A major challenge for marketing the internet of things? Explaining why it has value

[protected-iframe id=”28cc3374fd8657e015fdac6b1069b8d4-14960843-25766478″ info=”http://new.livestream.com/accounts/74987/events/2423781/videos/32379838/player?autoPlay=false&height=360&mute=false&width=640″ width=”640″ height=”360″ frameborder=”0″ scrolling=”no”]
Transcription details:
Date:
18-Oct-2013
Input sound file:
10-16 am Session 2_1002.MP3

Transcription results:
Session name: The IOT guru panel
Speakers:
Stacey Higginbotham
Javier Martinez
Mike Kuniavsky
Scott N. Miller
Peter Semmelhack

Announcer 00:00
… Principal Scientist at PARC, a Xerox company; Scott Miller the CEO of Dragon Innovation; and Peter Semmelhack, CEO of Bug Labs. Please welcome The IoT Guru Panel.
Stacey Higginbotham 00:16
Hi. We’re back. All right. This panel is awesome because unlike, maybe, a lot of people out there, everyone of these guys has actually built something, implemented an IoT system, a smarter city system, or writes a lot about it. No, you’ve built stuff, too.
S? 00:35
Yes.
Stacey Higginbotham 00:35
Yes.
S? 00:36
Theoretically or practically.
Stacey Higginbotham 00:38
Hence, the guru. Yes. This is going to be a lot of fun. What I figured I would have them do is each of them is going to introduce themselves and then we are going to go after that. They’re going to tell a story or tell about a time, their lessons learned in implementing or building or doing some kind of system. You guys should totally listen up because I have learned many things and, quite frankly, I’m scared. We’ll get started with Javier. Would you like to introduce yourself?
Javier Martinez 01:06
Yes. Good morning to everyone. My name is Javier Martinez and I am the Chief of Business Development at Libelium.
Stacey Higginbotham 01:14
And Libelium is?
Javier Martinez 01:15
Libelium is a wireless sensor network company, technology company based in Spain.
Mike Kuniavsky 01:22
My name is Mike Kuniavsky. I guess, I’m wearing two hats here. One is that I’m in the innovation services group at PARC where what we do is we essentially reduce the risk of our clients adopting novel technologies in a variety of ways. I’ve also co-founded ThingM which is essentially an Internet of Things start-up and we’ve been making products for six years, something like that.
Scott Miller 01:50
Cool. I’m Scott Miller, CEO and Founder of Dragon Innovation. We help companies go from the prototype to high-volume manufacturing and also have a cool hardware-focused crowd funding site.
Peter Semmelhack 02:05
I’m Peter Semmelhack. I’m the Founder and CEO of a company called Bug Labs. We got started in the open source hardware space and has evolved now into what’s now called the Internet of Things. We have software and hardware that helps people innovate and bring new products to market.
Stacey Higginbotham 02:19
All right. Let’s get started. We’re going to go right down the line. Javier, let’s start with your pitfall, your horror story.
Mike Kuniavsky 02:28
We found out that implementing or deploying an Internet of Things or M2M solution, that one thing is conceiving this solution. One thing is conceiving this business case, presenting it, and actually selling it. A totally different thing is the actual deployment. For example, we deployed a wireless sensor network in a city in Spain for a smart city initiative. We found out that although the technology [called?] proposal was horizontal – it was meant to be given a response to a global initiative for the city – we actually found out, the city itself, the municipality, was not aligned with this horizontal approach. That is, we were implementing four different solutions based on our technology which was involving different departments within the city, within the municipality. What we found out is that there hadn’t been a communication between them. There hadn’t been the right communication between departments. We were finding obstacles when we were installing and deploying the equipment. For example, the mobility department. There was a parking solutions, a smart parking solution, the mobility department had not warned the lighting department that we need to install our equipment on lighting poles. We go this obstacle to, as I said, to install the equipment and to get the solution running.
Stacey Higginbotham 04:14
Make sure that, when you sell something, that everybody is on board and all of the different – especially in smarter cities – everybody knows that you’re coming to install things.
Javier Martinez 04:23
There’s no sense in having the right technology, the right technological approach, a horizontal approach, if there’s no such alignment with the end user.
Stacey Higginbotham 04:33
All right. Mike?
