One of the promises of the Internet of Things (IoT) is that everything should talk to everything else. These talkative “things” include sensors, consumer appliances, home automation systems, and even connected vehicles. The frameworks through which such interconnectivity is arranged, controlled, and mediated – that is, how these things “entitle” each other to connect – is going to be a fundamental part of IoT. Managing this “entitlement,” which defines who can access your device data, and under what conditions it can be found and used by others, will be one of the major challenges for consumers and businesses.
The data from all these things will be valuable not just to the companies that deploy them, but also to people or companies operating in other domains. For example, your thermostat might talk to your neighbor’s weather station to determine an appropriate temperature setting, and then switch on the heating when your phone’s GPS tells it that you’re nearing home.
It’s this many-to-many and cross-domain aspect of connectivity that distinguishes IoT from earlier remote monitoring/control systems and M2M (machine-to-machine) systems, where only one organization created, owned, and used the data. In the IoT, each connection won’t be predetermined; these things should be able to structure their conversations on the fly, in an automated and ad-hoc manner. But this raises a number of questions and concerns around privacy, interoperability, and data-access privileges.
First, how do all these things automatically find each other? In the web world, discoverability (the ability to be found quickly) has been key to growth, but this isn’t necessarily the case in IoT, where many private things will not want to be found. Having a heart monitor at a home can be useful to your doctor, but you wouldn’t want an unknown company to be able to access that data. Knowing you have a heart monitor, a blood pressure monitor, and a dialysis machine could lead to unwanted targeted advertising. Things must define how, when, and by whom they are found.
Once they’ve found each other, how do they “speak” to one another? This has led to a focus on “interoperability,” which is getting things to talk in a common language. Yet even human beings have been unable to converge on a single language to share information. And where we are speaking the same language, we have different ways of describing the world. Some of us use Celsius and others use Fahrenheit. The same is true for our “things”: they are used in a variety of contexts. There may even be a disincentive to adopt a common protocol, data format, or parameter, since that can restrict inventiveness.
Then what happens when the data from one thing is used by another out of context? Imagine a thermostat making a decision based on a nearby temperature sensor, even though that’s measuring the temperature of a baby, not the room temperature. Developers of sensors can’t pre-define all the potential contexts in which their devices will be used. So how might their owners help refine or restrict the conditions in which data is employed?
Most importantly, how do people control which things can access the data generated by their things? People have different comfort levels when it comes to making their data available to others. One person might be happy to make his real-time energy usage profile available to an energy supplier, because that supplier offers competitive rates or services; but another person might not want to release that information. Real-time energy usage data can provide intimate information, including behaviour patterns – i.e., when people are present and active in their homes.
People are also nervous about how their things report contextual and geolocation information (where the things are and what is near them) — especially in the wake of revelations about how the NSA in the U.S. has used such metadata. Some might be happy to let others to see their real-time geolocation (if they’re using a mobile app that helps friends find each other in crowded locations), but others will want to keep it private.
On the web, we have largely ignored these issues. The data that we’ve grown used to making public has seemed innocuous. So people don’t protest too much when their personal profile or usage data is employed by companies like Facebook and Google to deliver highly targeted advertisements. But in the Internet of Things, the data generated by all these things tends to be much more personal and commercially sensitive; and the way it’s consumed (via non-screen devices) means that an advertising model isn’t necessarily going to be sustainable. These challenges don’t mean that IoT is impossible, they just mean that it will be complex — technically, socially, and economically.
Source: hbr.org