AWS Podcast – Developing an IoT solution to keep clean water flowing

With AWS Partner Twisthink, charity:water found a true collaborator that recognized its core business is a vital mission.




Having a clear vision about the IoT solution needed to keep clean water flowing for millions of people was a big step forward for non-profit, charity:water. Of course, the execution turned out to be challenging. Charity:water first tried collaborating with leading IoT sensor manufacturers to get existing technology tailored to its specific use case, which is monitoring hand pump functionality in real-time, across remote communities. When that didn’t work, charity:water had the gumption to try to develop the solution themselves. But the org faced frustration and stalls before finding the right partner. With AWS Partner Twisthink, charity:water found a true collaborator that recognized its core business is a vital mission. Together, they developed an IoT sensor built on AWS that allows real-time monitoring of remote wells around the world, slashing repair time by years. Listen to the podcast to hear about the development process and learn about the IoT roadmap to reshape the approach to meeting the global clean water crisis.

Check out the solution and contribute:
Get help developing your own IoT solution:



Bob Niemiec 0:00
Welcome, good morning, good afternoon, wherever you may be, um, greetings to you from, from myself, Bob Niemiec. And our panelists who you’re going to meet in a moment, I just want to start by saying, you know, thank you, thank you. Thank you in this world of webinars, coming fast and furious, to take a little bit of time out of your schedule to join us on this journey. We hope it to be useful and memorable and encouraging to you, as you plow through on your own digital or IoT experience. I heard a leader recently say America is ready for takeoff. And I would just pile on by saying, Well, if you and your team preparing for a digital takeoff while you’ve landed on the right webinars, so we warmly welcome you. We’re excited about what we have to share. And let me just start before I go to the panelists to give a shout out to this team of AWS and twist think, coming together as one. And some of you may have heard the announcement of three weeks ago, where we’ve, in a sense, framed and formed a team. And we’re, we’re hitting the field together to help organizations accelerate their own digital journey. So today’s title of building scalable IoT solutions for outdoor rugged environments is part of our linking arms and trying to help you and your own journey and to get to market faster. So there’ll be three themes that we’re focused on, you see them there, the powerful process of Human Centered Design, and how do you apply that to IoT, a few comments and words of wisdom on latest technologies, and then kind of a little bit of the proof behind it all with two real world examples. So again, my name is Bob Niemiec. I have three great panelists. With me, why don’t you start Kurt by introducing yourself?

Kurt Dykema 2:31
Hello, everybody, and welcome. Again, thanks for being with us for a few minutes today. My name is Kurt dykema. I’m managing director at twist think I’ve been with this team for 21 years. And it’s been an exciting journey every step of the way. My background is in electronics design and system design. Twist think I spend approximately half my time engaging with new clients, and the other half of my time supporting our key strategic accounts. Ben. Thank you, Kurt. And thanks to everyone for being with us. Today, it’s a pleasure to be speaking with you and sharing my background. Then over 10 years in the software industry, cross GRP supply chain CRM domains and for the last number of years focused heavily on Cloud. I’m very enthusiastic about the cloud space and the potential I see the technologies they’re unlocking for customers. At twist, think I focus on the holistic dimensions of our cloud deployments as they touch the projects that are working through our portfolio. And look after those pieces. Austin, take it away.

Austin Ashe 3:40
Thanks, Ben, Kurt, and Bob, pleasure to be here with you all. And with everyone on webinar. I’m Austin ash. And I lead our strategic partnership strategy for IoT, which incorporates a whole host of edge, and cloud services across a number of different partnership domains and experiences across the globe. So excited to be here to talk about what we’re doing with Swiss bank.

Bob Niemiec 4:05
Great, thanks to all three of you. And again, just by way of clarity, we all know AWS, Amazon Web Services, the global market leader in cloud and digital technologies, and have been at it for 20 years. So Austin as a part of that team, we just want to warmly welcome you to this call and to joining the two other colleagues here from twisting, twisting for those of you who don’t know, a 20 year old professional services firm, all about helping organizations innovate, accelerate and grow, and doing so through the power of custom IoT solutions. So that’s where the relationship and the partnership of our two teams coming together as one made a ton of sense. And we’re excited to now share with you where we’re headed So let me advance to the next slide. And bear with me. There we go. So let’s get to work. And let me start with Kurt lawston. Can you explain what human centered design is, and how it’s used in the world of IoT?