Mike Kuniavsky 04:34
I think I’ve got an example from either one of my hats. At PARC, we develop a lot of really very, very new technologies. The giant challenge is explaining the value of those technologies. It’s one thing to develop a technology. It’s another thing to essentially create a commercial explanation for what the value of that technology is. I’m sure everyone has experienced this. One specific case for us is that we developed this printer electronics technology. We essentially print. We can print displays, sensors, logic, and batteries onto a thin substrate. How do you explain that as a commercial product when it’s just really like very much a very powerful but fundamental technology? What we did there is we essentially said, “We’re going to find a partner that’s interested and has an adjacent business.” Then, we’re going to work with them. That partner is Thin Film which is Norwegian company and they’re developing a whole range of products around these technologies. They’re essentially educating the marketplace on our behalf about what the potentials are.
Stacey Higginbotham 05:48
It’s super cool. It’s like you can print this for, like, cents and they stick them on like serial boxes and then the serial box knows where it is.
Mike Kuniavsky 05:55
Right.
Stacey Higginbotham 05:56
And, you know where it is.
Mike Kuniavsky 05:57
It’s, in many ways, like the next generation of RFID but in an entirely non-silicon way. You’ll avoid the entire silicon process and create something that’s flexible, small, and functional. Very cool.
Stacey Higginbotham 06:12
Your lesson there is find a partner who can help explain.
Mike Kuniavsky 06:16
Yes, exactly.
Stacey Higginbotham 06:17
Like I just did there?
Mike Kuniavsky 06:19
Correct [chuckles]. It’s a similar thing in terms of explaining the value of augmented objects. One of the things in the Internet of Things, you get a lot of it but you get this problem where you have an augmented version of something, a connected version of something versus a non-augmented and connected version of that. The challenge there is just that when consumers look at it, they go, “Oh, why should I pay four or five times as much and, equally as importantly, do all these extra work? I never had an app for my smoke alarm before but, now, I have this other thing that I have to manage, I have to think about. Why should I do that?” The way that Nest is doing it – a number of people we think is doing it – is they’re doing it through design but they’re also doing it through user experience design as well as industrial design. They’re also doing it by really creating incredibly clear use cases for that.
Stacey Higginbotham 07:31
All right. Product design, actual products, getting people to buy them before you can do any of that. Scott, what do you have to think about?
Scott Miller 07:41
All right. In the spirit of Halloween, I’ve got a manufacturing horror story which pretty much ended our stay at iRobot. Before starting Dragon six years ago, I spent a great ten yeas at iRobot. Like a lot of our clients, we were really focused on just getting the first thing built. We had really tight schedules, expectations from the press, and so on, that we just needed to launch the product and really didn’t think too much about quality. We though it was important but we’d figure it out later.
Stacey Higginbotham 08:09
Wait. You should tell people what iRobot– for the few who don’t know.
Scott Miller 08:12
Yes. Sorry. iRobot is a consumer electronic company that builds the Roomba. It’s a really cool vacuuming robot. They’ll clean your floor for you so you don’t have to. It’s a highly-integrated five degree of freedom electromechanical device. For a robot guy, it’s like heaven. It’s just really entirely put together. We’ve been scrambling to get this thing put together and meet the schedules. At that time, I was based over in China and get a call on a Saturday afternoon saying, “Hey, Scott. One percent of the robots are failing after 48 hours.” This is like the worst thing you want to hear. If they all blew up instantly, that’s awesome because you know you’ve got a problem but these intermittent things like, for step one, it’s always discredit the tester. No. That’s a bad idea.

[Laughter]

Don’t do that. I was like, “All right. I’ll take this seriously.” I get on a boat, go into the factory. Three hours, I lay and sit down. Lo and behold! These robots are failing after 48 hours. We have a whole floor of it, about 100 of them running around. At the time, we had built up 50,000 of them for Home Depot which– if you think of 50,000 robots, that would fill more than this room. That’s a lot of robots and you really don’t want to rework 50,000 robots. What we did is built this code called the Blue Light Special where, everytime one of the components would fail, the blue light would go on. Just looking at a sea of 100 robots, all the blue lights were going on. It’s like, “Oh, this is really, really not good.” Because that’s tens of millions of dollars a product. Fortunately for us, after four long days of pretty much no sleep in hot China, we did come up with a software solution to solve the mechanical problem. It was a little switch which was bouncing too much and [?] fix the product and ship it. The lesson we learned there is you got to take quality really, really, really seriously and not deal with it as a by-product but design it in upfront.
S? 10:01
Have you already shipped all of the 50,000 robots?
Scott Miller 10:03
No. Thank God. We haven’t done the final inspection. If there’s one engineer at the factory didn’t raise his hand because it’s a Saturday. He wants to go home. Nobody wants to say this might be a problem. We would have been– the returns would kill you. It costs ten times as much to get a return.
