IDEA 2021 Conference: Self-Powered Sensors for a Large Scale IoT

Speaking at the IDEA 2021 Conference in Austin, TX on September 29th, Everactive Director of Product Marketing Rafael Reyes discusses self-powered wireless sensor technology and how it is the key to achieving a large-scale Internet of Things (IoT). 

Hello! Welcome to the IDEA 2021 Conference. My name is Rafael Reyes and I’m the Director of Product Marketing at Everactive, and it’s a pleasure to be with you today to share the recent technology developments on self-powered wireless sensors for the large-scale IoT and [to] discuss what it means to steam trap maintenance and its impact to the overall steam system. 

So let’s start. This light captures a little bit of the context of the Internet of Things. The idea is that we hope that devices will get disseminated into our world and into our environments so that those environments can become smart. We hear a lot about this in many different contexts. For example, in smart manufacturing where we hope that sensors keep track, not only of the goods that are being produced, but also the equipment and the assets that are doing that production, helping to maintain them in a predictive fashion so they can operate as efficiently as possible, saving resources and achieving high levels of reliability. Smart health is certainly a hot topic too. The idea being that ongoing continuous monitoring of our bodies can do a better job of producing positive outcomes by getting information to stakeholders like our physicians. And this idea just perpetuates across all these different segments: only if we’re able to put electronic devices into the world in a way that is ubiquitous, we can get access to information and connect that information to the people who need to use it, so that we can make informed decisions and have our world operate in a way that is smarter and more efficient than is the case today.

Self-Powered IoT Data

In order to realize the greatest impact of the IoT we need to have pervasive monitoring. Think about this as a world where we have real time visibility across all assets. What that allows us to do is not just to focus on individual applications and improving their efficiencies – like for example steam trap systems or rotating machinery – but be able to aggregate all that information at a facility level or a regional level in order to make better decisions on how to operate those facilities. So the real benefit comes from pervasive monitoring or monitoring all assets, and not just a fraction of them. Is this really occurring? Unfortunately, this large-scale Internet of Things is not coming to be as quickly as people were predicting.

For example, this slide shows a few predictions that were made in 2012. That year IBM Watson predicted by 2015 there will be a trillion IoT devices. But that didn’t happen. In 2017, the prediction was down to 50 billion devices by 2020. But that didn’t happen. And then, in 2018, the prediction had been reduced further down to 25 billion devices by 2025, again pushing down the number of devices and pushing out the delivery date. The large-scale Internet of Things seems like it’s still coming, but the progress to that large vision has been very slow.

Barriers to Progress in the Internet of Things

So why is it slow? Well, I have a few ideas. I think that the first big reason is that batteries are incompatible with this huge scale internet of things. Back-of-the-envelope math can allow us to see the problem that batteries introduce. Suppose we wave the magic wand and produce a battery that has a lifetime of 10 years. We can’t really do that in most applications today. But if we were able to do it and deploy those 10 years batteries in a trillion sensors, you still need to change 274 million batteries in a day. If the battery only has a three year lifetime, then you are changing almost a billion batteries every day. The back-of-the-envelope math shows that batteries are not compatible with a huge scale internet of things. They require intervention by people and a level of maintenance that is not going to work. They will become not just a logistical nightmare, but also an environmental tragedy.

The second big reason is that the distribution of devices will continue to proliferate exponentially. You can see on this plot, which shows on the log scale, the number of units on the y-axis against time over the decades. On the x-axis, you can really see the proliferation of devices. Electronics have gone from PCs that were one per employee, to laptops with everyone having computers. And then in the post-mobile world that we live in today where each of us has in addition to a computer, a phone and several other connected devices. But still, we’re in the order of maybe 10 devices per person. In order to get to this trillion node Internet of Things vision, we’re talking about a much larger scale of adoption of about one device per thing. There’s absolutely no opportunity to get there with batteries. This really points to the availability of batteryless solutions and that eventually, self-powered operation by batteryless, wireless sensors is going to be how we get to this huge scale Internet of Things.

The third big reason is that this business of collecting insights from the physical world is a complex and complicated combination of technologies and services. It requires the implementation of assets, functions, and processes. From identifying the assets, to selecting the right sensor technology, to determining the right connectivity approach, and then finally implementing the most effective way to communicate insights – in a fashion that can create an actionable outcome to the customer with a clear return on investment. This large complexity, coupled with a large variety of applications, comes with a plethora of expertise that needs to come together for the full systems to work properly and create value. [The result is] solutions that are not seamless or secure to implement, maintain, and expand over time, severely slowing down the adoption of IoT vision across multiple sectors and applications.

Powering IoT Devices at the Microscale

So this begs the question: can we operate our IoT devices without a battery? Can we run it from a renewable energy source or from harvested energy? I’m going to call this “the renewable energy source at the microscale.” Rather than solar panels on your roof, think of a one centimeter by one centimeter solar panel the size of a postage stamp powering the IoT device. So is it possible to operate an IoT device from renewable energy sources at the microscale? To answer this question, we have to look at the supply and demand of power on an IoT device.

