IoT Device Development Gone Wrong and How to Fix It

by | Oct 28, 2020

Connected device development catastrophes can happen – we’ve seen our fair share of them! Why do they happen, and how can your products avoid doing the same? What are some buzzwords that you’ve been hearing, but don’t know truly how to apply the technology?

We invited professionals within our marketplace to discuss the best practices of developing the hardware and software aspects of a product:

Hugo, the Lead Technologist of Untitled Kingdom, is a software engineer with over a decade of experience. In his work, he believes that the quality of the code is the main key to success. He is a security evangelist and has spoken at multiple IoT and startup events within London, Poland, and Silicon Valley.

Alexey, the CEO of EMBIQ, is an expert in embedded and mobile applications since 2003.

Ron is a US inventor and founder with over 20 years of product development experience with an emphasis on hardware. He is currently the Director of Connected Solutions at Breadware, an IoT-focused design consultancy, and the founder of GroupGets.com, a hardware-focused crowd purchasing platform.

Click on “Ask a Question” to jump to a specific time in the conversation to hear an answer to a question you want to know!

00:05 Danny DeLaveaga: Alright, I think we’re live. Hello, hello, thank you so much for joining today. And I know it’s night and morning, so thank you.

00:12 Bartek Hugo Trzciński: Good morning everyone.

00:15 Danny DeLaveaga: Yeah, great to see you Hugo, Alexey, Ron. So I think we’ll get this kicked off with just some brief introductions. I’ll talk about each of you in turn, and then let you talk a little bit about the company and go from there. But is everything good? Your mics are working? Alexey, it looks like Alexey and Hugo, you guys are in your offices, it looks like?

00:37 Bartek Hugo Trzciński: Yeah, that’s true.

00:38 Alexey Shabalovskiy: We are.

00:39 Bartek Hugo Trzciński: Alright.

00:39 Alexey Shabalovskiy: Even though it’s late and no one’s here, we’re still at the office.

[chuckle]

00:44 Danny DeLaveaga: And then Ron, you’re in your office as well? Or you’re at your home, home office?

00:47 Ron Justin: Home office. Yes.

[chuckle]

00:48 Danny DeLaveaga: Awesome! So thanks a bunch for joining today. And I think I’ll go ahead and start… I’ll start with you, Ron. So Ron Justin, I’ve known you for quite sometime. I really appreciate you coming on to talk about bad things that could happen for IoT product development initiatives. But just to get started, you got your electrical engineering degree from UNLV and actually, now, you’re serving on the advisory board there, correct?

01:17 Ron Justin: Mm-hmm. Yes.

01:17 Danny DeLaveaga: And then you kinda just got jumped right into entrepreneurialism. Well, you worked as an angel engineer at TrackR, which is an IoT product.

01:27 Ron Justin: Mm-hmm. Yeah, a very early IoT product and in fact, potentially too early. Back in 2008 to be honest.

01:32 Danny DeLaveaga: Yeah.

01:33 Ron Justin: From a business sense, but it was a very cool concept.

01:37 Danny DeLaveaga: And it’s a cool story of how that company grew and then had some problems with growing, but still…

01:44 Ron Justin: Yup.

01:45 Danny DeLaveaga: So you got jumped right into the entrepreneurial side, and then you went back to defense, the US Department of Defense and a couple of different realms there?

01:53 Ron Justin: Yeah, yeah, Department of Energy. And in fact, that was always my main hustle right out of college. I went right out of college to working underground for the Department of Energy. Fascinating work, lots of cool physics experimentation. And then on the side, I was sort of working doing these entrepreneurial things with TrackR and whatnot.

02:13 Danny DeLaveaga: And then you got more embedded in the electrical space with GroupGets, which is a really interesting concept. So that’s a company that you started with some co-founders that’s still going, correct, on the side, as well as Breadware?

02:25 Ron Justin: It’s alive and thriving, absolutely. And it’s strange, and I feel almost guilty, but COVID was a boost for both GroupGets and Breadware, and we’ll get into that a little bit later.

02:39 Danny DeLaveaga: Awesome! And then the main role now is with Breadware as the president there, or it’s you were the president when I was kinda working with GroupGets, and then shifted over through the acquisition from StoneAge.

02:53 Ron Justin: Yeah, yeah, yeah, so a little bit of a title change, same role. So I’m the Director of Solutions. I help do the solution architecture of IoT projects, and then draft up the contracts, work with the engineers, and sort of get our customers going.

03:10 Danny DeLaveaga: And just to recap, Breadware is an electrical engineering and embedded development product development firm located in Reno, Nevada. And you guys are thriving and growing even today, correct? I actually got to see some of the guys there recently.

[chuckle]

03:29 Ron Justin: Yeah, yeah, obviously, we have to thank you for being a founder of Breadware, but the acquisition has been… It has been good for us. We’ve had some good opportunities this year related to COVID and making touchless interfaces and things like that. So things that we never really foresaw happening are… We have four projects now due to COVID.