Stacey Higginbotham 10:24
There’s lots of lessons there. There’s the be-nice-to-your-engineering-staff, quality control. Okay. Peter?
Peter Semmelhack 10:32
Yes. This is more of a perception versus reality lesson. It’s for anybody who’s looking to invest in the space media, entrepreneurs. One of the things that I hear fairly often is either this or some variant of this which is, I think, the next Facebook is going to come out of the Internet of Things. That’s an easy thing to get seduced into thinking because we’re going to have 20 trillion devices connected to the internet. That’s a giant number. We can build an application, 99 cents times 20 trillion. Big number. The problem is that– you think about it, though. A lot of people don’t recognize this thing which is the Internet of Things. It’s not the internet. We got a guy like Mark Zuckerberg who could come with this really cool application for social networking and it could go viral very, very quickly because we all had on our desktops, for most part, a Wintel device. An architecture that was made popular over many, many years. An Intel chip, a Windows operating system, standard browsers, standard communication stack over standard network. Everything was standardized. One idea could scale with zero friction. Let’s say, Scott and I came up with this great, new idea for diabetics. It would really help them live a better life. We knew we could sell this to lots of diabetics but the problem is that there’s no they’re there. I can’t use your iPhone or your Android device because there’s no glucometer on that thing. What has to happen in the Internet of Things is that things have to go first. You have to get them into the market and then you can scale. Right? Therein lies both the challenge and the opportunity because you have to manufacture it. You have to get it out. It has to be successful. Then, you scale. A lot of people don’t recognize the thing part of the Internet of Things phrase. It’s the complicated piece. You’ve all heard hardware as hard. That was because it’s true. I feel like you serve yourself well by really digging into what does it actually take to get a thing into the market and then scale it because it’s fairly sophisticated thinking [inaudible].
S? 12:31
Can I jump in?
Stacey Higginbotham 12:32
Go for it.
S? 12:33
I was actually going to redouble that and say that, as a software person going at a hardware – because I was into software for a long time – you have this misperception that once you can compile something and get it to run once, that’s awesome. You’re done. It works. That’s absolutely not the case with hardware where not only do you have a percentage of the atoms don’t work like the rest of the atoms but–
Peter Semmelhack 13:00
[Crosstalk] Physics.
S? 13:00
Right. You also have this problem where, when you make 100 of something versus 100 thousand, you have an entirely different set of processes. You have an entirely different set of thinking that has to happen around that which is incredibly hard.
Stacey Higginbotham 13:17
To take this another step further, if we want to get the things out there and they’re going to be built well, we got to convince people the value and I’m not sure in my home if I can think of this and that’s where I’m going with the horizontal integration. How do you create platforms? Do you create platforms? How do you create this ecosystem where I, as the consumer, can buy a glucose monitor and then, later, go buy a heart monitor and know they’re all going to work together? Because that, to me, is the biggest stumbling block.
Peter Semmelhack 13:53
I guess, that’s a huge problem because in that statement is a business model problem. Everybody wants to own the stack. If you look at health as an example, you got Medtronic. You got Johnson and Johnson. You got all these guys who really are doing a great job and they’re particularly vertical whether it be diabetics, heart condition, or what-have-you. If you want those things to talk to one another, it’s not going to happen because, again, everyone wants to get out of the thing business and they want to get into the services business. They want to give the thing away and charge you for services. I think there’s this inherent contradiction between what the internet did because no one owns the internet and what’s going to happen in the IoT space because everyone wants to own the IoT platform. One of the things I’d like to say is before we had the internet, we had the information super highway. We all know how that worked out. How a platform becomes a market where everybody benefits is a real interesting question there.
Stacey Higginbotham 14:49
What do you guys–
Javier Martinez 14:49
We found from our experiences a couple of ideas. First of all is that end user, whatever the end user is – an individual, a [inaudible] in business – wants zero complexity. The Internet of Things and the value chain within it is a large thing. It’s very comprehensive. It can be a complex thing but the end user is not looking for that. It’s looking for zero complexity, “I want a solution and I want this solution implemented and satisfy my need.” There is a need for all the agents involved, all the players involved, to find their space, to find their specific role. In our case, we’re not aiming to champion this phenomenal platform with another one. We are focused and we’re specialized on that small part of the value chain. Another idea is that we cannot trim a value chain. A value chain has to have all the parts. If we all want to be focused on one part that is the service part that is creating the largest revenues and nobody is taking care of the parts before, nothing is going to be done. No solution can be created.
