Brought to you by VANTIQ
Episode 15
Implementing IoT In the Real World
IoT strategy is one thing, but actually implementing projects involving software and hardware is definitely something else. Join real-life IoT practitioner Cliff Robbins from GearForge in this detailed discussion of how to bring IoT projects to life.
Blaine
Mathieu
Chief Marketing +

Product Officer, VANTIQ
Cliff
Robbins
V.P. of technology and Resident Data Scientist at GearForge Software

Blaine: Joining me today is Cliff Robbins, V.P. of technology and Resident Data Scientist at GearForge Software. Cliff is a longtime programmer, engineer, and software architect who learned his leadership skills through eight years in the U.S. Marine Corps. Thanks for the time, Cliff. We’re going to have some fun today I think.

Cliff: I can’t wait. Thank you so much for having me here today, Blaine.

Blaine: Great! Well, I know this is going to be a little bit different than many of our other interviews. They have been with consultants and folks talking at a very high level. As we can tell from the array of IoT devices and sensors and the oscilloscope behind you, you’re an actual, hands-on practitioner in building IoT systems.

Cliff: You’re spot on right there, Blaine. I love getting my fingers dirty. I can talk high level just like many other folks, but I like getting my hands on. You can’t see over here, an entire solder station to my right. And this is just my tinkering space. I’ve got a whole workshop in the other office section back there.

Blaine: All right. Well, maybe, at the end of the interview, you can give me a soldering lesson because I’m horrible at soldering. I make a mess of anything I try to solder.

Cliff: You’re better off just letting someone else mess with it. [Laughter]

Blaine: I think you’re probably right!

Well to start, Cliff, tell us a little more about your GearForge, GearForge Software, what are they all about?

Cliff: Sure. So GearForge Software is really built around the concept of how do we take and not add just the high level from the IoT standpoint, but how actually then help companies implement IoT. And it’s beyond just IoT because we also encompass things. There’s typically going to be some custom software that needs to be brought into it, consulting (specifically around technical consulting), and then also, once you have it connected, what do you do after that?

There’s a lot of other things too. Like the other consultants we talked about: you got your legal, your pricing, the whole big ball of stuff that’s all encompassing. That’s really what GearForge’s focus is on: really coming in, assisting firms with trying to understand that and then being able to actually have a practical implementation. So it’s not just fluff. It’s like, we’ve got this, we’ve defined it, now let’s actually build something.

Blaine: Interesting. How did you get to where you are today? How did you get to be the guy who knows how to do this thing you’re doing?

Cliff: Oh sure. Probably a very diverse background. You mentioned earlier the Marine Corps. In the corps. I actually worked on an F-18 so I did avionics, electronics. The core taught me a ton of electronics. And then, ended up going into computer science after that.

I’ve always had a heart for electronics. IoT, of course, is the perfect blend of electronics with computer science couple that with data science. I love data and understanding the meaning and the story behind the data and how you can actually drive transformation.

So, you combine all of that, it’s the perfect storm. It’s just a love all of these components. It fit really well to be able to bring them all together.

Blaine: Interesting. Where would you say the market is in terms of its maturity for these solutions?

I’ve been going to IoT conferences fairly aggressively for a couple of years now. And even a couple of years ago, it seemed to me it was always the same examples, the same couple of use cases, even the same companies were doing some leading-edge stuff. But mostly, it was a science experiment or a POC.

Now, I see a lot more companies doing real things, obviously with the help of you and GearForge. Where do you think this overall market is in its maturity cycle. Are we still early days or is it in the mid-market? What’s your take?

Cliff: Yeah so it reminds me of back in 2007/2008, mobile space. Back then, it was very early on. Most companies were trying to figure out is this a space that we need to be getting into. That was really the question. That’s really early market. So, you’re early adopters.