03:53 Danny DeLaveaga: Awesome! And so Alexey, the last time I saw you was is actually in Lublin, which was a ton of fun at your office, which is amazing. So that was a really cool trip. Got to go out to Poland and see you guys. And Lublin is a really amazing town. It’s kind of a university town from what it looked like. Tons of development going on there.

04:13 Alexey Shabalovskiy: It is. It is a lot of students here and a lot of student inventions and ideas together with students here in Lublin. Yeah, and that was very nice to see you here in our office Danny.

04:27 Danny DeLaveaga: Yeah, yeah, I was very impressed. It was a lot of fun. And so let me give you a quick background, and correct me if anything is off, but you got your Master’s of Computer Science from Wroclaw University of Science and Technology. And then you worked in Symbian OS, tech application development for Psiloc, before going into Sentivision and doing embedded drivers. And then actually heading a division at Gadu-Gadu. Can you tell me a little bit about that?

04:57 Alexey Shabalovskiy: Of course, yeah. So my, well, professional career started from mobile apps before they called that mobile apps ’cause well, it was Symbian, and it’s so many years ago. And then I went further into, well, firmware driver developers and embedded systems. Later on, still worked on mobile devices, which are most closer to embedded devices I was able to reach. And well, right now, I have a company, which is specialized with embedded devices, firmware development and together with that, integrations, mobile applications as well, web services. So a lot of software and firmware, actually, this…

05:53 Danny DeLaveaga: And so then… And EMBIQ, this is a… So around 15 years old now?

06:01 Alexey Shabalovskiy: Not so many. We are 11 years old.

06:04 Danny DeLaveaga: Okay, 11. Great! Very established. Yeah, in the mobile technologies, web services. Actually, you guys have a small hardware division that I saw as well, which is amazing. And so it sounds like…

06:20 Alexey Shabalovskiy: In fact, the hardware division is growing and growing. We are very much excited about that because we see the trends more frequently requires some piece of a hardware, sensor, microprocessor connection channel to be connected into the software part, and so we do that.

06:42 Danny DeLaveaga: Awesome. Well, thank you so much for joining us today. And, Hugo, thank you as well. Coming from Krakow, correct?

06:50 Bartek Hugo Trzciński: Yeah, that’s correct.

06:52 Danny DeLaveaga: Yes. And I was unable to visit your company, but we’ve seen you multiple times here in San Francisco as well as actually over here in Reno, Nevada. And so, let me give you a quick background and then let me know, but… So you have a degree in Applied Computer Science from AGH University of Science and Technology in Krakow before going into iOS development and design with Chwilemi? Or… And was that also in Krakow?

07:24 Bartek Hugo Trzciński: Yeah, so basically that was a… I was looking… I got into the Apple development because it got interesting when the first iPhone got released. So I was looking for a company where I can actually start doing something around Apple. So there was a guy in Krakow who was actually quite famous in the community, and I asked him about the opportunities to work with Apple device. And he said that, “There is a company in Poland and that is my company, so you can always come and visit.” So that’s how it all started. And particularly like 13 years later, the company still exists, but we changed the name to Untitled Kingdom at some point, and we’re doing pretty well. So that’s definitely one of the best emails that I wrote in my life.

08:03 Danny DeLaveaga: Oh, amazing. So you’ve kind of grown with this company since its beginnings?

08:07 Bartek Hugo Trzciński: Yeah, the Polish name Chwilemi, which is like moments wasn’t good when we actually went abroad with other partners to find, so we’ve thought that it’s actually might be a good idea to changing name. And actually, the name is quite a funny story as well because it was a shower idea that every project that you start on the computer basically, doesn’t matter if it’s a programming project or like Word project, it’s always called Untitled. And the Kingdom was there, the connection, because we were working basically with a lot of UK customers at the beginning. So that was the merge of the two ideas of the founder under the shower. So that’s how we came up with the idea of Untitled Kingdom.

08:47 Ron Justin: I always wondered, that’s very clever.

08:50 Danny DeLaveaga: Yeah, it’s amazing.

08:52 Alexey Shabalovskiy: Really on point.

[chuckle]

08:52 Bartek Hugo Trzciński: Thanks.

08:52 Danny DeLaveaga: And for Untitled Kingdom while we talk a little bit about that, it’s a well-developed software, iOS mobile, Android applications design firm. And then you guys have a lot of expertise in FemTech, actually, in digital health, software development that meets FDA, HIPAA, ISO 13485-type requirements. Correct?

09:16 Bartek Hugo Trzciński: That’s correct. So we jumped on the IoT train quite early with the mobile development. It was very interesting for us to have something physical that we can interact with while doing the software development. So that’s how we get on. But after some time, we decided that actually doing some weird and different applications, like 17th million version of Facebook. It wasn’t that interesting for us so we decided to focus on actually building the products that can have an impact on like every day people’s lives, good impact on the planet. So it’s trying to focus more on digital health. And the first products that we’ve done, like for example, Elvie, which is like seven years old on the market, which was the first… FemTech at the beginning wasn’t even there at that point so it was just called IoT product. But FemTech is one of the things that we do quite a lot, which is basically female technology that helps women with their lives, but it’s part of the bigger digital health that is the main focus of our company.