Kurt Dykema 5:22
Sure, let me take a stab at that it’s a, it’s a great question. And it has a lot of layers to it. But if you boiled it down into one word HCD is a process. The process is valuable, and powerful. Because it’s really good at unlocking new business value. And in our experience, the process if properly executed with the proper team, it works every single time. Unfortunately, the process is not a simple checklist that you can download from the internet. And follow a simple equation to unlock that value. It’s more difficult than that it’s more nuanced than that. In order to unlock the power, our experience has been that HCD needs to be practiced by a cross functional team of designers and technologists. This team has a large variety of tools available to them as they do their work. The net effect though of running this process with the right team, is deep engagement with the stakeholders that you are trying to serve. This deep connection is needed in order to understand the various types of people who are touched by you, or your clients offering in a word HCD helps us develop empathy, deep empathy, once we can experience empathy with those folks we are trying to serve, the team starts to see patterns. Sometimes the pattern is a pain point. As you talk to people, you hear the common complaint come up over and over again. Those are the easy ones. And those are can be very valuable, too. But sometimes the pattern is deeper, and then you need to dig for it. And it ends up being not a pain point. But in so if you actually ask the folks that you’re trying to serve, they would not be able to verbalize what the unmet need is. But as you practice Human Centered Design, and you go through the process of interviewing and talking to and being immersed in the world of the people that you’re trying to serve, sometimes you see a pattern of an unmet need. And those are really exciting, because those can be really valuable, too. So our our team has been trained to use this process to answer the question, What new digital experience deserves to be built. The application space for Human Centered Design is really large. In fact, it’s almost limitless. We’ve seen clients use this to explore what the what the next kitchen appliance needs to be. But we see its power in the world of IoT. And that’s where we’ve been focusing and, and trying to serve our clients over the past five to 10 years. So at a high level, Human Centered Design is a process process is powerful, but the process needs to be facilitated by the appropriate team. Austin, from your perspective, how do you view HCD? Thanks, Kurt.

Austin Ashe 9:06
Great question. You know, at AWS, we consider ourselves customer obsessed. And what we find more often than not, especially when it comes to IoT technologies and digital transformations is that the majority of the challenges and the complexities are actually not technical in nature, whatsoever. They’re really come they really stem from people and from culture, and from as you laid out, an understanding of the stakeholders involved in a particular use of a product. And so what we encourage is very much a leadership alignment stakeholder alignment, commitment behind solving a challenge versus building something that may not be solving a real problem, maybe solving a perceived problem. And so understanding the those who those stakeholders are, the value that they get from the proposed experience is absolutely critical to driving adoption of new IoT devices.

Bob Niemiec 10:30
Great, thanks. Thanks, Kurt. Thanks, Austin. Kurt, let me just throw it back at you and say, Can you share how HCD informed us of the challenges with IoT in harsh environments.

Kurt Dykema 10:47
So as we serve clients in various markets, let me try to, we try to shine a light on how HCD informs the team. So one of our clients manufactures product that exists on mining conveyor belts. So we’ve spent time on mining sites in Western Australia, walking the miles of conveyor belts and watching them run and talking to the folks whose job it is to keep those conveyor belts spinning. Another client who manufactures power tools, they wondered, I wonder if there’s value and wirelessly connecting power tools to the lighting system at the construction site? So we spent time walking a construction site and talking to the various trades, who were working on that site and exploring this idea of what if the tools could talk? What would they say what would be valuable, maybe not a client whom manufacturers a wireless lighting control system for parking garages. So we spent time in parking garages, walking a parking garage, I remember, I remember one interview where we were talking to a female executive, who told the story of how she would come out to the parking garage at night after work, and explained how frightening it was to face a dark unlit parking garage. We’ve spent time in the maintenance room of hospitals, talking about their water treatment and heating systems. All these experiences are a great way to sensitize the team, to the needs and opportunities to serve those folks in those environments in a new way. And applying the powerful tools of the cloud, the powerful tools of data being sensed and processed, and bringing new value. Hopefully, that gives a little bit of insight on how HCD informs us of the challenges.

Bob Niemiec 13:17
Yeah, great job. And again, you hear the theme of eight CD, unlocking the head, and the heart of key key stakeholders. That’s what that process is all about. Now, let me just throw it at Ben. Ben, you got to get in game here. And let me just say to you, what are some of the latest technologies being used in the world of IoT?

Kurt Dykema 13:42
Yeah, thank you, Bob, I think Kurt painted a compelling picture of you, how do you how do you get to where the value is created? How do you get close to the stakeholders that, you know that spirit of gumbaz, they say in the Agile domain, in terms of technologies that leverage their leverage in this space? Certainly, a lot of times, you know, when you’re looking at how do I unlock value from this industrial ecosystem, or from this specific application, there’s an element of sensing right into the sense the environment and need to measure things, I need to be able to capture data. And so that’s kind of where things start the processing on devices. You know, in some cases, it’s very sophisticated, you have algorithms that are running. They’re processing the data that are determining what insights need to emerge out of that data, right on the edge device itself. In other cases that’s pushed to the cloud. And the analysis is bigger and broader. But certainly this idea of not only data capture through sensors, but also the ability to process that data in an edge context. Switching gears and looking at connectivity, right? You hear about cloud, you hear about the connections and you’re also in an IoT realm about just how critical security is right? How many vulnerabilities have we heard about and so a secure connection to cloud is absolutely critical. In some cases, where you have devices that are running in rugged environments that are battery powered metering that connection to make sure that connection frequency See matches up and we don’t drain the batteries too quickly, can also come into the picture. So conductivity edge to cloud. And once you get into the cloud domain, obviously in our partnership with AWS, there’s key capabilities that are unlocked there, you hear scale brought to the forefront, when you talk cloud the ability to plug in and not have to worry about, okay, if my business grows, we’ll have to purchase servers, we’ll have to expand my data center, etc. But know that I can scale effortlessly on the back of the capabilities that AWS provides. Additionally, innovation is a key area that EBS continues to just pile forward in a pace that’s hard to keep up with, you look at the amount of services that have been released from you know, two or three services in 2002, to where we stand today with over 200 service offerings in the AWS cloud. And it’s turned into just a fantastic ecosystem, where that innovation that AWS is piling in, becomes unlocked for your applications and use cases. And you can continue on through DevOps, you know, looking at flexibility and extensions and things like that. Austin, I’ll give you a chance, I guess, to weigh in from from a cloud perspective and EBS perspective, what additional dimensions Do you wanna bring forward?