Personally, I think from an IoT standpoint and just the overall connectivity and technology standpoint, I think we’ve probably moved past a little bit of that early stage. We’re not mid. I think we’re right at that interim space right about that is where I personally think we are. Now, we’re seeing there are some B2C but we are seeing B2B has a lot more opportunity compared mobile, in this case, for some revenue generation. So, I think there’s a lot of opportunity and value creation from a B2B standpoint.

Blaine: Hmm interesting. I think a lot of the early buzz around IoT was the B2C devices and the Fitbits and all that kind of stuff, but I also agree with you. It seems like a lot of the money and the traction has shifted more to the B2B side and how you can improve business operations, processes with the use of IoT devices. Sounds like you’re seeing that too?

Cliff: Yeah. I mean, it’s interesting because you’re seeing companies that are just hardware companies like manufacturing firms. Now, all of a sudden, they’re technology companies and it’s something they’ve never had to before. It comes down to even invoicing. .

Now , they’re a platform as a service or a software as a service company. They’ve got to change their mindset and hire new people to actually morph into that realm. It’s a big change and it’s a real challenge for those leadership teams.

Blaine: That’s a good point. Let’s circle back to the change management side and that’s a whole other big topic, right?

Not to get too strategy geeky, but as most folks know in technology circles, when you’re trying to cross out of that early adopter phase into the mainstream market, you hit something that Jeffrey Moore called the chasm; this part of the market where the early adopters aren’t good reference customers for the mainstream market. The mainstream market needs to see other mainstream customers using this technology. But there aren’t any. It’s all these crazy early adopters.

How are you feeling that at GearForge?Are you still being approached primarily by early adopter organizations that want to be on the leading edge of IoT? How are you finding your clients? Do you have to evangelize and teach about the possibilities of IoT or are they finally starting to come to you saying, “I know what I want to do. I just can’t do it. I need the help of doing it.”?

Cliff: That’s a great question. It’s a mixed bag. And from that standpoint, Blaine, it really is a mixture. You’ve got some that are three years into a project. They are knee deep and they don’t even realize it’s IoT. They just know it’s a smart something or other. They’re hitting a wall and need a technical person to help guide them through the process.

We have other folks where we’re coming in and we’re pitching the idea. They’ve been to conferences, things like that. They know it’s possible. Now, [they want to know] how to capture, how to turn this into a revenue stream, into a real business. But, like you said, they hit here and are like, “We have no examples to go off of.”.

On a side note, we do get calls from folks that say, “Hey! We’d like to do blah blah blah. Think you guys can have it done by next month?” And we’re like, [laughter] “No. You’re dealing with hardware/software. Let’s be realistic from a time standpoint. Let’s be realistic from a time standpoint.

But, I think it’s interesting now because it’s shifted from those very earlies and now you’ve got folks that are, just like you said, know there’s value, they can make a business case for this. They need help, though, of getting that. In many cases, most of these firms don’t have a CTO.

Typically, they’re outsourcing most of their server work, everything else because they just have laptops, things like that. That’s just a service they’re outsourcing. They don’t have someone that’s a technologist to be able to walk them through on actually how to do it.

Blaine: Right. And I imagine, even larger companies, large enterprises that have significant IT departments, still don’t have experts in IoT devices and building IoT systems.

Cliff: You’re entirely right. We’re talking to 450 million dollars manufacturing firm. So, it’s not a small company by any means. Basically, their internal staff, as far as their technology firms, they know all about process control, anything else. But that was just factory. They didn’t know anything else about actually combining their data sources and be able to start bringing things together to figure out ways.

Even thinking beyond that of how do you turn this into a revenue stream wasn’t brought up, but the interesting part is I think they have about 30 use cases. They just don’t know how to actually apply those internally because they’re focused so much on their technology stack. It’s the forest in the tree scenario. They can see the tree in front of them, but they can’t actually see the entire forest.

Blaine: Well thanks for bringing up use cases because that was where I was going to go next. Are you seeing any commonality in use cases across your clients or the companies you’re working with? And If you have any examples, that would be great.

Cliff: Realistically, some of the main one’s we’re seeing it’s not so much cutting the cost per se but a lot more of how I create new revenue streams. I would agree top line revenue. Not so much worried about cutting down on expenses. So, we’re seeing more so on top line revenue.