10:17 Danny DeLaveaga: Alright. Well, thank you. And I’m really honored to be here with you, three experienced veterans in the product development industry, and to be talking about what happens when things go wrong. So, yeah, I’ll just jump right in. And I think I’d like to address this question to everybody. But, yeah, in general, kinda what are some things that you’ve seen? What are some war stories? What happens when things go wrong? And maybe we can… Yeah, we’ll start with you, Hugo. And we’ll just do a little bit from everybody, and then we’ll dive in a little deeper.

10:53 Bartek Hugo Trzciński: Sure, thanks. So, yeah, we’ve seen quite a lot of stories of people coming to us with different ideas, different stages of projects. So this is always fun to get to know how the project started and why you are on that stage and how actually you plan to develop something else. But quite often, there are some problems that we want to… But sometimes it’s already too late because the hardware development is already quite far in production, so changing something in the work usually, it takes a lot of times. We need to try to figure out the ideas how to do something correctly, even though the basics, the fundamentals of the projects are not actually doing pretty well. So for example, again, with like choosing basically the wrong protocol at some point might be a big issue, so people tend to use the technologies that they’re quite known to, or at least they hear about those somewhere in the IoT world.

11:56 Bartek Hugo Trzciński: So for example, Bluetooth is one of those technologies, which is very popular, is very cool, and we like to use it as well. But it’s not always the best idea to integrate Bluetooth in your projects, and especially if you tend to control something live with sending loads of data. So we’ve been there, we’ve done that. And unfortunately, sometimes we had to spend quite a lot of time, which was actually quite interesting for us but maybe not the best for the partners because we still had to do some loop, some… Coffers basically for the code to do that to improve the quality of the connection and the connection speed. Once again, that is not meant to transfer a lot of data, so that’s really very limited, especially with the BLE. And people tend to use that at the beginning of the projects, which is not a very best idea without actually doing a lot of research. So that was on one of those. And thanks to our friends from Punch Through actually, we were able to increase the transfer speed for Bluetooth Low Energy, quite significantly.

13:04 Bartek Hugo Trzciński: So, for example, reducing the firmware updates from, like, 27 minutes to, like, seven, so that was pretty cool and impressive for us, so… A lot of things can go wrong, so we definitely need to do a lot of research before actually starting going into them…

13:19 Danny DeLaveaga: Oh, yeah. So, the choice of communication protocol really needs to be looked at carefully when getting started, just to understand the volume of data that is being passed, the OTA, how is this going to actually interact with the customer’s device, and I think we’re going to definitely come back around on this topic. Yeah, so Ron, what have you seen, what have you seen as…

13:44 Ron Justin: Well, I’ll build on the BLE example. So, a lot of problems I’ve seen are when somebody chooses a hybrid mobile app framework like React Native to do a Bluetooth communication to hardware, but they want a very reliable background connection, they want really fast data transfer, and what we’ve found is, it’s hard to do with a hybrid framework. It’s just better to go native. A lot of people don’t want to do native app development and maintain both an Android and an iOS app, it’s complicated, it’s more expensive, but if you really need a very great background connection and fast data transfer, we find it’s just the best way to go when you’re talking to hardware.

14:32 Bartek Hugo Trzciński: And usually when you choose the hardware, you actually need to still support the native part, so instead of one technology, you end up with three, so that’s another downside, so I completely agree with this one.

14:44 Danny DeLaveaga: And Alexey, yeah.

14:48 Alexey Shabalovskiy: Yeah, yeah, so answering directly to your question what can go wrong when you choose wrong protocol. Well, everything can go wrong and in particular, absolutely, I think Bluetooth here is a most favorite technology about some war stories or some troubles in implementation. I had lots of them. We took a part of resolving millions of different issues with Bluetooth and yeah, the thing which you guys mentioned that is sound research before choosing technology is absolutely first thing to do before you choose your global solution.

15:38 Danny DeLaveaga: So, Alexey, when companies are creating or using connected products, what kind of things should they be thinking about in the beginning in terms of certifications or should they be thinking about that? How should they go about selecting a development, for a product development initiative?

15:57 Alexey Shabalovskiy: Sure, in first place, I would say it’s very much important to find a partner or somebody who is experienced in the field to which you’re going to build application for, like when you’re trying to set up some set of sensors in some specific environment. It’s worth to find a company which is already experienced with that kind of sensor or that field of knowledge. Speaking about certificates, well, certificates are important while you already have chosen your technology which you are going to use in your solution, but mainly at the first, the initial stage of your product, at concept stage or feasibility study. It is very much important to not choose a particular technology or particular use case for that technology, but rather to look at your business need, your functional needs and then to fit technology to functional needs.