Austin Ashe 16:13
Yeah, thanks, Dan. You know, at AWS, our goal is to really provide the the fundamentals of IoT to any entity looking to build and deploy IoT devices. And it really comes in into four forms. One is technology and services, as you mentioned, and the other is partnerships. On the technology front, we have device software, that allows companies and entities to securely connect their devices, gather that data, and execute some of those processes that you were mentioning locally. An example of this is free our toss, a trusted kernel that has a very, very tiny footprint, and can allow for devices and outdoor rugged environments, for example, to run all of these advanced algorithms and communications, as you’re mentioning to the cloud without draining battery in a safe, secure manner. We have control services that allow you to control and manage a wide fleet of devices. So can you trust yourself to deploy not just one device in the field, but 1000s? Millions and all over the globe? And can you can you trust that you have your hands on on on the heartbeat of all of those devices. And then of course, there’s analytics services that we that we bring to the market to help our customers gather those insights and abstract some of those more complex insights out of the data needed to drive value for that implementation. And so those are the fundamentals of the technology side. And then on the on the partnership side, we’re investing in building a vast ecosystem of partners that have deep, deep domain expertise in IoT. And these companies range from Silicon companies, like NXP and SC, which I know we’ll talk about later, to OEM companies to s eyes and ISVs and installers and that whole edge to outcome ecosystem of partners enrich the experience of anyone trying to build an IoT device or deploy them. And so as that as the technologies continue to emerge, and the solutions get more sophisticated, understanding that there’s not one entity that can do it alone, not even AWS can do it alone, we need partners to is an important realization for entering these kinds of projects.

Bob Niemiec 18:52
And that scale. Go ahead. No, no, no, no, go ahead.

Kurt Dykema 18:57
As I say that scale that you talk about Austin, you know, in the investments, it’s very evident, right, you watch the IoT space over years, and you see things that you know, even three or four years ago would have been custom solutions that were built, now brought into the platform and available as managed services so that customers can focus on my core business, and how do I bring my model and my application, know to market versus having to do you know what Nativists loves to call under differentiated heavy lifting, you’re building the repetitive services, that again, and again, need to come forward and IoT solutions. Interesting. taking that same mentality, we’re trying to lower the bar for our clients tap into AWS, as experience tap into our collective experience on deployments. And we’ll talk more in a minute around a platform called Oris, which brings those things forward as an accelerator platform of pre build assets to help customers speed up development of their custom cloud IoT deployments. And I think that that’s key in this space, right? How do you move quickly, you try and build on the shoulders of giants and take advantage of those things that are already out there. So I think that that partnership and the benefits that come forward there, it’s, it’s just such a strong picture. So if I could just tie this conversation back to the original HCD process, I think Austin said it well. And when he said, most of the time, the problem isn’t technical feasibility. The power of the cloud and the power of the latest electronics offering can do amazing things. But what HCD does is it says, Okay, before you try to apply technology, first figure out what’s desirable from the humans that you’re trying to serve. And then ask the question is what they desire technically feasible? So you can, you can compromise your ability to innovate, if you apply the technical feasibility problem, too early. And what HCD does is it says, All right, let’s assume for now that most things that we could ideate on are technically feasible, let’s focus on desirability first.

Bob Niemiec 21:09
Great, great, all three of you are doing a great, great job. You should be on perhaps Meet the Press this coming week. But nonetheless, let’s move on. Let’s press on and, and just for the participants, again, just an encouragement that I’ll let them off the hook to easy submit a question that we can throw at them, starting at 1245. So with that, let’s move on to the next slide. Where again, we promised real world examples. And so we have two that are teed up now that are alive, they’re in the marketplace, they’re delivering digital good today while we speak. The first one on the left is with reference to the organization flexco, global leader in belt conveyor systems, and you’re gonna see details behind this organization and their transformation into the digital age. But then, right alongside of that, again, under the theme of both of these examples of digital good, you have teed up as a real world solution being delivered by Charity Water, a nonprofit with a mission of bringing clean drinking water to everyone, to everyone on planet Earth. So, two great examples. Let’s move to flexco. And again, I’m just going to open it up for the three of you to work with each other to unpack this example, and bring it to life for our participants.