A good example is a manufacturing company. They want to create an entire line of monitoring for a particular type of product that they produce. But, they don’t know how to get there. Very sharp, they’ve got engineers and everything. Very intelligent individuals, but they don’t know how to do this other piece. So, one of the big things we are telling them, when it comes to hardware: prototyping, sourcing, getting demos, and everything else is going to take time.

This is not something that you’re going to have done next week. We’re estimating at least a year to two years of just getting your hardware prototype nailed down.

Blaine: And what is the use case in this is? Is it like a manufacturing line or what is it?

Cliff: This is actually a monitoring system for a particular line of product that they produce. What they are going to do is they are going to monitor this across the board (for others even) and be able to predict failures based off of what they’re seeing. Essentially, being able to use some machine learning, things like that, upstream, to be able to do that.

Blaine: Perhaps to enable a real-time field service use case?

Cliff: Well, it’s across the board. They can say, “Yep. Now take that item out of service.” Either we can come in or your own technicians can come in and fix this. They’re going to provide the monitoring of that. You can’t scale that from a human standpoint, so it’s got to be technological.

Blaine: It’s interesting you say that creating new revenue streams is the value that you’re seeing for many of these scenarios with your clients because I was at an IDC conference 3 or 4 months ago where they were talking about their research data on IoT projects. They actually found that most of the early projects were more about cost savings.

I wonder if that’s because if you’re doing an IoT project with your internal IT group, the internal groups are very focused on optimization and IT is always focused on cost management and trying to squeeze more out of the stone. Whereas, if they’re bringing in an outside expert like you, it’s for the reason that they want to do something new, they want to do something transformative, generate new revenue. I’m just theorizing here.

Cliff: Yeah! I think you could be onto something that makes total sense to me because, internally, they think they can figure this out, but you get to a certain point and a lot of them are like, “Nope. We need to bring someone else in. We just don’t have the expertise.”

Blaine: Mm hmm interesting. Any other common use cases you’re seeing? We’ve got the machine monitoring, enabling real-time some kind of field service application, predictive maintenance. What else? Anything else?

Cliff: We had a fun product we’re doing right now and it’s a vending machine. It’s not a beer vending machine or anything like that, trust me. But, it’s totally different. Again, a lot of these, I can’t – I’ve got NDAs and they have patents and everything else that they are working on. But, it’s a pretty wicked vending machine that they’re working on for a very large franchise in the United States here.

So, it’s been a really fun project. They tried to go as far as they could with a model and so they said, “Here’s this. Can you make this work?” And so I started work on it with a machinist, more of an industrial design individual and I didn’t know about the design that he had. We ended up having to tweak quite a bit of basically everything they brought us. And had to redo a lot of, almost all of it actually, just because it just didn’t fit from a prototype standpoint once you actually built the actual vending machine, it just didn’t work.

Blaine: Hearing you tell that story reminds me of just a couple episodes ago we interviewed the CEO of Hoplite Power who was actually the first VANTIQ client.

They have a series of vending machines across New York State and soon expanding around the U.S. for actually vending battery packs for cell phones.

Cliff: Sure. That’s perfect.

Blaine: You put in your credit card, dial in on the kiosk. You get a little battery pack that you can take with you. You don’t have to leave your phone plugged in. You can take it with you. So, managing the real-time state of these machines: the batteries as they’re being charged or taken off and returned, maybe even somewhere else – all that done in real time.

He was describing to me the process they went through to get from their prototype. They’re about to launch v2, but it sounds very similar to what you’re talking about here.

Cliff: There’s a lot involved. And it’s sad. We’re like, “You know consumers are probably going to try to wreck this thing on top of that.” So how do you balance that use too?

Blaine: Yeah. Really, really cool. What have you seen in the real world as some of the main barriers to success for IoT projects? What are tripping companies up? What are slowing them down?

