Hi, this is Wayne again with a topic “Things and the Internet: Rachel Kalmar, Misfit Wearables”.
Talk about toward an internet of things or toward we’re all we’re already heading towards internet of things. I’M going to talk about toward an open data ecosystem and toward because we don’t yet have an open data ecosystem. We don’t have it because there are a lot of really hard and complicated issues, and the reason i wanted to talk about it today is because there are many issues that our communities need to be discussing, and so i wanted to highlight some of those and talk To all of you guys as well that way, so my name is rachel. I’M a data scientist at a company called misfit wearables. We make elegant wearable technology, i’m wearing one of our products, it’s the shine tracks activity and i was really excited to join misfit because, okay, because i really like data and i like sensors – and i wanted to try out all of these different devices – and i wanted An excuse to do that not only to get data from them, but to figure out what activities are, which one is good for, and what can we do with that? What are the the user engagement strategies that these companies are using, and so i’ve been wearing about 10 devices for about six months now, and what have i learned? Well, i’d really love to tell you guys a story about data driven behavior change about how, by tracking my data i was able to be more aware of my behaviors, and i could change could make subtle changes in my activities.
That would, let me be more active. That’S not what i learned. What i learned is that i can’t get my data. I have 10 of these devices. I have 10 logins to different websites and apps and at the end of the day i can’t get time resolved data from any of them. Some of them, let you download your daily totals some are for free. Some, you have to pay a subscription fee. Other ones have apis that give you a snapshot of your data in time, but you can’t get time resolved data from any of these, so you might be thinking okay, big deal, how many people really want a csv file of their step counts? I’D have to say you’re right, not that many people do, but what we do want is real time interaction and we can’t have that unless we have access to time resolved data. What i want is for my internet to get glacially slow. If i haven’t walked enough. So i have to go outside or if i’m having a cardiac issue, i want my doctor to know about it in real time and not later so you guys all know about moore’s law and that connected sensor devices are getting smaller and faster and cheaper, and we’re Able to do things that we weren’t able to do before our world is becoming better connected.
Our devices can talk to us. We can talk to our devices, they can talk to each other. It allows us to have things like wi-fi, toasters and automated traffic flow. So we don’t have to pay attention when we’re driving and at some point in the future, smart medicines and drugs that can figure out what is going on in our bodies before we even have any symptoms. So we have this exponential explosion of data from the very personal level to the global level, and it’s allowing us to do things that we’ve not been able to do before.
We have this grand vision, the internet of things and just as mobile computing and smartphones ushered in a world of possibilities that none of us could have imagined. The internet of things will do the same thing, but we’re not going to have that. We can’t have an internet of things unless we make some changes in the way that we deal with data instead, we’re going to end up with the apple internet of things and the google internet of things and the philips internet of things and the fitbit and the. Why things internet of things? But that’s that’s not what we want. We want one internet of things. We want our devices to talk to all our other devices, not just the other devices within our silos. So why do we have this? Well, we have a new economy. The internet of things in the internet of things, the economy is one of data. Data is the currency that our devices use, but we’re still pioneers in this age, and we don’t yet have all of our infrastructure set up. So if our devices are using different currencies, that makes commerce really hard so like with money with data, if we don’t have the right infrastructure, our currency is useless.
So in this context the conversation shouldn’t necessarily be about ownership, but of control. We don’t we’re not talking about physical things like records and like books, we’re more in an age where we have digital media like mp3s and other things that we can download and data is more like an mp3 file than it is like a record. And so what does it mean to own data? Whoever owns data is the person that has it, and it also doesn’t change in value. Multiple people can have access to the same data and it’s not based the value, isn’t based on scarcity, in the same way that it is in with money.
And so if the conversation is about control, then what does it really mean to talk about data ownership? The better conversation is to ha is to talk about something like control, so just it is in the way that artists and musicians and authors can use a creative commons license to dictate the terms around how people should use their work. Maybe that’s the kind of conversation that we need to be having about data. So let’s say that you built this crazy machine and i press a button and something happens who should get the data? You built the machine, but i’m the one who generated the data by pushing the button. Now.
Why is this complicated? So many of you may have heard the story about hugo campos he’s a patient that has a medtronic cardiac defibrillator and he can’t get access to his data now on one hand, we as patients should have access to this data. It’S our body after all, but if you gave me access to my data, would i know how to interpret it? Would i have the right, contextual information and background to be able to interpret it accurately or what happens if i misinterpret my data and i make a bad medical decision because of it who’s liable? Is it my fault or let’s say that you get access to your data and you store it on the cloud somewhere and somebody else accesses it and either intentionally or inadvertently, manipulates your data, and then you make a decision based on that, you can see how this Has a lot of potential to go wrong, so i’m not saying that patients shouldn’t get their data or that users shouldn’t get their data. I’M just saying that there are a lot of really tricky privacy security liability issues here and people are definitely working on these, but i’m going to table that for now. So let’s say just imagine, i could get a skeleton key that would allow me to get all of my data. Well awesome am i done. Maybe does your wi-fi toaster speak the same language as your washing machine? Most of our devices aren’t even speaking the same language.