Kurt Dykema 22:55
So what you see on the screen right now in one slide is an 18 month journey. So yeah, we’re gonna try to fly at the speed of sound here and give you a feel for what does it look like to run this process, identify an opportunity and build a solution and then deploy it for flexco. So starting at the left, you see some folks gathered around a conveyor belt system out in the field. And underneath that you see what we would call is a journey map. So the very first event that we hosted on behalf of our client was a workshop. So flexco is a global organization, they have employees all over the world. They flew in to one location, I think there’s about 30 folks from all over the place. And we organized our work around the creation of a journey map. So in this case, the journey map was the journey of a conveyor belt system from the time that a need is identified to the design of a conveyor belt, the deployment of a conveyor belt all the way through to when the mining site is no longer operating, and the conveyor belt is turned off at its end of life. And this was a gave us structure for our conversations. But the goal really wasn’t just to create a journey map. Although our client found that a valuable process. The goal was to understand the different types of stakeholders that were involved in the journey. So after we met these 30, folks, and we got to understand what they do and how they do it, and how they deliver value on behalf of flexco. We could start to see we could start to understand and articulate different stakeholder groups in the business ecosystem. Stone of flexco, we even started to see some patterns in stories that they would tell us from the field. And we use those patterns as inspirations for a sketch. So on the second column, you see what we would call a visual research tool. It’s a series of high level sketches, and intentionally vague and intentionally high level, but still yet very powerful. So the first event led to the second event. And then we took those sketches and we went into the field with our client. And we walked those mining sites, and we met the actual humans doing that work. And we showed them these sketches, and we said, Hey, what about this? And hey, what do you think about that? And we listen really well. And our favorite response is really, wow, that’s a great idea. Or, boy, is that stupid? Why would you do that? It has no value, we love the really positive responses, and the really negative responses, the mediocre mediocre responses in the middle, they’re not so interesting to us. Okay, so that’s the second column. As we did that field immersion, a new pattern started to exist, pattern for a new offering, based on data that didn’t exist yet. But that’s okay. It’s, it’s fine to identify needs that needs to be met by customer electronics. Sometimes it makes business sense to build new things. Sometimes it doesn’t. In this case, a simple sensor mounted to a conveyor belt created a ton of business value for our clients. So our team got to work and designed, validated and launched a custom sensor battery powered, cellular connected device that mounts onto the frame of a conveyor belt and brings new data to the cloud. Ben want to unpack a little bit, the cloud back end and infrastructure that goes along with this offering?

Absolutely. Thanks, Kurt. My favorite part of what you just said, is calling it simple because I think we all know that what may appear to be simple has a number of layers to it. And certainly like Kurt mentioned, you know, in the application, being ruggedized, being in that remote environment, there certainly are design challenges that come into that and how that moves forward. The connectivity to the cloud in this case, right? There’s personas that need access to the insights that this device is unlocking. And so secure connectivity, being able to provision devices. In this example, there was a mobile experience that would go with technicians into the field, to provision and set up new devices and to launch them and to get them connected. And so there is a dimension of the solution there that’s mobile focused. And that would help that technician who is on the move, and working in these remote sites. However, there’s a number of personas that were serviced through a web user interface as well. And it’s rather small on the slide here, but you can get a picture of kind of that map that shows me my fleet, where my device is located, what’s their status? What actions do I need to take? And how do I need to move on top of this data, in with this application, there was heavy edge processing. But there’s also a number of algorithms that run in the cloud, that are looking at data over a longer term perspective, generating intelligence, you know, based on not just what happened in the last minute, but what’s happened over a longer period, in predicting based on that. And so there’s certainly a heavy element of user interface that comes to the table from a mobile and web perspective, while also the intelligence in the data that really drives Okay, I have an insight here that I can unpack and then I can talk to you. So in terms of results, when you look at this, right, being able to track the uptime of conveyors, be able to look at equipment virtually, and connect to it without having to send a tech out on the ground, physically, to inspect it, and also react to things that are happening in real time. Certainly, there’s proactive maintenance that emerges. There’s also reactive patterns were unforeseen circumstances happen. But getting access to that and knowing that that’s happening in real time is absolutely key. Also equipping flexco to look ahead and evolve into new business opportunities. Austin, I don’t know if you want to layer on from an AWS perspective, but I think those are some of the key dimensions of that holistic solution.