First, let’s talk about the demand side. How much power does an IoT device consume and what is the resulting lifetime, if I were to power that device from a battery? The graph on the left shows power of an IoT device on the y-axis and time on the x-axis in two scenarios: one powering from a AA battery and the other one powering from a small coin cell battery, like the one you might find in your car key fob. In order to get a one year, or even a ten year lifetime on those batteries, we need to be operating at the power range of one hundred microwatts or ten microwatts average power. This still means that after one to ten years, you are disposing of the device or changing the battery somehow.

Let’s look at what you can do from the supply side with the microscale harvesters. The table on the right compares a few different energy harvesters and how much power you can expect from each of them starting with outdoor light, which is really an abundant energy source from an IoT device perspective. You can harvest about a milliwatt of power from a square centimeter photovoltaic cell. That’s about the midpoint level on the demand graph on the left. But harvesting from the other sources such as human motion – like heel strikes in a smart shoe; or vibration like rotating machinery; or thermal, like harvesting through their temperature gradients; and finally indoor light, like the bulb in the room. The amount of power you can harvest from those sources is between a few hundred to maybe down to a couple tenths of a microwatt per square centimeter. So putting all this together, I have drawn this harvested power floor around 10 microwatts. If you could develop an IoT device that has an average power of 10 microwatts or less, you’ve got a really good chance at operating that device continuously from these small amounts of renewable energy sources at the microscale.

Pervasive Remote Monitoring

So how can pervasive remote monitoring be enabled? It all starts with batteryless technology. We designed our own in-house custom circuits by leveraging the ultra low-power circuit design expertise that we have developed over the last dozen years or so. Then [we] created circuits that include sensors, data processing chips, power harvesters, radios and memories – really all the components of an IoT device. We operate the system-on-chip under a 10 microwatt power budget. From there, we will build a full stack solution.

We integrate those chips into self-powered nodes or devices, and power them from heat or solar energy. We connect them with an always-on network that is deployed in a customer’s environment. That network connects to the cloud where we can perform additional analytics and drive insights that are valuable for the customer. That full stack solution is targeted to specific solutions in the industrial space, focusing on applications that have the maximum impact to industrial facility operations improving their efficiency, reliability, and cost.

So digging a little bit deeper, the full stack solution starts at the bottom with our custom design chips which includes an always-on, ultra-low-power receiver; an energy harvesting power management unit; and an ultra-low-power digital processing unit. So we can do data computing at the edge node. The chips connect directly to energy harvesters and operate those circuits continuously from harvested powers.

This is a key differentiation between the typical battery operated device, which often does not operate continuously. It operates on what is called a “duty cycle,” or operates intermittently. And the reason it will operate intermittently, is because it is trying to save power and extend the lifetime of the battery. In contrast, you can run the device from harvested energy power sources and you can operate continuously. And [you] don’t have to transmit the data intermittently. The self-power nodes are then powered by these ultra low-power custom chips designed to operate from harvested energy alone. There’s no battery inside so they have exceptionally long lifetimes of multiple decades and can continuously sense and compute at the edge and then communicate through our wireless network back to the cloud.

We have also developed an always-on network that takes advantage of those ultra-low-power receivers, which are listening all the time. The low-power network enables scalability to a thousand nodes per web gateway, with millisecond latency down to a node or even groups or subsets of nodes. And then the top level is the interface to the user. It includes cloud analytics and a dashboard that allows users to gain visibility into what’s going on. And through an insights as a service model, it allows users to define alerts and notify them when an asset has failed or needs maintenance so the user can go and address that situation right away.

The self-powered sensors send data wirelessly every minute at the 915 megahertz frequency range, which provides a strong sub-megahertz communication with great penetration. The gateway can then communicate to the cloud via Wi-Fi, Ethernet, or LTE. Over 90% of our customers communicate to the cell tower via LTE. The data is then stored in the cloud and can be displayed via a web-based dashboard on a mobile or personal computer device, allowing you to access the data anywhere you can run a web browser. You don’t have to install any software and you don’t have to configure any software. The dashboard can send notifications via email or text messages and communicate to other EAM or CMMS systems via our public API.

When to Use Self-Powered Technology

Sending and receiving data is the most power hungry activity of an IoT device. A key enabler of this self-power sensor’s communication is the low-power wireless network protocol. Here you can compare our Evernet low-power communication protocol with other popular wireless network protocols. The breakthrough technology in the design of radio transmitter and receiver allows the product to lead on sensor density, range, latency, and penetrability, while powering only from harvested energy, which is a significant breakthrough.

So for what use cases are self-powered sensors technology best suited for? The first constraint that you must evaluate is location. Since the self-powered sensors communicate to a gateway, they are better used for in-premise applications. Self-powered sensors work best in applications that have (1) a large volume of nodes or monitoring assets, (2) where those assets are either on or under the monitor, and (3) there is a quantifiable action that can be deducted from the data collected. You can think about these three elements in a Venn diagram and the best use cases will be at the intersection of these three elements.

