you can see exactly how long the machine’s running rough for. We can also click and drag on any one of these areas to zoom in so now we can get down to our individual measurements and you can see these are up to the minute…so 53, 54, 55. Charted alongside those vibration levels we have our temperature measurements as well. So the ambient one comes from the sensor body that’s that 118 and then the surface comes from the thermoelectric generator, so wherever we’ve placed the tag we can chart that as well and you can see during those spikes when that motor is kicking on it does get quite a bit warmer.
So this is one visualization for the data another one we use is those frequency magnitude pairs some actual spectral analysis. So if we got this alarm, we know we’re peeking out here for something to look at. I’ll go into our FFT view and we can click to highlight just that peak and then we’ll see down here the frequency magnitude pairs for each of those sensors or axes I should say at that time. So we know that our most severe was around the tangential, I’ll scroll down there we can see here that’s where our biggest spike is. I can also click and drag to zoom in here and see if there’s something worth looking at a little bit of low-level vibration there but the big spike is here right around 3200. So your vibration analyst from there can say ‘hey you know it looks like given uh where we’re seeing these peaks the frequency at both the magnitude, uh you know let’s let’s go check on this’ and given context around the equipment you can determine exactly what repair needs to be done. We store a lot of metadata here in the cloud so it can give you more insight into you know what it is you’re in for when you head up to take a look at that pump. In the case of this one, we’re just storing a little bit of basic metadata but we actually have tons of additional fields we can store like horsepower, the motor manufacturer, things like bearing part numbers, line frequency, all that stuff. So a pretty rich platform we can also set our vibration levels here as well. right so that’s a quick look at our software but most of our customers are responding to alerts right they’re not driving through that platform, looking for faults, uh they’re waiting to be told something’s wrong. so by placing these out on the edge and setting intelligent thresholds you’ll be alerted when there’s a problem and then you can focus on that.
So a couple of notes about what it is our customers love about us they’ve told us that we made their installation really quick and easy. And that’s true we work together with you we’re not just selling sensors we work together with you to make this service work. So you don’t need specialized tools or prior IoT knowledge to get up and running, we provide everything you need. We can do the installation for you or work together to make sure it goes smoothly. and as I said before that happens usually within a few minutes of opening up the box. other things our customers love we don’t even bother their I.T. staff like 95 percent of our installations today are using pure LTE from the gateway layer up, so we don’t have to spend a lot of time you know requesting ethernet drops or asking it for assistance and troubleshooting, we’re a self-standing solution. We sell this as a service and that de-risks it for you. so instead of making a huge upfront investment in the hardware and then trying to make sense of it and see if you get a return. we are incentivized to stay with you on that journey uh you don’t have to pay a huge upfront cost and try to write it down as the equipment ages, you’ll never be stuck with hardware you can’t use uh we’ll be with you there covering it to make sure it works so it’s an all-in model. If a gateway goes down we have a team that monitors those on the back end to make sure if there’s a sensor issue. Let’s say you have a premature hardware failure, we’ll ship you a new sensor. So for you, this means something that works completely right out of the box you’re not adding this new maintenance task of replacing batteries in order to go out and get that data and you get access to 24/7 continuous monitoring and alarming on all your equipment, which makes it a lot easier for you to understand what’s going on in context.
Now I want to talk a little bit about what we aren’t what we are. There are some of our competitors that say hey ‘You know AI and machine learning predictive analytics, we’re telling you you can replace your vibration analysts with just a sensor.’ We don’t think that’s true. It’s our job to empower vibration techs, right? So we want to give you all the information you need to make an intelligent decision. When our competitors say that a sensor alone, you know somebody sitting on the other side of the world looking at sensor data, can tell you precisely what’s wrong we think they’re bluffing. But for our existing customers, they’ve seen us as a force multiplier so they can screen from anywhere this code I liked since covid hit we’ve had fewer personnel on-site it’s hard to keep up with the monthly routes now this person monitors remotely and when they see a problem they can respond as they need to. so it’s a force multiplier in that way we also think it’ll help your efficiency; when you are on-site you’ll know exactly where to go and what to do. So this quote ‘my employer’s money and my time is much more well-spent screening than running around and trying to find problems with manual measurements.’ From what we’ve seen talking to vibration analysts you know if you’re running a route like a monthly route probably 80 to 90 percent of the measurements you take are going to come back as being within spec you know expected behavior. In our opinion, those are wasted trips if you’re going to go out and measure something that’s good you know, why bother right? By putting sensor sensors out there at the edge they’ll tell you when something’s bad and you can focus your attention on only what’s bad.