Austin Ashe 29:33
Yeah, the couple things I’ll mention here is this story is so fascinating because what we what we’re showing here is how a company that is 110 years old, was able to take us traditional, analog, rugged device, work with twist, think and connect that device to the internet. For all practical purpose. purposes. And you all did it through this invention of the sensor. And the the fascinating thing behind it is the new business models that that come out of this the new outcomes around safety, as you mentioned, Ben. And and you know where where we were excited to participate in this is of course providing the underlying building blocks that helped twisting focus on the outcomes that the customer desired. They can bring together their specialty expertise in embedded software and systems and end to end architectures. And the customer didn’t have to worry about is the operating system secure? Is the connectivity secure, is the system going to be able to function in multiple areas around the world. So AWS was able to bring forward those building blocks for for twisting to utilize things like free our toss, that lives on that as on that NXP chip in that picture there, connecting that device through a ublox connectivity module, which is one of our partners, and connecting that to AWS IoT Core, so that you had an elegant way to show flexco how all the devices were going to be to be managed. So partnering with with twist thing can be able to bring NXP and ublox. to the to the story along with our edge and cloud building blocks. It was fascinating to see how you all brought all that together to bring flexco elevate to market.

Bob Niemiec 31:38
Yeah, awesome, great job by all three of you, we’re gonna in the interest of time, we’ll move on to the second, a real world example. But again, if I could just for the followers for the participants, you’re hearing this reoccurring theme of the incredible power and value creation that comes through a reliable process called Human Centered Design and or design thinking. Don’t take it for granted, it will allow you to change the world, Kurt and Ben and Austin, how about a few words on Charity Water to

Kurt Dykema 32:21
Charity Water is a nonprofit based in New York City. And they do a great job of bringing the need and inspiring action to solve the problem of delivering safe drinking water to the world’s population. So you could broadly call the the economy that they participate in the water sector. What a lot of people aren’t sensitive to is that the water sector generally has not solved the problem of continued maintenance for the infrastructure that is installed to bring this safe water to people. It’s relatively straightforward to understand and inspire donors to fund the delivery of a new hand pump or a piped water system. But I think if you stop and think about it, you’d realize that that infrastructure is going to age, it’s going to break. And it’s going to need care and feeding. The sector isn’t really good at monetizing funding and supporting maintenance. Okay. So again, back to our process, the first thing we did on the left there was we gathered folks working in the sector together for a workshop and got to understand the process of identifying a need for water to delivering the water and then maintaining the infrastructure after the water starts to flow. And what we realized was that the ability to efficiently efficiently maintain these systems was very difficult. The water points are very hard to get to. There’s very little communication to and from these sites. And so this pattern emerged of Hey, what if we could simply just know whether the water was still flowing out of a hand pump or not? Is there value there? How much more efficiently could we maintain and keep that precious water flowing if we could know anywhere in the world Whether the pump was still working or not. So again, after understanding the stakeholders a bit, putting them into primary stakeholders and secondary stakeholders, one of the primary stakeholders that rose to the top was these maintenance teams, these mechanics, who are charged with keeping the water flowing. So that led to an IoT offering. And a portfolio of sensors. And what we’ve been working on here at twisting for the past year or so is a custom sensor designed specifically for one particular type of hand pump, called the Indian mark two. So this particular sensor, very easily installed, take four bolts out of the pump, slide the sensor and put the four bolts back in, this sensor then will detect water flowing in that cavity and report that. And if everything’s working fine, it’ll simply report its data once a week, but it has algorithms running on it. And if it detects that, hey, something is different, it will, it’ll notify the cloud quickly so that we can we can get mechanics on site. So the first offering of data and sensing was simply water flowing. But then we set a goal to even do more. And we said, What if we could not only detect whether waters flowing? And what if we could detect and measure the health of the mechanical system that is pumping.

And so we theorized on that a little bit, we did some experiments, and we said hey, to the most common failure modes are cracked, cracked pipes, or failing seals. And if that’s happening seems like it would take more human energy to get the water out than it would for a healthy pump. And so we devised a way to measure the human energy going into the pump, what we did was we mounted a permanent magnet at the top of the rod that’s used to pull the water up. And we put a magnetometer on our circuit board. And with those two things, we could detect and measure the travel up and down of that rod. And if you can quantify the travel, you’ve effectively quantify, fight the amount of human. So we now were able to measure a simple metric, how much water is being delivered? And how much human energy is it taking to deliver that water? And sure enough, that metric is effective and useful in predicting a pump that’s going to fail. So now not only can we measure if water is flowing, we can also predict when there’s going to be a problem.

Bob Niemiec 38:03
Then or Austin, any comments to follow after that?

Kurt Dykema 38:07
Yeah, just really briefly, Kurt, I love that story. It’s so compelling to kind of see the technology used for good in this case, and being able to keep impact people’s lives in such a positive way to just click dimensions I want to hit on right, of course, when you talk about an organization like Charity Water, their focus is on impact. And how do we make this need? How do we solve this need and scale for this need. And so thinking about a cloud offering that goes with this, the the idea of I need a huge engineering staff, I need a lot of technical complexity to keep this running. It’s not compelling for them, they’re looking for, you know, a lightweight offering, but that an offering, they can scale as water sector grows as their impact girls in the future. And we were able to achieve those goals with AWS, focusing on services that can scale that can meet that demand in the future as the fleet grows, while at the same time making sure that they’re efficient, they’re managed, and we keep that cost of ownership, low cost. I don’t know if you want to pile on with any additional dimensions. But I do want to hit that quickly.