17:10 Alexey Shabalovskiy: So, at that moment you probably wouldn’t need any certificates yet. At the implementation stage, of course, and that depends on technology which you have chosen. For example, if you chose that your BLE device or IoT device has to be connected with Apple device, Apple headset, and Android headset, you would probably look for a company which has MFi certification and can support you with some extension of API or SDK. But before you decide how you’re going to use your product, well rather look for experience than certificates.

17:53 Danny DeLaveaga: Okay, great, so look for a company that has built something similar to what you’re trying to build.

18:00 Alexey Shabalovskiy:Right, right, but if you find, if you’ll find a company which built exactly similar thing which you’re going to build, think twice if you want to build it.

[laughter]

18:10 Danny DeLaveaga: Oh yeah, yeah. Alright, well, so, we’ve got some questions from the audience that I’d like to shoot at you and I think I’ll do this one to you Ron and then go to Hugo, but we would love to hear a couple of stories where things went wrong in a project and how they were fixed. So does anyone have any stories that they can tell around things that go wrong? I know Hugo you brushed into some mobile application and some Bluetooth issues, but yeah, let’s go ahead and start with you Ron.

[laughter]

18:45 Ron Justin: Not sure I wanna say here, but… Oh jeez, I mean, I’ll speak in general terms, just something we see a lot, like a theme we see a lot. Clients always want the smallest device possible, right, if it is wearable, doesn’t really even matter. Small seems to be generally desired by most clients, but there’s obvious trade-off for small, that means you’re gonna have a small battery, less battery life. Also means you’re likely going to have non-optimal RF performance, you know, we’re in analog world. Most IoT projects have wireless. Wireless is analog. Antennas love ground plane. If we have to make a very tiny device, that means we’re going to have a small ground plane, we still cannot break physics in 2020 and therefore, your RF communication just won’t be as good with a tiny bit of ground around the antenna. So, these are the general starts of how things can go wrong. And the solution, at the end of day, it ends up being, “Oh, we need to make it a little bigger.” And so we try to communicate that upfront to try to avoid that. And so that’s my general war story.

20:03 Danny DeLaveaga: And communicating upfront. What do you mean by that? Does the initiative, in most cases, they have a prototype, or when does it come to benefit?

20:14 Ron Justin: Well, when it gets something from an industrial designer that’s the size of a dime, and the marketing team says it’s gonna last three years on a tiny coinsaw and it’s the size of a dime, and then we’re like, “Whoa, whoa, whoa, we need to expand this out a little bit,” or communicate that the battery life is gonna be a week as opposed to years.

20:38 Danny DeLaveaga: Awesome. So thank you. And so, in the design phase, the earlier the better. Talk about these things. Talk about the pros and cons. What you’re getting into if you do need to make these big compromises in size, or in battery life, or in RF performance? So yeah. Hugo, do you have any stories around product that you could relate?

21:02 Bartek Hugo Trzciński: I can actually relate… Yeah. I can definitely relate to that, but I could tell maybe a story without saying too much. So we’ve been also working with different protocols, even slower and even less robust, way more less robust than Bluetooth. And the problem was that the device that we’re controlling basically was accepting a lot of commands. So the development team that stayed behind the hardware was actually doing whatever they were told to do, so they implemented all the functionalities, all the possible commands that we needed to interact with to be able to actually control the device.

21:41 Bartek Hugo Trzciński: But once we got into the actual development and controlling the device, it turns out that making those calls takes a lot of time. And because those are just simple commands, we need to do like 16 in a row, which was taking four seconds. So instead of actually creating one command to do all of those internally, I would have to call like 16 of those. So there wasn’t like anything wrong done by the hardware team because they were told to prepare those commands to be able to control the device, but no one thought that at the beginning, it was gonna be the outcome for the product and for the end-users.

22:18 Bartek Hugo Trzciński: So from now on, we actually try to go through the whole process for the user, and try to see what are the things that the user is gonna be interacting with. Because if the firmware update is gonna take like 30 minutes and it’s gonna be done once a year, that’s fine. But if the user wants to perform a simple single command to a device, and this is gonna be taking seconds and blocking the UI for him, that’s definitely not good for a user and for the product. So the communication here was actually correct, but it’s still good to look from the outside, from the non-development perspective, but from the user’s perspective, how it’s gonna look and what is needed to be done to those situations. So that definitely was a lesson for us. And now, in a development process, we have a meeting when we try to mimic the user’s behavior and try to figure out how the application is gonna behave, yeah.

23:12 Danny DeLaveaga: Awesome, yeah. So, Alexey, have you seen anything go wrong?

23:23 Alexey Shabalovskiy: Yeah, absolutely. Absolutely. Like a few years ago before iBeacon became popular, we had a solution of asset tracking for one of our customer. And well, that was the case when somebody heard about technology and found application for the technology based on the technology itself, not on the market demand. And well the issue was that obviously, iBeacon behaves so sometimes, it loses the range and at that stage of BLE technology, it wasn’t able to be connected. And so, I think everyone here experienced that kind of things with the BLE devices.