Cliff: Some of the biggest problems that we see is [companies] trying to get too grandiose at times. It’s like, “We’re going to do all this!” Well, my personal opinion is you need to narrow it back down and try to improve the concept. It’s no different than if you’re bringing a product to market, stuff like that. You need to validate your audience. Model the business factors and marketing factors first before you get there.

And in many cases, you can almost test some of this stuff prior to doing. It’s part of that lean-type model that you might go after. I really would focus on trying to do more of a prototype or smaller scale just to get it out there. Because otherwise, you’re going to dump a ton of cash.

This is no different than bringing out any product line. Bring it back down. Prototype it. Leverage those lean mechanics. That’s part of the biggest thing. It goes back to chasm that you’re talking about. A lot of them think they’ve got to have all this. And really you need this much to validate that. So, let’s start with that first.

Blaine: It’s back to your vending machine example we’re both talking about. Guarantee you’re going to do it wrong the first time.

Cliff: Hundred percent.

Blaine: Or not optimally, in any case. So, start small, do it fast, learn a bunch, and then go bigger.

Cliff: Exactly. You know there’s going to be v2 and 3. You’re not trying to build the Cadillac here. Let’s get so we can get a functioning system. Make it so it looks good. Packaging still is really important. But under the covers, maybe you have to have a service guy come out there a little bit more often to make sure things are running correctly, get some metrics. But, you’re going to be building additional things.

Blaine: Yeah. All right, so starting too big of a vision is barrier number one. What else are you seeing as barriers to success in IoT projects?

Cliff: I think the other part is really just finding individuals that can actually do everything that needs to be done. So, talent in this marketplace is probably one of the second biggest that we are seeing right now. Employment in general, trying to find an engineer just right now is such a tough marketplace. So, that’s why we’re actually hearing is the second biggest barrier is really just talent.

Blaine: And thus, you started GearForge to help address that issue! [laughter].

Cliff: Yes, that’s helpful. But that’s a really big problem a lot of folks are facing is they want to do something, but they can’t find people.

Blaine: Well, I know what you’re talking about is not just self-promotional because prior to this interview in my notes what I wrote is my thought starters for the main barriers were complexity which relates to your starting point and then knowledge which relates to your comment about talent. So, literally my two thoughts and your two were identical on this.

Cliff: I think that’s what you see. Like yourself, with VANTIQ, you’re seeing a platform that’s geared to help make it easy. You dive deep into the covers and get into the guts of it. But really, even if you look at Amazon or Azure, both of those are really trying to make systems there are a bit more simplistic. I mean, those are still maturing at this point and there’s still a lot. You could dive deep into the tech side issues with that, but there’s still a lot of those pieces to get those sorted out.

Blaine: Yeah. Well, and then you add the complexity that this is a mix of hardware and software. So, most of the discussions we’ve had on VANTIQ TV have been shifted more toward the software side of technology, talking about AI, machine learning, and real-time applications and those kind of things. But, IoT clearly brings hardware into the mix. Again, judging by the shelves sitting behind you.

Cliff: We’ve got a client in one industry. Because the example is it’s pretty scary. We got brought in as a technical consultant because they were really struggling on bringing things together. We came in like, “Oh my gosh. There’s a lot here.” They had overpromised, as far as delivery is concerned.

We said, “Do not deliver. You should not be delivering this product to your client right now. Based on where you are at, you’ve done zero testing as far as I’m concerned.” Anything else they went into the stars. We actually explicitly said we’re just pulling this out. If you deliver, we are really concerned about overall safety of the device. It ended up catching on fire at the client’s location.

Blaine: What was the client name? [Laughter] Don’t worry about it.

Cliff: So they pulled it back and were re-going through all of this, and still now we’re getting better. But, I’d say we’re probably moving out of alpha into beta where we’re having problems with I2C buses were they’re getting crashes and exceptions that causes the entire thing to just stop working; snares where they didn’t anticipate that they needed more memory in their actual chips, so now they’ve had to upgrade their chipset. And, because of how they had structured the board, they no longer have I2C. I have to use serial communication.