Austin Ashe 39:07
Yeah, I’d love to chime in here and just say that this was an absolute delightful project. It’s a port twist think on. And the story is just absolutely incredible to see how it was thinking Charity Water, transforming the lives of human beings around the world that need need access to something like clean water. And I think the thing that was was most satisfying about about working with with you on this program is again, just helping provide the basic building blocks that you needed to bring to this system so that you and Charity Water could focus on the results. And some of the technologies that we use in the flexco example were used here, right like free are toss an operating system that allows you to securely connect to the cloud, and a very simple footprint in a reliable way so that you can get the data off the water pump off the magnetometer as Kurt was saying, and send that data through a service that AWS has some called simple email service to the technicians informing them, hey, there’s a pump that could be going down, you should go check it out and make sure that you adjust the the tension or the bolts or whatever needed to maintain it this pumps integrity before it breaks, and just allowing you to take advantage of of st micro, our partner for the silicon and ublox for the connectivity. And again, being able to bring that together in the right building blocks in the right fashion, stitching it together. So that Charity Water can deliver a pump with 10 years worth of data for under $300 in any location around the world, which makes us extremely economical and scalable for communities around the world.

Kurt Dykema 40:58
There’s that breadth of services, right, the incredible power to not just IoT, but also extended to all these other parallel spaces.

Bob Niemiec 41:06
Yeah, to the comment earlier about going digital is a team effort and a team sport. No one company can in a sense, do it alone. So great comments there. And again, it it’s worth mentioning as we transition to this question, just you know, the kudos to both flexco and Charity Water, setting out a bold vision and then taking the action to get there. That’s that’s no small feat. And again, our compliments to them in that regard. So here we go. We’re in the homestretch. You guys are holding up? Well, and the questions are looming. They’re stacking up here on the side. We’re over 1000. Right. Now, we might have to extend this webinar. But here’s a here’s a, here’s a softball pitch of a question. What advice would you give to someone starting their IoT journey? What’s one? I know I’m catching you called with a question. But what’s one Golden Nugget that you know immediately comes to the forefront of your mind.

Kurt Dykema 42:13
Don’t skip the the messy? First step of understanding your stakeholders? Don’t assume that you understand what they need. Go and talk to them. invest the time, invest the money to go and talk to understand your stakeholders first.

Bob Niemiec 42:35

Kurt Dykema 42:35
I love that one. Yeah, I’ll go quick before Austin steals mine. Hmm. So as a firm believer in the Agile process, thinking iterative thinking MVP, thinking, how, how can I pull forward the prototyping of something? How do I get, you know, kind of this fast feedback early in the process? And very often you get into product management into like, okay, let’s develop a product, let’s scope it out. The temptation is to think well, if I could do this, this, this and this, think of the impact I’d have. And I think my advice would be how can you start as small as possible? Like, can you prove the value of that? And then how can you think iterative on top of that, right, and there’s a little bit of magic there of saying, What do I do first? What can scale with that? But in that way, you let yourself earn success in a small space and grow over time? I think that’s just critical in this realm.

Bob Niemiec 43:25
Awesome. Awesome. Anything else?

Austin Ashe 43:28
Yeah, I think what I would offer is that whenever you’re undertaking an IoT journey, whether you’re developing something brand new that’s never been connected, or taking something that has been connected and improving upon it, there are three laws of IoT, that you just want to embrace and understand. And those are the law, the law of physics, the law of the land and the law of economics. And you can’t violate any of those three, the law of physics is just recognizing that getting data from a device to the cloud in a safe, secure manner, anywhere in the world, especially in an outdoor rugged environment, is not an easy thing to do. But with the right team, as Bob said, and the right tool set, it can be done in the right way. With the right latency and the right cost parameters. The law of the land is just making sure you understand your government, local regulations and any safety regulations so that you’re not violating any of those. As you bring about something that works. That’s great. But what if it violates this law of the land over here, all of a sudden, you don’t have an IoT solution. And then the law of economics is straightforward, but but the key here is, it kind of echoes what Ben and Kurt say is really try to avoid a solution. Make sure you’re working backwards from your customer understand the value that they’re going to get out of the IoT product. And that’s what will drive the economics behind the cost of the of the cloud infrastructure, how much compute you do at the edge, what kind of connectivity you put in. To place all of those tech decisions come out of understanding these these three laws. Fantastic. Couldn’t agree more

