In this episode of the IoT For All Podcast, Richard Beranek, Managing Partner and Engineering Lead to Brash Inc, joins Ryan Chacon to discuss custom versus off-the-shelf IoT solutions. They cover what off-the-shelf and custom IoT solutions are, combining them, risks and things to consider, factors that influence the decision between them, and advice for companies starting their IoT journey.

Episode 333's Sponsor: Wilderness Labs

Meadow by Wilderness Labs is the only IoT developer platform that supports full .NET on microcontrollers out there. Our mission is to make building IoT hardware as easy as creating web or mobile apps. Get 10% off when you use the code IOTFORALL at our online store at store.wildernesslabs.co.

About Richard Beranek

Dr. Richard Beranek has a Bachelor of Engineering in Aerospace Engineering and a PhD in Mechanical Engineering, both from Carleton University. In 2012, he co-founded GaitTronics to commercialize SoloWalk, a robotic walker used in early mobilization rehabilitation. He would go on to form Brash Inc. in 2017, leading the product design and development of new devices, consumer electronics, robotics, software, and IoT products.

Interested in connecting with Richard? Reach out on LinkedIn!

About Brash Inc.

Brash Inc. is in the business of collectively finding creative solutions to interesting problems. Whether it's starting the next big idea or revamping an existing one, they partner with clients to deliver customized solutions from the initial scribble all the way to market (and anywhere in between). Their team of innovators include designers, engineers, software developers, data scientists, and researchers that are not only technical experts in their respective fields, but also well-versed in the startup journey and the various stages of product development. Alongside with you, they dig into developing products that fuse design and engineering to create intuitive customer experiences.

Key Questions and Topics from this Episode:

(00:31) Introduction to Richard Beranek and Brash Inc.

(02:16) What are off-the-shelf IoT solutions?

(04:11) Custom IoT solutions

(05:40) Hybrid approach: combining off-the-shelf and custom

(10:52) Impact of legacy infrastructure on IoT solution choice

(12:02) What other factors influence custom vs off-the-shelf?

(21:06) Advice for companies starting their IoT journey

(22:22) Learn more and follow up

Transcript:

- [Ryan] Welcome Richard to the IoT For All Podcast. Thanks for being here this week. 

- [Richard] Thanks for having me, Ryan. 

- [Ryan] Yeah, it's great to have you. Before we jump into our conversation that we have planned, it would be great if you could give an introduction about yourself and the company you work with to our audience.

- [Richard] Yeah, so I'm Richard Beranek. I'm a managing partner at Brash Product Development Inc. So we're a turnkey product development firm. What that means is we bring together a wide range of engineering, industrial design, and software development services. We do everything from electronics development, mechanical engineering, UI/UX design for software, enclosure design, and web mobile software and bring that all together in the projects that we do.

We have a couple of areas that we work in regularly. Medical devices is one, IoT is one obviously as well, consumer electronics. But otherwise we're pretty agnostic in terms of the projects we take on. As long as it's interesting, we'll bring those services together and build it out. 

- [Ryan] That's fantastic. Yeah, we're familiar with kind of the IoT side of the work you all do, and you have some impressive stuff out there. So, I wanted to jump into some things that I'm sure have been learnings from a lot of that experience that I think our audience is going to get a lot of value out of. 

And the main area that I wanted to talk about was off-the-shelf versus custom IoT solutions and what the difference means. I think there's a lot of talk in the market right now about the value and importance of end-to-end solutions. But when it comes to finding the right solution for your company, depending on who you talk to, they will have off-the-shelf, I should probably use that in air quotes, solutions that you can adopt and then ones that are custom. And then there's probably like a hybrid of some in some nature, right, where you have a large portion of the solution is ready to go, but a piece of it needs to be customized. 

So when we talk about off-the-shelf versus custom, if we start with off-the-shelf, what does it mean for an IoT solution to be considered off-the-shelf? What are the benefits, drawbacks, challenges, how does that fit in? And then we'll jump into the custom side in a second. 

