At the recent W3C Workshop on the Web of Things in Berlin (workshop report blog post coming soon!), one of the obvious and non-trivial questions of course was: what, exactly, is the
Web of Things? And how does it relate to the
Internet of Things, which is another term that is used quite a bit (and quite a bit more often than WoT, at least for now)?
To me, this question eventually will be answered by one of my beloved Wittgenstein quotes:
The meaning of a word is its use in the language. However, there always is some time until use (and thus meaning) converges, at least in a substantial part of the language community. And it seems like we are not at that point yet. So here is one attempt to say what WoT is and isn't, and I am sure some people will agree, and some will disagree.
We had various discussions during the workshop, both in the forum and in informal conversations. My answer always was a very simple one: The difference between IoT and WoT is the same as the difference between the Internet and the Web. This was always the fun part for me, because that was always my introductory question in my Web Architecture courses, and surprisingly few of our students knew the answer. The answer comes in one sentence:
The Web is a resource-oriented information system that is based on uniform interactions with globally identified resources, and uses the connectivity provided by the Internet as its interaction fabric.
With this definition in mind, WoT can be defined rather easily. It follows the principles of Web Architecture, which in today's Web are mostly embodied in URI and HTTP as the two central pillars of identification and interaction (a.k.a. REST).
This definition also allows to rather cleanly separate the three main communities that were part of the workshop:
- What I call
Connectivists, who mostly care about
the last mile, i.e. how to connect things such as sensors. These I would put squarely in the IoT camp, and while they are of course essential for WoT to even exist, they are not so much concerned with Web architecture.
- What I call
Interactionists, who mostly care about WoT according to the definition above, and who don't really care so much whether sensors are directly connected or via gateways, as long as there are well-identified resources with uniform interaction models and self-describing representations.
- What I call
Modelers, who are mostly concerned with modeling the WoT space. This is what I would call the
Semantic Web of Things (SWoT), according to how the Semantic Web community relates to the Web.
It remains to be seen what (if any) activity will happen as a result of the workshop. These were very interesting two days, and sometimes it was challenging to establish a shared vocabulary when discussing with people there, because many of us had and have different perspectives and definitions in mind.
Maybe treating the IoT/WoT separation in the same way as the Internet/Web separation might help, because that latter one is well-defined, and thus could help to come up with a simple and well-defined separation. But in the end, it will all go according to Wittgenstein: WoT will mean whatever the majority thinks it means, and maybe over the next one or two years, such majority will actually form. I don't think we're there yet.