24:21 Alexey Shabalovskiy: Our solution was to review firmware and to review connection types and the broadcasting types. So we proposed a change in the device firmware. In that case actually, it wasn’t device manufacturer, but it was an external company. But we developed concepts for them how they can change their product to improve stability of device recognition and connection, and they introduced that fix, and will release it for a broader range. And yeah, actually, the solution was quite happy. Right now, we have a lot of asset tracking based on BLE, and that particular solution was sold to another company. So it was a war story with happy ending, yeah.

25:15 Danny DeLaveaga: That’s good. So I have another audience question with regards to testing actually. So if the early prototypes, at some point, don’t ignite on fire and they were tested, they weren’t tested enough, is this correct? Yeah, this is for the group. If anyone would like to take it.

25:39 Alexey Shabalovskiy: I would say regarding the testing itself, testing hardware, it’s very interesting, a very interesting thing because there are a lot of different types of how to test hardware, starting from unit test when you test a firmware on a stage of development through automated test when you have connected device to your hardware and you roll some activities on that hardware on and on again. And on the robotic test when you have your hardware, you have a robotic arm and do some physical activities on that device. And absolutely, then that’s actually our most favorite field of hardware development, is to test it until it [26:28] ____. Absolutely, that’s a good idea to test it. I tested a lot of different stages of development, the usage and the real usage, so…

26:44 Danny DeLaveaga: Also, talk a little bit about the difference between hardware and software testing and the types of testing that you do it at Untitled Kingdom. Hugo?

26:52 Bartek Hugo Trzciński: Yeah, for sure. So one funny story also, we do a lot of testing and the QA team is also growing pretty quickly. But because we’re working with a lot of IoT devices which are making a lot of different noises, beeps or other sounds, it’s usually quite funny sometimes staying in the open space and hearing all those things for a whole day, so we’re also testing ourselves and our patience. But for sure, especially for the software testing, we start early with the unit testing so to be able to…

27:28 Bartek Hugo Trzciński: Say that whatever we’re gonna do according to the documentation is gonna work, so there’s another thing, so when there is no hardware to test with, we also have a method for that, so usually for the IoT projects, we try to create some kind of simulator that basically mimics the actual device based on the documentation that the partner provided us with so once that is done, actually when the final product is ready and we got a hands-on that as well. It’s a very good two-point verification, so because we were able to connect to the simulator that was built on top of the recommendations and then we got actually the real device, it’s a good point to check if everything is correct or the documentation was correct.

28:20 Bartek Hugo Trzciński: Usually, it takes less than a day to make sure that we can connect to the device and everything works. So that’s pretty good. Of course, the team is creating a lot of user stories trying to figure out the worst possible monkey testing in the app figuring how that can happen, which is always a funny story. But also recently, we tried to also do a lot of hardware automation testing with the hardware. So we basically try to build our own CI internally, I think Alexey said before, that we basically have some kind of mechanical arm or some kind of Arduino programs that are trying to make sure that the device is in the correct state because sometimes, again, you need to somehow make it discoverable or press a button or something like that. So if there are no [29:07] ____ that we can interact with, we try to figure out something of easily to be able to test that automatically because again, sometimes doing a workup for example with LV took 7 minutes and we had to do like seven of those, again takes a lot of time, so we need automate that for sure.

29:30 Danny DeLaveaga: Alright, well, thank you very much. I guess you kind of briefly touched on it, but on the security side of things, where does testing there come in, on the hardware and the software side, maybe Ron, you could talk a little bit about this to start.

29:50 Ron Justin: Sure, regarding we try to take security first on day one with the design, so we try to choose chips with secure enclaves, chips that support encryption, protocols that support encryption. We use the tools that are available to us. But at the end of the day, we do not think that it’s prudent for the design agency to also be solely responsible for the security testing, so we typically will rely on a third party to get an independent security audit. And I just think that’s the best practice going forward.

30:28 Danny DeLaveaga: Alright, so I think we’ll jump into a couple of more questions, and then I do wanna definitely give everyone the chance to talk a little bit about what they’re seeing in the market right now with product development. But before we get into that, let’s go ahead and talk about successes, where are you seeing success right now with product development initiatives that you’re working on? So let’s go ahead and start with Alexey.

31:01 Alexey Shabalovskiy: Yeah, right. About the general trends in product development in terms of IoT, we see that a few years ago, it was pretty much popular to have different gadgets based on wearing devices. It was popular, and it’s still pretty much okay but right now we see the trends that a lot of smart cities, smart agriculture, smart manufacturing, and smart sensing productive maintenance and so on and so on. A lot of that kind of inventions are heading to the market and having a real demand on market actually, and in terms of that kind of solutions what we had recently, we’ve built employee optimization and monitoring system for greenhouses where… Well, the solution was intended to optimize workflow in greenhouses where you have a huge area to cover with huge precision of where the employee is and what he does actually and whether he does, what he intended to do.