- [Richard] You know, when it comes to off-the-shelf, I would give two main subcategories to that. One is hardware and the other one being software. On the hardware side, is it, the common hardware pieces in the IoT solution is the sensor, whatever piece is giving you the data that you want. And then there might be some communication infrastructure. Gateways, cellular communication, Wi-Fi communication that's getting that sensor data onto the cloud. In terms of off-the-shelf, is there a hardware piece that does that for you and gives you the functionality that you want in terms of getting the data that you want with the precision accuracy, update frequency, and all of that. And that goes both for the raw kind of sensor data that you might be taking and as well as the communication infrastructure. Those kind of go hand in hand in terms of those functionality that, the software bit of that is then accessing, displaying that data the way that you want it and whatever other logic you may want to run on it. 

Seeing the data is one thing, but in terms of web dashboards, mobile applications that allow you to access that data, there's so many requirements that go into it from how users are accessing it, and how you control that if you're setting up any kind of additional logic like alerts.

The off-the-shelf solutions are what hardware is out there that does that for you. And then on the software side, what are you able to get that plugs into that hardware that shows you the data the way that you want it to. 

- [Ryan] And on the other side of this, when we talk about custom, I know there are varying levels of customization that might be needed or required for a solution. How does that usually fit into the conversation? When is that something that companies lean more towards versus off-the-shelf and what are those benefits, drawbacks, kind of things that we should be thinking about when we hear the word custom IoT solutions? 

- [Richard] Yeah, and that's certainly when a company like Brash comes into play because most of the people coming to us have exhausted what's available off-the-shelf, and they've crossed that barrier and said, okay, whatever's out there isn't working for us, or it's not going to meet our functional requirements, and so we need to customize it to a certain degree. And that comes in a lot of different flavors on both of those subcategories that we mentioned, both hardware or software.

So, it can mean that, look, I need to measure a specific thing with a specific sensor at a specific rate, and whatever existing IoT sensors that are out there can't do it. It can be that I want a very specific dashboard solution or logic on the software side, and I want to marry that.

And it could be a hybrid as well, as you mentioned before, where there might be a lot of the pieces out there. There might be a sensor out there that does what you want it to do, but you want, you know, a customized dashboard that shows it in a different way, or you need to develop different firmware for that sensor to get it to operate exactly the way that you want it to.

And so most solutions are not either purely custom or purely off-the-shelf. They end up being a kind of a grab bag of what's available out there and then putting it together into a working solution.

- [Ryan] I know people when they hear the word custom, they often think, immediately think that this is going to be much more expensive or the process is going to be having to start from scratch. But it sounds like from your conversations, it doesn't necessarily always have to be the case. 

- [Richard] Yeah, there's certainly a whole spectrum out there. We've had customers across the whole spectrum where we've had people that wanted to start with a blank slate, wanted absolute control over the hardware communication back end and front end and wanted a fully enclosed ecosystem that way. And that's certainly your highest development cost path, right? And the trade off is that you get the functionality that you want, and you get to choose how every part of it works, and you have that kind of control. 

On the other end of the spectrum, it could be a customer that already has a dashboard. A lot of enterprise customers already have, you know, existing dashboards and tools to process and visualize data. So it's more we want a very specific sensor, right, and have hardware developed, or it can be even as simple as existing sensor, existing hardware, but the firmware just needs to be tweaked a little bit to get it to work the way that they want it to or work on a different communication protocol. So all of those kinds of situations arise. 

- [Ryan] And are there more risks associated with one or the other? I imagine with more off-the-shelf, it's tried and true, potentially has more, it's gone through the paces to be tested and then scaled out to be more off-the-shelf versus custom.

Is there any kind of long term drawbacks, long term benefits that we should be thinking about when evaluating the difference between custom and off-the-shelf? 