We’Re in the wild west of the internet of things we have a tower of babel. Here our devices are speaking different languages. How can we take data from multiple devices and try to do interesting things if they’re, not even speaking the same language? Now, going back to my example, i’m wearing all of these devices and they all track steps. What is a step anyway is a fitbit step, the same as a jawbone up step.
Is that the same as a misfit shine step, and we do activities that are not just walking? What does a step mean in the context of swimming? Is a step the same for a backstroke as it is for butterfly how about rowing or yoga? How many steps are those worse worth? It’S clear that this is a limited metric, so each company has tried to fill in the gaps by coming up with their own system, so the fuel band will give you fuel points. The jawbone up counts. Your number of active minutes. Fitbit gives you an active score so for any given device.
It gives you a more complete picture of how active you are, but if you want to try to compare across devices good luck, so maybe we need some standards here. This kind of sounds like early days of the internet, where we had many different platforms. Maybe we need to come up with some standards, but the problem is that we’re still prototyping, we don’t know the right questions to ask we.
So this is just one one context in the health space we can count steps, but we don’t know what to use them. For medical applications might be a little bit easier, but if we’re talking about the internet of things, we don’t want to restrict ourselves by tying ourselves into standards too early. What we do care about, though, is that our devices can communicate with each other. So before we get to having standards – and we can’t have true interoperability without that – but what we can do is say: okay, we agree to have the apis and to document what we’re doing thoroughly. So we want our devices to communicate with each other.
One question we have is: how do you balance being competitive as a company with being open? I want to be able to give people access to my data, but i also know that i, if i’m a company, i need to be making some revenue. So i want to focus on a particular type of hardware company, so the kind of hardware company that i’m going to talk about is hardware companies that build the device to get access to data that they wouldn’t otherwise have so for those companies um. How do you balance this? You? You want to build a device, so you can get data. You want to do right by your customers by giving them the access that they deserve, but you also need to be sustainable as a company, and so i’m just going to walk through a couple or a few different examples of business models that allow you to share Your data and i’ll go through some of the pros and cons of each so one of them be say: pay per api call and companies like google and qualcomm do that.
There’S a clear value proposition on both sides. But it’s hard to do that. If you don’t have a big enough reach, another model is reciprocal data sharing, and so this is a model that jawbone just announced that they’re going to be using and i’m really excited about it. So the way that it works is that they have two-way integration. With a closed set of other apps or services – and this is cool, because both companies trade data – and so it’s clear why it’s valuable on both sides.
It’S a problem, if you’re not part of that closed set. But this is one way that we can start to move forward. Another option is pay for data download and companies like body, media and fitbit charge a subscription fee, and let you download your data as a csv file makes it a little bit hard to interact with other apps or services.
If you do things that way, another option is having a free api for trusted parties and again fitbit uses this as well. It’S great for your trusted parties, but it doesn’t scale. You can build your own platform, and this is like that that apple app store or google play or run keeper, and so, if you have enough traction on your platform, this can be a great option. But it’s also hard to start out.
Another option is say: open data share sharealike. This is like creative commons for data. You can have a system, that’s completely closed. It makes it much easier to integrate throughout your system.
You don’t have to worry about making things accessible. The flip side of that is that nobody else can use your data. You can have a system, that’s completely open and so examples would be like zeo or the personal genome project or in the past, twitter or makerbot, and the advantage of that advantages of that is that it’s great for driving community engagement. The downside is that there may not be a clear revenue model, so there are a number of different models that people are using, which one is the right one.
Well the jury’s still out, and it depends on what you’re trying to do. Maybe it’s a mixture of these, maybe it’s none of them at all. So where do we go from here? Well, if we want our devices to talk to each other, we should start out by having us as humans talking to each other. Now i have the great luck to live in the bay area, where there are awesome communities of people that meet regularly to talk about what we’re working on.
One of them is a sensor meetup that i run. Another is the hardware startup meetup that nick runs and they’re actually hard ups hardware, startup meetups in many cities in the united states and i think outside as well, and so it’s great to be able to talk to other people working on projects. And this is important, because our vision that we have of this internet of things is much bigger than any one of our individual companies and even though most of our companies might not make it we’re going to learn a lot in the process and as a field. We can leverage this and use the community to be a rising tide, to lift all boats after all, isn’t that why we’re here thanks .