So, really some struggles around all this because they didn’t actually try any type of testing of the hardware, and the person doing the hardware didn’t even have one of the products. They were shipping it back to the factory.

Blaine: Well we have just managed to scare everybody off doing an IoT project and we might as well end the podcast now. [Laughter} It’s just, when it comes to hardware, you ship it out. It’s not like software where I can put it to cloud – a bug – fix it- I’m done. With hardware, you ship that, it’s in audience’s hands now. It’s much more difficult.

Are we getting to a point where standardization is starting to kick in on the hardware side?I’m talking messaging protocols like MQTT and other standards or is it still all over the place; every manufacturer’s got their own way of getting data out of the device or out of the sensor and it’s a free for all? Where are we on that spectrum?

Cliff: I think we’re still early stages. We see standards starting to rise up, but it still is very early. Even beyond that even beyond that, just trying to discern which protocol you should use for transmitting information. Our cellular carriers are still trying to determine which technology they plan on rolling out between Sprint, AT&T, and T-Mobile.

The larger players are still sorting through that. I think I’d like to see the process control industry come into this a little bit more. They’ve already got some really strong IDE standards that I think they can apply. I think they’ll eventually get there. I think they’re more concerned about market share pulling away from PLCs and things like that.

Blaine: Interesting. And that’s actually probably a good segue into your thoughts on 5G and MEC (multi access edge computing). What’s your general thought on the impact that will have on the whole IoT ecosystem and how soon that might have an impact or maybe it already has?

Cliff: I’m really, really excited about the edge computing side. I think we’ve got to get there. As you can see, I’ve got a lot of devices back here. Azure just told me they’ve got one of their edge devices, actually. So, I think it gets delivered this week. Even Microsoft is working on the edge type stuff.

You’ve got to have the edge computing because you know fundamentally that sending everything back without some sort of cancellations or some sort of machine learning at the device or the actual component, you have to have that. It’s one of those where we’re not quite there yet. They’re realizing they need it because you will lose connectivity from that standpoint.

As far as the 5G, I think that’s going to be great. It’s really more of the lower-level bandwidth that I’m really interested in. The problem with a lot of those is the pricing model. It’s still really freaking costly – ten bucks a month. Most clients now are saying, “Alright I’m going to get a $10 sim card for a month.” That’s great. Now what happens when you’ve got a thousand of these devices rolled out? You’re still going to pay $10 a month? That seems excessive to me.

Blaine: Exactly exactly.

Well, let’s shift to something that almost always makes its way into IoT-related videos and that’s augmented reality.

There’s inevitably a video of a technician wearing a headset and then the devices are spewing off data. Everybody has got their video. I’ve seen these so many times. Are you involved in any or even potentially touching any projects that are there yet or where do you see that fitting in?

Cliff: I’ll give you my two cents on what I think about augmented reality. And then, I’ll touch on that.

I think there’s two great applications for augmented reality. Not that I personally use either of these, but games and porn. Those are two spots for augmented reality right now, your best use cases with regards to what you’re describing.

See the hard hat back there? That goes with the safety glasses and the boots. I can’t imagine actually a technician going into a mining facility with one of these augmented reality headsets. I don’t know if they actually went into the field and actually tested these things out. If you actually would go into a mining facility, that’s not going to work. I don’t understand how you’re going to make that function in a real site.

I struggle with real life, as far as that is concerned. Sure, I can see how they walk into a copier and then you try to sort through that, but I am willing to bet the technician that has been on the job for 15 years can probably diagnose it a lot faster.

Blaine: Yeah well that’s right. I think, in a mine, you gave the worst-case scenario use case.

Cliff: But it’s a real scenario!

Blaine: It Is.

I was actually at the Field Service USA event a couple of weeks ago and there was a lot of augmented reality vendors in there selling mostly hardware solutions to connect up to software and IoT devices. In the discussion I had with them, a lot of it actually was about the culture. It’s not about the technology not working. I put on some of these hardhats with integrated eye pieces and yeah it works! It actually works.