- [Richard] Yeah, certainly when you're buying something off-the-shelf, it comes with the support of the vendor, and you're buying a solution and a product that's being supported by an existing vendor. So I think that's the biggest part of it. If there's updates, if there's bugs, the vendor is dealing with all of the issues associated with that, and they're there to deal with that for you. 

As soon as you get into something custom developed, you become an owner and a manager of that product or that solution and everything that comes with it.

For hardware, that can be quite involved, right? And if you're building an IoT sensor, there's going to be a PCB there. There's going to be an enclosure there. Those need to get manufactured, assembled, packaged, sent out. Quality assurance has to be done on it. You're managing that supply chain and everything that goes along with it. So that can be quite involved. 

And for software, it's similar, right? You're developing a piece of software, you're running it, somebody has got to make sure that that software is upright and that it isn't crashing and dealing with bugs and dealing with support cases. You have, need, depending on what you're developing and how involved it is, you end up owning all of that infrastructure and needing to control it. 

- [Ryan] And you mentioned with the custom, you become the owner, need to manage it with, or I guess my question is for an organization that might go that route, do they need to have a certain type of internal team in place to support going custom or, and then on the same, on the other side of this, is a similar question, is there a type of infrastructure you need in place personnel wise to help decide if one is even capable versus the other two to take on for an organization or how does that play into the decision making process? 

- [Richard] Yeah, certainly if you're developing and deploying custom hardware or software, you need a team that manages those aspects to a certain degree. It doesn't have to be internally necessarily. A company like ours handles the design portion of it and to a certain extent as well the support of those products, right? If there's changes that are required or if bugs come up. Then there's the day to day operations, right? So for hardware, for manufacturing, you can outsource a lot of that and just get the final product delivered to you, and you're paying a certain premium for somebody to handle the manufacturing logistics and the quality once you know that design portion is finished. 

So that can be outsourced. It doesn't need to be internal, but you do, there's a minimum amount of kind of logistics overhead to deal with that. And similarly with software, most software, as long as it doesn't get too complicated or scaled too big, can run on its own in the cloud for the most part. But bugs come up, you want feature changes, all of that. And you need some minimal level of DevOps as well. Again, a lot of that can be outsourced. If it's a simple project that doesn't require a lot of scale, you can outsource a lot of that fairly easily, but then if it does scale, right, if it's really big, then you can hire that team internally as well. 

- [Ryan] So if we look at it from a different angle outside of the personnel and the team you would need, what about existing structure from a technology standpoint in legacy systems, other types of technology they already have implemented, how does that factor into the viability of an off-the-shelf versus custom kind of decision? 

- [Richard] It's a limitation on both sides whether you're going custom or off-the-shelf because that integration with existing systems, whether it's hardware that you're already using or software infrastructure that you already have in place becomes a requirement in that decision tree, right? And it can push you in one direction or the other, right? If you're already in a certain ecosystem for your IoT and software, then there might be off-the-shelf solutions available to you there, or it might be limiting in the sense that, hey, you really want to stay within this infrastructure that you know how you're managing your data and visualizing it, and so you need to customize whatever new thing you want to build out to be compatible with that. It ends up being a driving requirement in either direction. 

- [Ryan] That makes a lot of sense. And to round out this portion of the conversation, are there other requirements that influence the decision between going custom versus going off-the-shelf? We talked about the team structure internally. We talked about the existing infrastructure from a legacy system standpoint. I imagine cost, time, a lot of other things are factoring into decisions. What are the ones that really lead the way in the discussion that you have with companies?

- [Richard] Yeah, cost and timeline is always a big decision factor in terms of any direction that you take. But then that always gets traded off with the required functionality. And then there's, we can propose always different hybrid models in terms of how much custom development we do and the trade off with that functionality. 