The initial applications of the self-power sensor technology include a steam trap monitor and a rotating equipment monitor. In both cases, the insights generated from the data collected created quantifiable actions that lead to reduced downturn, maintenance costs, energy consumption, and unsafe actions. Currently, there are over 6,000 sensors operating live at several U.S. and Canadian industrial operations in food and beverage, pharma, institutional, and defense organizations.

Case Studies

Steam TrapZooming into the applications of self-powered sensors on the steam trap monitoring use case, the steam is used in a wide range of applications. Steam traps are devices that separate the steam from the constant condensate in the steam system. You all know that. And they are a mechanical device that wears over time and are a common point of failure for steam systems, which you probably all have experienced. At any point in time, 15-20% of the steam traps on a steam system will fail, and contribute to a lot of waste in energy, water, and Co2 emissions.

When they fail, the steam trap monitor or STM  collects data from thermistors in the steam and condensate sites of the steam trap. The steam trap monitor (STM) uses that information to determine the state of the steam trap including blow-through and intermittent blow-through.

The steam trap monitor (STM) powers from the waste heat of the steam pipe via a thermoelectric generator that can power the sensors from as little as 15º Fahrenheit of temperature difference. [It] then sends that information to the cloud every minute. On the right side, I’m showing some projected returns. A typical factory could run between a few hundred traps to several thousands, and perform annual outages to determine which traps have failed. In this case, we assume a facility with a thousand traps. Since we offer this solution as a subscription service, there are no upfront hardware costs and each steam trap monitor (STM) can be installed in under five minutes.

By being able to replace a failed system steam trap right after the failure happens, rather than waiting until the next annual audit to determine that, the facility could save over a million dollars every year, with a payback of three months and a five year ROI of 4.2x. Which also translates into an annual CO2 savings of 28,000 metric tons. That’s the equivalent emission of effectively taking off the road 6,000 vehicles for one year.

In this case, a study at an Atlantic Coast Conference university campus, we can get into more detail of the benefit of steam trap monitors with self-powered sensors. At this university campus, we monitor 500 steam traps with an average PSI of 60 pounds, a steam cost of $10 per thousand pounds of steam, and an orifice size of seven thirty-seconds of an inch. The installation cost of 500 steam trap monitors (STM’s) is $11,500 and the annual cost of the monitoring service is a $125,000 or $250 a year per steam trap monitor (STM). Every year, this prestigious university campus saved $313,000 on energy by being able to replace the steam trap right when it fails, rather than waiting for the annual audit. This results in an annual net energy savings of 188,000 with a five-year ROI of 1.48x and a payback of 5.2 months. In addition to that, the university saves 3,350 metric tons of CO2 every year.

These are some of the quotes that STM has gotten, from Phil Reynolds, Hershey’s Maintenance Manager. He was able to best articulate the tragedy of a battery power sensor when he says,: “If the IoT system requires battery placement, it is a failed system.” And he was also able to provide additional insights into the challenges of keeping a condition-based monitoring program up and running over time, which is a difficult task to achieve. He says: “When you implement a system, you want to ensure that it is sustainable over time. When your system starts to run itself, you know you really got something going.”

The Future of This Technology

The self-power sensor technology is currently in its second generation. It supports solar and thermoelectric harvesters as well as temperature, relative humidity, light, acceleration, and magnetic sensors. In the plan, is the development of electromagnetic field and vibration harvesters as well as pressure, acoustics, ultrasound, and gas sensors. These new developments will enable new applications for the self-powered sensors including filter monitors, corrosion monitors, gas monitors, and heat exchanger monitors – all working seamlessly with the same hardware, software, and networking technology. There is the feasibility to continue to improve the technology into a third generation, which will compress the form factor to the size of a cookie and enable a one kilometer non-line of sight range improvements.

The fourth generation will aim to compress the form factor to the size of a postage stamp and enable up to three kilometers no-line-of-sight range to communicate directly to a cell tower. This will unlock additional applications and use cases like asset tracking, commercial connectivity, and connected homes and wearables.

Steam Trap Monitoring Benefits

Steam trap monitoring blowing trap
Steam trap monitoring blowing trap

What does this mean to you? First, from taking the sensors out of the box, to applying it into a steam trap, and seeing real-time data flowing to the clouds is minutes – usually under five minutes. The minute you install a device in your asset, it immediately powers up and starts talking to the cloud.

Second, no additional maintenance. There are no batteries to recharge and replace and the device will work seamlessly for multiple decades. You apply the self-powered sensor once, and you only go back to that steam trap when you know there’s something wrong that needs to be fixed.

And finally, to maximize returns we want to make sure we can demonstrate a return on your investment for you and the people in your organization, who are going to insist on it before you set out on a new project like installing cutting edge technology sensors. Therefore, we have a calculator on our dashboard that helps us determine how much money you’re saving by applying these self-powered sensors to your steam traps. The bottom line is you will save money and reduce emissions with self-powered sensors. Period.

Thank you very much for your time. Please feel free to add your questions to the chat message. We will read your questions and provide an answer soon. If you have any questions about our products and services feel free to contact me by email or visit our website at

Again, thank you for your attention and have a great day!