Bob Niemiec 45:07
love the reference to the three IoT laws? Okay, so we’re building up to some q&a. Now, before we do, I just want to pause here for a moment and make the participants aware of you’ve heard reference from all three panelists to this exciting asset that we call Oris. And so Oris, again, was a proactive effort, that team twist think engaged in, in the middle of the pandemic in 2020. And we put it down for all the right reasons. And and you see that this slide is really a call to a, we’ll be unpacking that in much greater detail a week from today, same time, same place, perhaps some fresh faces, to throw into the mix, but, or is simply put is an IoT accelerator, and can be applied to any environment, harsh or calm. And so make note of that. And please join us if your schedule allows, we would, we would appreciate it. So with that, we now turn our attention to questions, I’ve got a stack that have been given to me, I know pre reading, no screening, so you’re going to get a cold. All right. And I’m confident you’ll be able to respond effectively. Here’s one from the great state of Florida, what options exist to keep an increasing number of devices connected, while in outdoor hard to reach environments, give me some options.

Kurt Dykema 46:57
Well, cellular is a great option, Wi Fi is a great option, there are more on the horizon. Within the AWS offering, there’s all the tooling you need to scale large fleets, you may need to put some time and effort into equipping your team to manage large fleets well. Not only do you need to operationalize the cloud, you also need to operationalize the fleet that you’ve deployed into the, into the physical environment. But those are all problems that have been solved. And so like Ben mentioned earlier, you can step on the shoulders of giants and not have to invent or to any take any significant technical risk to deploy connected devices via cellular. And there’s various forms of cellular now that you can look at depending on how much data and how much how much cost you can bear for the for transporting that data, Wi Fi. And then Ethernet is a great option too. But I think too, depending on the density, right of that fleet, and how many sensors you have in close proximity, you also see patterns like lorawan, where you have a gateway that’s managing multiple sensors talking to the cloud, you see those things deployed as well. And to kind of pile on curve. To your point, I think you look at the holistic aspects of how do you manage a fleet like that? The platform, obviously, and having, you know, a platform that gives you the intelligence, the DevOps solutions, to make sure that that platform is healthy, right, much less the devices themselves, but then also thinking about the experience of managing those devices, how do I get insight into their health? And how do I make sure that individual device by device within the fleet, things are in a good state, and the more time that’s spent on that holistically, you know, you can reduce and offset your time to maintain and to drive in the future? So looking at things holistically, there, I think is very key. Right? Great.

Bob Niemiec 49:08
Let me throw another one at you, gentlemen, are most of your customers using a cloud platform? or developing a custom cloud? Talk to me on that?

Kurt Dykema 49:23
I can. Absolutely. Yeah, I think you see different positioning as you look across customers and across verticals, right. There are circumstances where the economics of a deployment, there may be folks who say, you know what, I really have a fairly simplistic use case, I don’t need custom hardware developed, I can use something off the shelf and I can deploy that maybe using a sass platform on the back end. In other cases, the needs are very unique. There are very specific constraints that come to the table, or on the physical device or on the conductivity or on things like that. And it just naturally leads down the path of saying, let’s develop something custom can meet these needs because we see the longevity of the business case. Also, this is someone’s in an industry vertical, industrial vertical, excuse me, versus a consumer vertical. You can also see different dimensions emerge there in terms of how do fleet sizes. And you know, when I’m running millions of devices on a SaaS platform, how does the cost model scale and the business case scaled us since earlier point, versus saying, if I can build some assets, I can spend some engineering time and I can own that platform. Over time, if my device fleet scales into the millions, my fixed cost of what I’ve built here is relatively standard and doesn’t scale as much. So the economics can come into that picture as well. I think maybe one closing thought is the fact that T co isn’t just your software costs. It’s your support ecosystem around it. It’s all these other factors that come to the table to say, How do I build my support my platform, I don’t know if others want to layer on to that, but there’s some reactions there. I’d like to throw out an observation relative to economics. Austin, maybe you could give us a sense of scale from from the AWS perspective, but from my See, when I look at the amount of investment and energy being put into the AWS offering, to me, what that is doing is dropping the barrier for clients to own their own cloud infrastructure, there making it easier and cheaper to do that every quarter. And so if for any one of our clients who are looking at getting in this space, we would recommend taking a hard look and doing the numbers that owning your own cloud infrastructures simply because Amazon is doing a great job of making that easier and cost effective.

Austin Ashe 51:59
Yeah, I would, I would echo that, Kurt. And I’ll offer some some statistics. For our listeners, um, you know, we see, on average, a 415%, ROI, over five years, from investing in a digital transformation of the product or the process. And so having that cloud capability there that is flexible and able to, to grow and scale with, with those company needs is what drives that that higher return. another statistic for you is, on average, we see around $4 million of company added revenue to their to their new business after they have built and deployed their cloud solution. And so we encourage people to think of it that way, that it’s not necessarily a cost or a burden. But it’s an actual opportunity to transform the way that those business operations work. So not only are you that 415% ROI, $4 million averaging in added company revenue per year, you know, and where it comes from, is this notion of leveraging the data that has never been leveraged before, I mean, what we see is on average 67% of data available goes unleveraged. And that is just an astronomical number to think about 67% of your data goes on leverage, does that mean? Well, getting access to the data, having the ability to control it, and manage it, own it, and make it actionable with your evolving business model is key there. So we encourage and try to help people understand that owning your own cloud infrastructure is actually quite advantageous and a requirement if you want to continue to innovate, and reinvent yourself and reinvent our company. So powerful numbers here. That’s fantastic.