But otherwise, it comes down to starting with the core requirements of the application and kind of all the standard IoT questions that come with that. What do you want to measure? How frequently? How do you want to communicate that data? In what conditions? What environment are you operating in? Those are the hardware questions, and then the software ones are what functionality do you need, what data do you want to show, who gets to see it and how, are there, is there other logic that you want to build in, such as alert notifications or automated decision making?

And then you take all of that in, and you have a conversation of some things are easy to do custom, some things are easier to do off-the-shelf and then trading off what is the time and cost to implementing each of those. 

- [Ryan] And I imagine that you can approach it where hardware and software can be looked at independently as, even though they are intertwined obviously with the end solution, but if you go custom with software, it doesn't necessarily mean you have to go custom with hardware, correct? Like you could do off-the-shelf, one custom with the other, as long as they're compatible and they have a company that can help integrate them in together.

- [Richard] Yeah, and that's a great thing with kind of where IoT has gone over the past few years is that interoperability is ever increasing. And so most hardware providers provide open access APIs to the data, right? It gets sent to an MQTT broker, you have the data there, and then that can be processed by whatever backend or software that you want, and you have the ability to access that data. So there is certainly a separation between the two. 

That being said, vendors are variable in terms of their documentation and support, right? Both on the software and hardware side. Sometimes it's really easy to get stuff to work. Sometimes it isn't. And having that end-to-end control is really valuable.

In terms of, the best user experiences end up being the Apple model, where you do control everything end-to-end, and you control every part of it, and you can adjust every part of it if you need to. But that usually comes at a cost, right? But yeah, if you, if that's really one of the aims, is that everything works together seamlessly, and you have a great user experience, then usually that's the direction you need to go to. Bringing together different pieces together works, but often there's hiccups here and there that come along with that too. 

- [Ryan] Yeah, absolutely. And have you had companies come to you that have maybe gone down one path, gone the custom route, and it didn't work out, and they came to you and said, hey, can we just figure out that there's an off-the-shelf solution or vice versa? Hey, we went off-the-shelf. We deployed something. It really didn't do what we wanted. We need another level of kind of customization to this. Can we go the other route now? Does that ever happen and if so, what does that look like? 

- [Richard] The latter happens pretty much all the time. The former not so much, right, where because the decision process usually for somebody who wants, looking to deploy an IoT solution is, hey, let's look at what's out there. Let's try it out and see if it meets our needs. And so usually customers that are coming to us, they've been through that and they say, hey look, we've tried what's out there, and we know what's available, and we know that it doesn't meet our needs, and that's why we're looking at a custom solution.

The other way around, I think doesn't happen so much, where you start custom, unless something new comes out, you know that whatever's in the market is not going to meet your needs. 

- [Ryan] When it comes to making the decision on going custom versus off-the-shelf, what oftentimes makes one a better fit than the other? And how do companies go about from the early stages to decide which path is more a fit for them? I know we've talked about a lot of different factors that play into this, but in those discussions you have or, and as people are listening to this that may be in that stage of their decision making process, what advice do you have for them to help evaluate if custom or off-the-shelf is a better fit for what it is that they're looking to try to do? 

- [Richard] Ultimately comes down to the functional requirements of the application, right? And we have, the IoT industry is so broad, right? There's so many different applications out there. And every time, you know, we speak to a customer, it comes down to whatever they want to do is really specific and peculiar in their way, right? Because you think there's so much general functionality in IoT, but then every application is so unique in its own way. And then those kind of general solutions might break down for that unique application. That's usually the story for those things where it's just unique and specific in that way and for that reason, it has requirements that need a custom approach. But usually we scale up with customers as well, right? It's rare that you move to straight to let's build a custom thing straight away. You can often build a proof of concept prototype using development boards, off-the-shelf pieces, right, to really test out your thesis and test out the design and make sure that you're taking the right direction. 

So you're not going into full custom development right away. Often you're leveraging existing tools that are prototyping or fully developed to try and get you maybe 70 or 80 percent of the way and test that out to make sure that you're, the architecture that you're selecting is the right one. 