Since we’re continuously reporting you’re never going to miss a measurement on equipment that is running intermittently. So we had an example where somebody told us where they needed to get vibration data off of a pump that tended to run overnight on the night shift. They called an operator in to turn it on and run it but even then, they were running it synthetically, it wasn’t genuine data in process. So from this quote here they don’t have to call in an operator to power up a specific piece of equipment, we can capture machine data in its normal running state as it’s happening. I think that’s an important distinction being able to see what it actually looks like under process, not some synthetic run that you did just to gather data. As you saw, we keep all of our data online so you can recall any one of those measurements in the entire history in the time our sensor has been on – even if the sensor is swapped out you know because it’s damaged or there’s a problem, we keep that data online in the dashboard so we can substitute a new sensor and you’ll still have continuous access to the previous records. So in this quote, I can look back at trend data over months or even years to see how the machine has performed under comparable loads over time. I think that’s important. We want you to be able to go back a year and see how that machine was behaving under a similar load back then. Some of our competitors will nickel and dime you for data access. We don’t believe in that because I don’t think you can make an intelligent decision if you can’t see all that data.
So this can feed all kinds of things around your environment right: that could be that we use the API to feed another business system and file a work order automatically, it could mean that you’re warehousing it in a data lake, you could feed it into part of your OEE data to understand your process you know from a complete picture holistically. So once this data is flowing you get complete access to it in the cloud from anywhere, so whether you’re standing right next to the motor, or working from your kitchen, or getting away from the plant entirely, you can still check in on your assets from anywhere. We think it’ll really change the way that people approach these types of tasks.
I talked a little bit about our service model before and I want to clarify here that we charge a subscription rate per sensor, essentially so instead of paying you know thousands of dollars upfront for the equipment we provide the hardware and you pay a fee to get that continuous data access and in that way, it’s de-risked right? If you take on our monitoring for some amount of time and it’s not providing value you simply discontinue the service and it’s on us to take our ball and go home. It’s kind of a new way of doing business a little bit different from how people have traditionally approached these problems but it makes it a lot easier to get started if you don’t have to make a huge investment up front. Many of our customers are doing this out of their maintenance budget because they know it’s going to be a recurring expense; so instead of having to go and get a big capital expenditure budget item funded upfront, they can roll this on and try it if it makes them more efficient, it’s really easy to demonstrate.
I should note we are not just a machine health monitoring company though there are other things that we do in terms of sensing around industrial environments. our first product was a steam trap monitor so for steam systems these can be a very costly point of waste uh when steam traps fail open they stop providing back pressure and allow steam to flow through the system that can be a very costly problem. uh machine health monitor is our newest product that was just released end of last year so if we start looking through the platform here you can see sources we can harvest from and sensing modalities that we have– the checks boxes are ones that we’re shipping today unchecked boxes are harvesters and modalities that we’ve qualified in the lab but we haven’t put fully into products yet. so the steam trap monitor is just thermoelectric powered we know we have a hot pipe so we scavenge that waste heat and turn it into electricity. in the case of the machine health monitor, it’s both of these as we saw a thermoelectric generator and a solar cell. In the machine health monitor, you get all six of these sensing modalities as well uh. so as you look down our roadmap here you can see as we check off these other boxes where else we’re taking our technology. differential pressure on either side of a filter to determine if it’s time to be swapped out, some of our customers are changing filters today on a time basis knowing that they’re going and you know throwing out perfectly good filters but they don’t have a better way to check how the filter is performing. corrosion so detecting changes in pipe thickness using ultrasound to detect corrosion. a similar thing for leak detection using ultrasound or acoustics to check for changes in compressed gas or air flows to detect leaks and also detecting fouling or plugged heat exchangers – another thing that’s on our road map there so.
Questions And Answers
so that’s kind of the overview Bekah,I know we’re gonna do some q a discussion and i’d left my camera there on the motor for a minute so I don’t want you asking the motor questions i’m here for you
BEKAH: we knew you were there peter but actually. I kind of liked being able to see that and um kudos to you for having a very first webinar where you showed us like an actual live demo, that was fun
PETER: you can ask the motor any questions you want but it’s pretty quiet I don’t think it’s going to tell you
BEKAH: I prefer to ask you questions but uh thank you so much that presentation was awesome um and I do have a couple of questions for you. So you had talked about of course your wireless sensors are batteryless. So um can you kind of explain a little bit more on how the circuits and radios are powered?