32:28 Alexey Shabalovskiy: Yeah, we’ve built the solution based on ultra wide-band technology. We covered all the greenhouses, huge area, the anchors built the wide band for that solution and provided the solution for our customer. And what they did well in their product, they did a huge research before starting building anything, it was a concept stage, very much… Well, very significant concept stage, it was feasibility study comparing of different localization technologies in different environmental, well, opportunities like humidity like plants around, like people, different weather conditions on the outside of lab and so on and so on. And at the end, it was POC, and POC proved that the concept is right. So right now, they’re happy and they are using the solution which they wanted us to build for them.

33:43 Danny DeLaveaga: Great, thank you. And you kind of brushed up against a couple of things. So you’re doing some, it’s not machine learning, but you’re using ultra wide-band protocol to detect nodes and people moving and then doing some analytics based off of that, but what about AI, the buzzwords, ML, 5G… What are you guys seeing in the market, what are your predictions around these kind of technologies? Yeah. We’ll go ahead and go with you Ron.

34:13 Ron Justin: I think we need to slow our roll on 5G a little bit, it’s a great buzzword, if you wanna buy a 5G module for an IoT project today, it’s gonna cost you $200. Most IoT devices can’t handle that budget, unless you’re doing a luxury car or a smartphone. Also the coverage, right? Reno has 5G towers, but then you look at the fine print, they’re not the real millimeter wave 5G, it’s just a slightly better version of 4G based on the bands it’s using and… This is the T-Mobile network. So I think the marketing is way ahead of reality on 5G, unless you live in select markets and you can afford it.

34:58 Danny DeLaveaga: Okay, yeah. How about you Hugo, how many of the devices that you guys are working on or the initiatives have some sort of AI or ML type work or part of the software?

35:16 Bartek Hugo Trzciński: Yeah, so maybe less of the AI. We actually have one particular with AI, but machine learning is the more common one. So it’s quite understandable that all the IoTs, all the wearables generate a lot of data, so sometimes, for example, again, with LV, there was actually no studies before those devices for women, so there was no data how actually human body behaves. So that was actually a good chance to use that data and try to understand how we can make those workshops for the women better and make them perform even better month after month.

35:54 Bartek Hugo Trzciński: So this is definitely… That’s something, it’s not maybe new or it’s not gonna be new in the future, but it’s definitely growing because the amount of data that we have is growing every day. So this is definitely something that’s gonna be worth checking out, if you’re still not on the train of machine learning, then you probably should be pretty soon, I guess.

36:15 Ron Justin: Yeah, I agree.

36:17 Alexey Shabalovskiy: Absolutely.

36:20 Bartek Hugo Trzciński: The other thing is actually sometimes generating too much data, which might be a problem as well, because you need to transfer that data and also store it somewhere, so it’s also a pretty good idea to review what kind of data are you storing, sometimes the data is actually under some local legislations, so to make sure that when you store that data, you actually protect them pretty well. So it’s again wise to be sure what you’re doing and why you’re doing that, so it starts with why. So it’s very good idea to understand what you’re doing and why.

36:56 Ron Justin: Absolutely, and I also think if you can do some of this machine learning at the edge, so you don’t have to transfer all that data and you can do more filtering locally, right, with tiny ML and some of these new generation RISC-V processors from GreenWave Technologies, the GAP8 that does, it is designed from the ground up to do machine learning at the edge on a battery, and so we have just successfully finished a project using that, and we were very impressed, [A] with the customer support, [B] with the performance and our customer is honestly thrilled.

37:31 Danny DeLaveaga: That actually jumps right into another question I’d love to kinda get in on. What are your teams working on right now? What’s exciting you guys with these product development initiatives that you’re working on? And we’ll go ahead and start with you, Alexey, and then go back to Ron and then to Hugo.

37:50 Alexey Shabalovskiy: Of course, speaking about the works right now here in EMBIQ well, mostly we work for companies who would not wish us to tell about what we do for them, but one of the most favorite, maybe my most favorite project which we do for ourselves with collaboration with external company is footfall analytics in public transportation, where we build an invention to monitor human flow of people, how they use public transportation, where they change stops, how long they wait for the bus and when they return, which days they use public transportation and so on and so on. So a lot of data or big data, this is what Hugo mentioned that one opportunity is to select the data which we want to collect, another opportunity is to process that data in a correct way, to generate feasibility reports about the public transportation which we worked on and…

39:06 Alexey Shabalovskiy: By the way, speaking about 5G and the narrow band technology, this is one of the technology which we are pretty much looking forward to have it, and be able to use it fully, because of abilities to process data on place, not on the device, but on the broadcasting stations. So it’s not possible yet, they advertise it, that it is going to be available, but so far it isn’t. Right.

39:47 Danny DeLaveaga: Thank you. Yeah. And Ron…

39:50 Ron Justin: We’re pretty excited about some of the COVID-related projects we’re working on to automatically disinfect surfaces and the air. We have another project where we’re adding LTE cellular to a really popular ventilator that’s being mass produced right now in response to COVID. So helping this ventilator talk to the cloud, communicate data to the manufacturer, also track it if it’s on the go because this ventilator is actually portable for people. So we feel good about that type of work that we’re helping people and responding to this pandemic.