- [Ryan] Yeah, I think it does a really good job of showing the power of IoT and what it can do for an organization during those pilot stages because as you mentioned earlier with interoperability, it does open the door for a lot of other solutions to be built on top of the initial infrastructure you set up to launch the first solution.

And we talk with companies about that all the time is once you get something in the door, and you start monitoring and tracking or tracking something in your business, and then you start to connect that to other things that you do on how that could be optimized or automated or get better insights into through other IoT solutions that you can bring in and put on top of the existing system you have, which I think is a big testament to where the industry is now.

- [Richard] Yeah, and absolutely, there's so many great tools out there as well to get you started and get you that 70, 80 percent functionality, so that you can try the different hardware, you can visualize the data, and maybe it doesn't give you everything that you need, but it gets you a kind of functional solution that you can try out initially, and then you can really make that decision well. Do I really need the customization that I need, or am I okay using these initial tools and this initial hardware in a first version, build out my application that way, and then really understand what is critical, what is not critical, and then go the custom route if it's needed. 

- [Ryan] Yeah, I feel like a lot of companies who have focused on building low code and no code tools try to advocate for that. Being able to do that themselves, a company can pick it up, try something out, and then if it works out, work with an organization to help scale, help add custom layers onto it and things like that. But we've seen a lot of growth in that area of the space I feel like over the last couple years or so. 

- [Richard] Yeah, and certainly, I think it's kind of part of the sales path for a lot of established companies as well, right? Where you have that initial solution that kind of gives you that core functionality, but ultimately most customers, especially as you scale up in their size, on the way to enterprise, they're going to want some level of customization, API access, or enterprise customers are running their existing ERP systems that they're going to want to plug that data into. And so that inevitably results into some amount of customization, non recurring engineering work that gets done for those solutions. One way or the other. 

- [Ryan] The last thing I wanted to ask you before I let you go is just any other general advice or thoughts you have for our audience who is looking to start their IoT journey, or maybe they've already tried, and they haven't had the greatest success. What's the best way for them to get started on the right path and or right the ship potentially and start to see success with an IoT solution deployment, just for, anything else we maybe did not cover that'd be worth mentioning before we go here. 

- [Richard] Yeah, I think it's taking the right stepped approach, and we touched upon that a little bit. And a lot of that is starting with do a thorough search of the tools that are available out there. Speak to the vendors. And then if something looks, it might not look like a 100 percent fit, but maybe even a 70 or 80 percent fit of what you need, it's worth trying out and doing, buying a dev kit and trying out that system and seeing what kind of data and functionality you get out of it. And slowly ramping up to maybe a pilot with that solution, or at that stage, you might realize that you want to go custom and then it's time to speak to a company like ours. And then do those trade offs of combining existing solutions versus fully custom and what you can do and slowly ramp up that decision tree. 

- [Ryan] And for our audience who might want to reach out after this conversation and listening to this podcast, touch base on anything we talked about here, just follow up in general, what's the best way that they can do that? Where can they learn more about what you have going on? What's the best path there? 

- [Richard] Best way is to come visit us at our website at brashinc.com. That features all the ways to reach us, and you get to view our full portfolio of work that we've done there in past projects.

- [Ryan] Well, Richard, thank you so much for coming on. I think this is a very timely conversation where our kind of organization, as well as many others, really believe we're in a spot where solutions are starting to really lead the way for the IoT industry. And the work you all do is obviously contributing to that.

So thanks for coming on and talking through a question that comes up very often, off-the-shelf versus custom, how to make the decision, pros and cons, long term, short term benefits, all that kind of good stuff. I really appreciate it and excited to get this out to our audience. 

- [Richard] Great. Thanks for having me, Ryan. I thought it was a great chat and yeah, it's great to get the information out there in terms of what's involved for developing custom solutions in IoT.