But, convincing a 55 year old service tech who’s been doing it one way for the last 30 years to then do it the new way, that’s the hard part, actually. It’s a culture change.

Cliff: Yeah you toss that hard hat and it will get broken and day one, “Oh! Looks like the equipment was screwy!” [Laughter] I’ve just been to the real factory floor many-a-times, and I struggle. You’re probably right: it’s more so culture. I would entirely agree. A lot of them are okay with the technology, it’s just the culture of it. I entirely agree with that.

Blaine: Yeah it is really interesting. At that same conference, I was talking to a senior guy at one of the largest aircraft engine manufacturers in the world. You could guess what company that probably was. We always hear about this notion of air engines as a service and all the IoT data that the engines are throwing off.

He basically said, “Yeah. Those engines are throwing off a lot of data. But are we actually doing anything with it? Not really yet. We don’t have a bunch of technicians on the ground waiting with their mobile devices, sensing what the engine is doing as it’s landing ready to tweak something.” They’re getting old fashioned service reports and pulling data out of a database just like they’ve been doing for the last 20 years.

We definitely have a ways to go in realizing the potential of these things.

Cliff: I entirely agree.

Blaine: It was actually a little bit disappointing. We were around the table sharing our stories. After the guy from this two-letter, well-known company told this, we all sort of went, “Oh”. We wanted you to live up to some of these fancy videos you’re putting out. But, the reality isn’t quite there yet. At the same time, on a more hopeful note, they know they have to get there. A lot of it is about solving the complexity challenge. It’s not easy to do this stuff. You need experts and you need tools and systems with hardware and software that can do it more quickly and easily than today.

Cliff: Exactly. And the hard part is, just like you were mentioning, it’s got to be part of that role, part of that job. It shouldn’t be a burden. Otherwise, you’re going to have such a barrier to entry.

Blaine: All right, well let’s wrap it up. I always ask our guests if there’s one element of conventional wisdom that they like to call bullshit on; where they think most of the market is going West and you think it’s actually East. So, what’s an area of conventional wisdom you think you are a contrarian?

Cliff: Well, we already talked about augmented reality and where that’s going. So, I’ll take another one.

Almost every time I hear IoT, it’s IoT and blockchain. I so totally believe that blockchain is way overrated at this point. I personally feel like blockchain has some very specific value propositions, but tying it back to IoT, every single time seems ridiculous to me. I don’t think they always go in hand-in-hand. I know of a couple of different companies that are trying to make this work. I think that’s really interesting, what they are trying to do. I think we’ve gotten a bit overblown when it comes to blockchain at this point.

Blaine: The funny thing about you saying blockchain is probably a third of our guests say blockchain. [Laughter} The other two thirds, a wide range of topics, but a third say blockchain. That tells you something about the state of blockchain in tech markets today.

Cliff: I’ve looked at the tech. Yes, there are some use cases and things like that, but saying that it goes that far across the entire span is ridiculous to me. There are other opportunities, other ways to do it without as much overhead.

What are some key takeaways or tips you would give to a business leader that’s trying to drive this IoT-based, real time-transformation of her business?

Cliff: I think that’s a great question.

Number 1, look internally for a project. Find someone who can internally do that totally and then get your executive or whoever is passionate, no different than any other project, to champion that. You’re really going to focus on a smaller-scale prototype project. Validate that, get the individuals behind it, and then you can start going from there. That would be my biggest takeaway when it comes to this.

Blaine: Great advice.

Well, Cliff, that wraps it. Thanks so much for chatting with me today. This has been a great conversation. It is truly fantastic to speak to a real practitioner who’s in the field making this future happen. So, thank you so much.

Cliff: Thank you so much. I’m really grateful for being here.

Blaine: You’re welcome!

Those Interested in hearing more about Cliff’s thoughts can check out his IoT weekly podcast on gearforgesoftware.com and you can reach out to me anytime at realtime@vantiq.com.