40:33 Danny DeLaveaga: And are you seeing a lot of that kind of work as well? The OEM that has a device in the market that needs to communicate a little bit more with the device or in communications via cellular.

40:49 Ron Justin: Yeah. All the time. Actually, one of my favorite projects, it was not a COVID project, is we digitally enabled a hose crimper for Gates Rubber Hose Company. It’s a machine that puts fittings on the ends of hoses. And this machine, the design was from the ’70s and the client could no longer even buy the electronics to build more. So we came up with an idea to add a IoT retrofit. So the cool thing about this was most of the machine didn’t have to be wasted. We redesigned the shroud that the user, the customer, could just replace the old shroud with a new one. And then boom, it has wireless connectivity to the Gates Cloud to get user guides, marketing information, help, all sorts of information. And this in turn allowed Gates to understand where are these machines ending up? How often are they being used? Who are my super users? Things like that. And that was something we were really proud of because it hit on sustainability. We didn’t have to scrap a bunch of machines, and these were in the tens of thousands. So that was a good example of how IoT can be successful.

42:04 Danny DeLaveaga: And Hugo, what’s going on in…

42:07 Bartek Hugo Trzciński: In this…

42:11 Danny DeLaveaga: Oh yeah. Go ahead.

42:13 Bartek Hugo Trzciński: Because of the whole COVID unfortunate situations, the digital health industry is actually booming even more than it was. And it was already booming quite a lot. So there’s a lot of things going on. So we try to do a lot of projects that basically take a lot of things that you need to do, with a doctor or with an appointment going somewhere else that… More and more things that you can bring home and do it yourselves is basically beneficial for everyone because it’s not that not that easy to infect someone else in the current period. And in the future also it decreases the amount of work that the doctor needs to do. So there is actually a lot of projects right now coming around the digital health. And we’re happy about it, and there’s a lot of interesting stuff.

43:00 Bartek Hugo Trzciński: But this is one of the things we do. We also experiment with the new technologies. For example, one of those is AR and VR. So we try to figure out a way to… It’s kind of internal projects, but not really. It’s a complicated situation. But we’ll try to figure out how to connect people in those crazy times, even though they are working remotely. So we’ll try to do something with the office and the VRs. We actually got the Oculus 2 today to play with. So there’s plenty of stuff coming. We’re actually planning to introduce the whole website page for our new [43:41] ____, basically aim to do that. So stay tuned, I will try to inform you guys when that’s ready so we can share more information about that.

43:52 Danny DeLaveaga: Awesome. And I have one more… Or I have another question. So when pulling a product through the development initiative, how important is it to run pilot tests locally? And what have you seen around, specifically hardware devices that have some software elements… What have you seen around the initial pilots before these products go to production, and engineering validation, and design validation? So I guess if anyone has any expertise around running initial pilot testing before taking a product to mass production, I’d love to hear a little bit about that.

44:42 Alexey Shabalovskiy: As for EMBIQ actually, the pilot stage is one of stage of development actually. Every product, whether it started in our company or it was born somewhere… In external minds. It starts from concept where you’re trying to figure out what you’re going to do and what actually you’re going to deliver. Then strong feasibility studies, so you’re trying to fit your idea into technological abilities. To fit technology into what you are going to achieve. This is very much important stage of project development. Then POC, when you build what you found that you want to build with chosen technologies. Then pilot, which you’ve mentioned, which is very much important to launch your product in the small group of people, and they’ll have to use it in real environment, not only in your lab. And to see how they use it actually, what are difficulties, what are problems, possible problems or maybe some hazards with the device when it stops working. Because it also happens because of the…

46:03 Ron Justin: And there is no other way to find that corner cases instead of doing pilot before you go to masses and produce tens or hundreds of thousands of devices and to start selling them. So, pilots, absolutely.

46:27 Ron Justin: Yeah. We build that into our contract, a pilot phase where we don’t iterate the design until the customer pilots, and we even highly advise them to use some beta customers to pilot. Not just their employees, but actual customers because they are gonna do things that the employees or them or us are just not gonna do, right. And so I think that’s a critical piece.

46:53 Alexey Shabalovskiy: Absolutely.

46:53 Bartek Hugo Trzciński: Can I actually ask how big are the pilots that your guys are doing?

47:01 Ron Justin: It’s like…

47:03 Alexey Shabalovskiy: In our case, it starts from 100 devices, but, okay, it depends on the particular applications. So if you have small wearable device you can hand that device to different people, and so you can have the pilot stage [47:25] ____ devices, I think. But if you build something for a smart city then it is quite hard to deploy each device in a real environment so you will probably stop at 10 devices in the real environment and see how it behaves in the real environment. This is our experience.

47:56 Danny DeLaveaga: Yeah.