Stacey Higginbotham 15:59
Who will have the power in that situation to set pricing? If you look at the app store model, Apple knew it wanted to get developers involved. It gave them a 30% cut. At the same time, it was dictating pricing to carriers and that sort of thing. There is a very real– I guess, it’s an open blank slate. Who will have the power in this kind of relationships? Does it come from the hardware guys? Does it come from your brand name? I don’t know. Design?
Peter Semmelhack 16:32
I think, if you look at Apple, Apple did something interesting. When the iPhone first came out, if you remember, there was no developer platform at all. They did this great job. As Apple has now proven, they can do it multiple times which is they sold a beautifully designed piece of hardware. It wasn’t even that great of a phone but it had this vision around it. It’s this awesome new thing. They sold millions from the first year and they turned around and said, “By the way, developers of the world, do you want to sell apps to these two million users? By the way, your $0.99 is what we suggest. We’re going to do the math. Two million, $0.99. That’s real money.” The developers came in. Right? They were able to do. They were able to start the fire. They sold the hardware. It was an existing device. People understood the value of its phone and then they created its ecosystem. They launched into this situation where they did have a lot of power because they said they are the users. That goes to my problem I mentioned earlier which is you have to get the hardware in place. You have to get something in place to keep developers interested. That’s such a challenge.
S? 17:35
I think that Apple is more the exception than the rule. Everybody tries to make a vertically integrated ecosystem that they own and they create this environment. Really, very few people have any success at it because, essentially, every consumer electronics company has their own version of this but, really, Apple is the only one that’s really genuinely successful. To me, the thing that is missing, as I said, it’s not about being to google something. It’s about creating the HTTP of something. That’s the layer at which that’s, if you’re talking about power, the real power is in the IETF. That’s who really– not here but in terms of who really enabled the web and that’s the layer here. It’s who creates the service discovery layer and makes it so open that there’s the incent of our people to create business models inside it.
Stacey Higginbotham 18:37
Okay. We’re probably not going to solve this today.
S? 18:39
[Laughs]
Stacey Higginbotham 18:39
Let’s move to this next issue which is big one and I don’t think we really talked about it yet. It’s about privacy and security when we start connecting things. In a city situation, I’m sure there are all kinds of issues ranging from surveillance to even someone hacking your parking systems. Most benignly, you would ever pay the park again or scheming credit cards. I don’t know what you could do.
Javier Martinez 19:07
I think the issue within a city is a city is a generator of huge data. Each day, cities generate huge data and the issue is how do we approach this data. Do we have all the data open? The data is from multiple nature, multiple traces, and involves different types of individuals and organizations. We’re for openness; but, where is that frontier? Where should openness finish and what kind of data should be open or not?
S? 19:46
We always go back. I love the saying from Paul Graham, “Build something that people want”. In press, we’re always balancing the value you get from exposing your data to the cost of– or the security of giving that up. If it does something useful then folks are maybe more likely to say, “Sure. I’ll share that.” It can easily go the other way if that trust is violated.
Stacey Higginbotham 20:08
Yes. I think, actually, we should probably make this distinction. Shame on me for not making it. Privacy versus security because they are two different things. If I’m looking at open data, I do not want the red light tends to– I don’t want those pictures taken of me. Not that I’m running red lights but I wouldn’t want those available. It would be like the mug shots online. Right? It’s just not something you want. That’s a privacy issue versus security which would be like hacking into the red light cameras to really ruin my day and give me a ticket because I have alienated someone somewhere who’s very smart. There’s probably different considerations here. Scott, for example, how do you build hardware that takes that into consideration? And, I’ll ask you how to build services that take those things into consideration.
Scott Miller 20:58
It’s an incredibly hard thing. I don’t think we’ve found the right– I think it’s very much in plan, an open question.
Stacey Higginbotham 21:08
I don’t know. [Laughter]
S? 21:10
I could tell you. One of our customers is Ford. You’ve seen a lot of interests now in telematics and getting cars online. There’s a lot of interest in things you can do. I do agree. There’s lots of potential things from safety to efficiency, to entertainment, and so on. The problem that the car industry has is this idea of driver distraction. Right? The big problem that Ford and every OEM in this country has is that, God forbid, they come out with some really cool new open platform. Somebody writes an application, someone dies like they [would just keep them up at night?]. How do we build a system that doesn’t open that? It’s a cost benefit analysis. There’s huge amounts of benefit but one thing will ruin it for everybody and, in many cases, they are not willing to take that step like OBD2 which is the interface. It’s totally locked down and they’re opening it in tiny bits here and there but it will be very interesting to see how they– because they, in many ways, are under a lot of pressure to open it. Like, Tesla, is doing some interesting things.