PETER: Yeah sure. So we looked at two of our harvesters there, basically, we need to find some source of energy that’s nearby and on motors or pumps. There’s often going to be heat so we always start with heat energy and by finding a warm spot just, by feel, or with a temperature gun, if it’s 15 degrees warmer than the ambient air temperature that’s more than we need. So we always start there because that’s most plentiful. If you’re in an outdoor application and the sun is visible you’re not covered over, that’s another great source of energy for us; so those are the two most plentiful. If we’re indoors, overhead lights like the room you’re in now or the room I’m in, that’s enough to power our sensor too. So just kind of depending on the application, you know if you think of a motor that’s in a very dark place then we’ll use the thermoelectric generator if there’s one that’s outside in the sun that’s a great source.
BEKAH: Well sounds like there’s a lot of different options which is great. So what does a typical wireless sensor deployment look like?
PETER: sure we uh so we look at um each machine train. In the picture, I showed I put there were actually four sensors down the machine train so one on either bearing of the motor and one on either bearing of the pump. That’s not uncommon for us to have inboard and outboard sensors on both the motor and the driven equipment. Not pictured is our gateway which is a very standard-looking IoT gateway. It’s about the size of a shoebox and it has a few antennas sticking out. Those are not batteryless, so we plug those into the wall somewhere. One gateway can support up to a thousand sensors and the distance between sensor and gateway is about 250 meters, so right around 800 feet. So we place one of those per zone in a building or area and it blankets a large stretch of the area with our network, takes all that sensor data, and sends it up from there.
BEKAH: Okay so kind of touching on what data is provided by your sensors and how it can be used and, can customers see more than just overall vibration levels? How are your customers using the service?
PETER: Yeah! So obviously the lowest hanging fruit when you’re doing vibration analysis is that vibration level, we show the overall vibration levels. We also have those frequency magnitude pairs uh so we give you some spectral data to analyze as well. Alongside that temperature data and that can be helpful too, there are customers of ours that track temperature data as closely as they track vibration data so each sensor hands you triaxial acceleration, it hands you those temperatures, I talked about and that magnetic field sensor which gives you important context around the VFD and what electricity is being fed to the motor.
BEKAH: Okay and so speaking for vibration levels, I had a question while you were doing your demo with the magnets: so with the magnetic base, I’m assuming that it doesn’t matter how much the equipment vibrates that magnet will still stay put?
PETER: Well it depends so there are some applications we’ve seen where there is it runs so rough that it could shake the magnet. So in those cases, we just epoxy down the bases and the bases are low cost, so we can send extras along with deployment. So if you epoxy it down to a motor and you end up swapping that motor out you can take the sensor off, leave the base behind and slide it into a new one. But you’re right, the magnet works in many applications but not just a vibration — there could be too much vibration in certain applications. There also are motors that aren’t magnetic, some of them are aluminum so a magnet wouldn’t work at all. So in those environments, we either screw it down with a stud mount or epoxy in place
BEKAH: So does your technology provide early bearing fault detection?
PETER: It does not. So the accelerometer today gives you data from six hertz to one kilohertz in that velocity domain. So early bearing detection requires usually higher resolution data, that’s something that’s on our roadmap in the future. For now, this screening tool is designed to tip you off on lots of other conditions below that stage one bearing texture, into stage two and stage three you may begin to see those types of faults in our sensors.
BEKAH: Well it’s a good thing to hear it’s to come yeah yeah it’s on the roadmap um can sensor data be integrated into existing customer CMMS or EAM systems?
PETER: It can. So we have a restful API that lives up in the cloud uh where the data is aggregated. So from there, we can use that API to take the data out to any other you know kind of parallel system be it your CMMS EAS work order systems historians all kinds of stuff. We have a public document on that on our website so if people are interested in getting started with our api, any measurement we take can be exported into another system down to the minute but we also can provide just things like events like when there’s a failure or a threshold breached so if you don’t want everything we generate you can subscribe to just the data you need in order to you know be alert and aware as to what’s going on in your facility
BEKAH: I like that that they can kind of pick and choose what is important for them and then only monitor that. That’s a pretty cool thing to have that capability to be picky and choosy of what you want to actually look at.
PETER: Yeah. I mean the funny thing about continuously gathering all this data is: there is some of it that is not necessarily going to be valuable at all times right. So since we’re going out and gathering a measurement every single minute we want to make sure that you don’t end up wading through all those measurements, and are only using the ones that are valuable to you.
BEKAH: Awesome! Well Peter, thank you so much for your time, and everybody I hope you thought this was as educational as I did because it was great . If you have any questions, definitely reach out to Peter and yeah thank you so much for your time!
PETER: Yeah! Thank you Bekah, this is awesome!