47:56 Ron Justin: Yeah, I totally agree it’s application dependent. I mean I’ve seen successful pilots with as few as 10 units, right. As long as actual customers have their hands on it.

48:09 Danny DeLaveaga: Right. And how about… It’s probably different for consumer products versus industrial products or smart cities. I mean, yeah. What have you been seeing in the FemTech world, because I know you have multiple products on the market, actually, Hugo, that we’ve seen come from Untitled Kingdom, and can you talk a little bit about consumer product pilot testing?

48:35 Bartek Hugo Trzciński: Yes, so this might be a bit more difficult because finding a user group, for example for like a breath pump is a bit more difficult than finding a user group for a watch because you need good timing, and also it’s a very personal time for someone to actually to have those tests performed. So those tests usually aren’t that big because it’s more about the flow of the application and see what that user needs, and there are more tests being done in the lab, because again, this is something more difficult to do and the clinical testing that are need to be done during the, for the medical devices, again, defines the user group that’s there. But I wouldn’t say that those are hundreds of thousands, usually it’s more like dozens maybe.

49:26 Danny DeLaveaga: Alright, and I got one more question from the audience. John is asking, have you seen any good energy-harvesting technologies that can eliminate batteries?

49:39 Ron Justin: I’ll jump in. I mean other than solar, I assume you’re talking about something for a tiny device. I mean look, how long have people been talking about harvesting WiFi signals, right? Years. I mean I’ve seen products 10 years ago win innovation awards at CES that claim to harvest WiFi to power your device, but they’re not on the market today. There are… I believe it’s coming. I see some promising stuff and I can’t wait, I really can’t wait. All this wireless energy is in the air, like 2.4 gigahertz, 5 gigahertz, let’s harness it, right. But nothing that we feel comfortable building a product on yet.

50:25 Alexey Shabalovskiy: Yeah, from our perspective, we heard a lot of energy harvesting on different… Well applications like harvesting energy from your smartwatch, when you’re moving and a smartwatch can try to harvest that energy back from your movement. This is the small application, the biggest application which we heard about, it’s probably already installed somewhere in Europe, but I can’t recall where. Where they harvest energy from highway and cars passing. They harvest energy from airflow to fire bulbs on that highway. So, the energy harvesting very much interesting. Like Hugo mentioned, the energy is all around us besides of wind energy or tides energy or solar energy. There is a lot of energy which we use, but we can’t harvest, but I think it’s going to be bigger in the upcoming years. Right now, we are technologically not so much ready to do that on big scale.

51:54 Danny DeLaveaga: Awesome, so we’ve got a couple more questions and then I think we’ll wrap up. I’d like to hear about any news, any events happening in your world. So, it looks like we have a question around pilot testing again in smart cities. How do you run these initiatives and moreover, who bears the expenses? What have you seen in your past initiatives?

52:21 Alexey Shabalovskiy: Well, I would say it depends. It depends on the application. For example, if you’re building a product which is intended to gather a lot of data and to gain some profits from data, you will install that pilot on your expenses. You will try to find how to fit that product and to collect that data from, well from the environment. But in case, if you’re trying to build something new, something which is important locally, then you have chances to gain, well expenses from your customer. For example, from city or from highway maintenance and so on. So, it very much depends on what you are going to launch and how you are going to monetize your solution.

53:24 Danny DeLaveaga: Awesome thanks. Well, okay, I really appreciate your time, Hugo, Ron, Alexey, and I… Yeah, is there anything going on that you’d like to talk about? I implore anyone who’s looking to check out product development initiatives to check out the case studies on all of your websites. You guys, have all launched products to the market, so. This has been a really informative session, and I’ve had a lot of fun talking about some of the problems and initiatives so… Yeah what’s going on, what can people tune into in the upcoming months?

54:00 Alexey Shabalovskiy:Yeah, right. Thank you very much for inviting us, and it was very much interesting and informative to talk about that different solutions on the IOT scene and device development, software development. Thank you.

54:15 Ron Justin: Yeah, yeah, thank you Danny, you did a great job. And one thing I wanna plug is one of our clients, this is fun for us… They’re called Shine Bathroom and their product, that we helped develop will be on the Shark Tank TV show, this new season, which I think starts literally tomorrow. We don’t know the air date but that’s something we’re excited to tune in on and see our product on TV. We don’t know if they got the Shark Tank funding or not, they won’t tell us, but we’re excited, tune in and check it out.

54:44 Danny DeLaveaga: Awesome.

54:45 Bartek Hugo Trzciński: That’s interesting. From our side, we’re gonna be releasing a new product, one of the new products of our partners, early next year, so stay tuned for that. And also for the Untitled website that’s gonna tell you more about what we do it and what is the point of as well.

55:05 Bartek Hugo Trzciński: Amazing. Well, thank you so much for your time, and I look forward to the next one.

Ioterra – Where you go for IoT Development

Find the right solutions and services for your needs from the most comprehensive marketplace for IoT projects.