Stacey Higginbotham 22:09
That’s the tension between these open versus the siloed systems. It’s not just a business model thing. It’s, kind of, an insurance and [security?]. Okay. Let’s go forward to thinking about for businesses because we focus a lot on consumers and in cool things you can do with this. Businesses having access to this data, how do you think they will buy into the Internet of Things. Is it something like GE’s platform where they’re like, “I have all these data. I’ll give this to you”? Do you think it’s a little more adhoc? Who’s going to win there.
Peter Semmelhack 22:48
I have a big opinion about this. I personally think that the Internet of Things is going to take off in the enterprise way before it takes off in the consumer space.
Stacey Higginbotham 22:56
Okay.
Peter Semmelhack 22:57
No offense to anybody doing smart home of anything like that but–
Stacey Higginbotham 23:01
I still love you.
Mike Kuniavsky 23:01
Looking from my entire career, home automation is always about to be huge. It’s like this is the year, home automation. It’s like artificial intelligence. It’s not to say a lot of good things are happening. It’s just, for me, I think the problem with it is ROI. Meaning, in business, you can say I’m going to apply this real cool technology. I’m going to save you 20% a year and, therefore, you can going to have to pay me this or I’m going to make you 50% more revenues and you’re going to pay me that. You can literally do that now much easier with this type of technology. In the consumers space, it’s a little harder. What is the value of turning the lights on and off remotely? Even things like safety, they’re good. What’s the real ROI? How much is that worth? Is it worth $300 or is it worth $99? I don’t know. I do think there’s huge amounts of opportunity. I think GE has got it right in the sense that the industrial internet– this is my opinion. I feel like a lot of innovation is going on in that space. I think we’ll see a lot of interesting things happening there.
Peter Semmelhack 23:58
My perception is that it’s going to happen in two ways. What the industrial internet is, is essentially people taking existing networks of devices they already have like trucks or turbines or meters and turning that existing network that’s currently physically-managed into a digitally-managed and digitally understood network. It’s essentially merging the physical network with the digital network. I think that’s much easier to roll out on a business scale when you have a physical network of things whether it’s delivery trucks or something like that, or washing machines or whatever. The other way that I think that it’s happening is that businesses are starting to see that there is competitive value in creating a specific hardware device that generates a range of data that allows them to compete in terms of the service that they have against other services. Hardware, even though it’s really hard, is actually comparatively easy relative to all the business problems. I think making a small sensor-based battery-powered thing that connects to the net is actually somewhat easier than it is to create a successful business and a service. I think services are going to be discovering but that’s a competitive advantage and they’re going to be creating hardware to get people to log into their services.
Stacey Higginbotham 25:23
Okay.
Javier Martinez 25:24
I agree with Peter that the Internet of Things is going to take off before on a business level more than an individual’s level because it’s easier to, as you said, to build a business case and to present a potential return on investment. In my opinion, eventually, the individuals market will tell us like killer applications that can be developed. We see that, for example, for businesses, we have not found – although our technology can be applied to multiple applications – we have not found a killer app, yet. Still, we see it easier to present, to build this business cases. I think, eventually, will be the same at an individual’s level.
Stacey Higginbotham 26:08
Are you saying that enterprise, king of, Internet of Things deployments are one-off kind of consulting things versus consumer can be scalable? Am I just–
Javier Martinez 26:18
I think you’ll find a scale in both cases. It’s just a matter of the potential end users saying, “Okay. This is going to be relevant for me. This is going to add value to my life.” For example, I need a health platform. A remote health monitoring platform could be an example of that.
Stacey Higginbotham 26:38
Right. Before that could have any value, it’s got to have all these other things in it. We have to actually all buy them so they could start building things for them that makes sense for the developer. Gosh! We may never get there. How long until we actually have really viable– just a one-word answer. How long do you think until we have a viable Internet of Things platform that we could point to and say, “Oh, that’s it right there”?
Peter Semmelhack 26:58
Ten years.
Stacey Higginbotham 26:59
Ten years.
Scott Miller 27:00
I’ll go for five.
Mike Kuniavsky 27:01
Yes. Three to five.
Stacey Higginbotham 27:02
Three to five?
Javier Martinez 27:03
I would agree on five.
Stacey Higginbotham 27:03
All right. We’ll see if Peter The Pessimist is correct [laughter] or we’ll all be excited in five more years. Thank you, guys, very much. I appreciate it. Give a round of applause for our guys.