Bob Niemiec 54:08
Sure, for sure. How do you accomplish another question? predictive capabilities within IoT?

Kurt Dykema 54:18
So predictive capabilities, right? I think when you look at that, I’ll take a stab at it. And you guys can later on, when you look at predictive capabilities, a lot of that is you’re thinking about this data that I’m collecting, right? And looking at the micro trends and the macro trends across the data to say, what’s, what anomalies are emerging, right? what’s changing in this data ecosystem? And using things like machine learning to say, not only do I have to know what patterns have been there, and what, you know, guided patterns, I’m anticipating the algorithm to design around the sport. But how do I let the power of the cloud go to work for me into discover as well to look at those patterns. So the predictive power A lot of times it’s how do you apply and use that lake of data that you’re building around these solutions to start to look for those patterns and to drive those forward, you hear about guided learning, and then non guided learning and, and those different terminologies that come to bear. And that’s really talking about those machine learning capabilities. Austin Kurt, you guys an ad. So the label we’ve put on our particular approach to that is data driven algorithm development. And like Ben said, it supports supervised learning or unser supervised learning, but I guess the concept I would throw to this group is the idea of ground truth. And for us, ground truth is a data set along with a verified observation. So for Charity Water, it was a dataset of magnetometer readings, and waterflow ratings, coupled with a report from the mechanics in the field who actually could verify Yes, this pump was failing. That is one element of ground truth. And that’s where the power is, if you can gather ground truth, then you can train your algorithms, and you can go from monitoring to predictive. But it’s really helpful if you have a way to gather ground truth. And those ground truth data sets are really, really valuable. Think of them also as a competitive advantage for your business. If you have a really rich data set that includes ground truth, and your and your competitors have to, in order to catch you have to gather ground truth. You have a nice, you had a nice lead on them in the market.

Austin Ashe 56:50
Great, great. Yeah, great. What so what I would just add real quick, Bob is that, from our perspective, the the bar is getting needs to be lowered for what it takes to develop predictive algorithms and analytics. A lot of people think of it as if I need to do this and do this right, I need to hire data scientists and an elite group of extremely skilled and smart people to figure out how to do this. And that may be the case in some circumstances. But what we’re trying to do is invest in services that help automate that task and make it simple to start with simple things like anomaly detection, you know, if my process runs between zero and 10, but all of a sudden, I start reading 12 and 13. I know that my process is technically out of control, I have it. And as long as I have the instrumentation and the ability to collect that anomaly data point, then I can start investigating, what is it that’s causing that that anomaly, and then you can start making predictions based on that on what’s going to happen next. So there is an iterative process. And it can be as simple as taking a service that almost requires no code to get your anomalies detected to something as more advanced as developing custom algorithms that do in fact, predict failures or other outcomes.

Bob Niemiec 58:12
Great, great job, I feel like want to be respectful for time are coming in hot. If we could. Let me just say with one minute remaining. You know, today’s NFL Draft day, there’s a lot of noise about who’s number one who’s number one, I just want to say to you three panelists that in my eyes, you know, you’ve performed like a number one. Alright, so you may have been drafted in different years, some of you are far older than a younger. Number one, like performance. So thanks to you three for doing that. And again, doing that on the fly, and off the cuff. And then let me just say to the participants who have hung with us for the 60 minute sprint, you know, again, well and where we started by saying thank you. Thank you. Thank you. You know, Austin referenced a few IoT laws. Well, there’s a, there’s a well known coach that perhaps we’ve all heard of, by the name of john wooden, that giant from UCLA, and he had a few laws that he use to navigate his life day to day one was be a tireless and endless learner. And for those of you who are participating in this, maybe that’s your example of following a woman’s law and encouragement to always be learning. And hopefully you can walk away from this with at least one if not several. nuggets, gold nuggets in your pocket. And then the other that I think I can say, spreads the cost us four on this call is always have a grateful heart. So on behalf of the four of us, we just want to say one last time Thank you for being part of the journey. Don’t forget about this other event focused on Oris in a week’s time, and yeah, our well wishes to each and every one of you. Thank you. Thank you. Thank you very much.


A monthly newsletter focusing on the intersection of global macro-trends, strategy, and technology.


Digital transformation,
with a twist.

Create unprecedented impact with Twisthink.
Sign up for our monthly newsletter focusing on the intersection of global macro-trends, strategy, and technology.
Sign up