IoT vs. WoT
November 25, 2014 at 4:57 pm #1461
I was reading an interesting discussion this morning that would be very appropriate for ForgeRock. The question was “What’s the difference between IoT and WoT?” (Internet of Things and Web of Things) At first it seems like a fairly benign and simplistic question, but the discussion around it was quite good.
On the surface, it would seem they are the same thing, but making a distinction is actually important. IoT, in this discussion, was described as the lower levels of the stack. The hardware, the physical layers, the connections, etc. That makes perfect sense.
WoT is the higher levels. The authentication, the ReST APIs, the data layers and the http protocol connections. The Identity and authorization and authentication levels lie in the WoT layers. The devices themselves reside in the IoT layers.
Anyway, I found that to be an interesting distinction. I’d mostly thought of the two terms as interchangeable, in large part. But making this distinction in this way makes a fair amount of sense in how it distinguishes what each part plays in the overall scheme of things.
Discuss. :-)November 25, 2014 at 9:55 pm #1463
I suppose I take a simplistic view of the subject – that IoT is the connecting of a number of previously unconnected devices, or simply put, creating a network. I suppose taking the thought further the WoT concept would be more the accessing of information over the network. So perhaps a “level in the stack” argument could be made.
Regardless, I still like IoT as a term to describe the effort. To best sum up my rationale, I quoth John Gage: The Network is the computer.November 25, 2014 at 10:37 pm #1467
Ha! I seem to remember that phrase from somewhere. Network was the dog, too, as I recall. :-)
I, like you, really saw the difference between the two as either semantic in nature, or just an attempt at a marketing differentiation. Cisco calls it the “Web of Everything” right now. <yawn>.
But in thinking more about it, it might actually be a (somewhat) useful distinction to make. The IoT is, as you said, the connecting of previously unconnected devices. The WoT is the accessing of information to/from those devices using (largely) web-based protocols. Overall, I still call the whole thing IoT, and WoT is just a part of the overall IoT ecosystem, but that’s just me. I’m not sure the nomenclature is entirely nailed down on these things yet. IoT is still largely the Wild West in many respects.November 25, 2014 at 10:45 pm #1468
Web of Everything? Wow, ok… That’s slightly hyperbolic.
Like with most things, the simplified view will be the one that gets adopted, mostly to the chagrin of those more deeply invested in the technology/industry. I’ve heard Richard Stallman rail on several times about why it should be called GNU/Linux. While I agree they have contributed significantly, it’ll never be a palatable term for the less invested masses. Linux it is.
For this? I do think Internet of Things has essentially already gelled as the term. It’ll take an especially cleaver marketer to change that I think.November 25, 2014 at 10:56 pm #1469
This blog post from Cisco, from a few years ago, was laying the groundwork for such a thing claiming that IoT was evolving into WoT. (http://blogs.cisco.com/sp/from-internet-of-things-to-web-of-things).
Now, I’ve been a member of the IPSO Alliance since it was founded — we were founding members and my friend Pete is the current president — and I don’t see any movement at IPSO towards a WoT from an IoT.
I’m not convinced that IoT has ‘won’ over WoT as I see a LOT of discussion around both. I think it diverts attention and makes things more difficult, but that’s the way things are. :-)November 25, 2014 at 11:13 pm #1473
I did the highly scientific Google Search check. IoT has 190m results and WoT has 176m results. That’s a lot closer than I would have figured. So you’re right. The race is on.November 25, 2014 at 11:15 pm #1474
Next up is to see what the Vegas bookie’s odds are on each, then we have some scientific basis. :-)November 27, 2014 at 12:42 pm #1518Simon MoffattParticipant
I was at the IoT World Forum this week and that topic came up! There were several presentations referring to the evolution of IoT. Starting at the Intranet of Things level (silo’d devices), through to Subnets of Things (internal organisations and enterprise device connectivity), then on to the now common IoT (multi devices connecting to public cloud services) and then finally a Web of Things, which is much more semantic and contains the analytics and service layers that interpret the the vast amounts of data that are being generated at lower levels of the stack. The WoT layer always has a heavy lean towards people and process, that are leveraging the underlying physical device layer.November 27, 2014 at 3:59 pm #1519Jamie BowenModerator
My take on it is this;
The internet is a vast amount of computers networked together using TCP/IP, while the World Wide Web is a way of linking the information on computers that are linked to that network using http, browsers and web servers.
There are many other protocols and networks that use the internet; email via the SMTP protocol for example.
As you’ve agreed the internet of things would appear to be the job of connecting devices to the network, using TCP/IP.
The web of things I have more trouble with. I don’t think it exists yet. Because of the nature of things, HTTP is not a good protocol for the Web of Things. It works for some elements but not others. This is because the WWW uses REST over HTTP and hypertext links to link resources over the internet. If we look at what a resource is, we see that it is typically a fairly static piece of information, e.g. a web page, a weather report, etc. That may even describe a thing pretty well, but it’s becoming evident that things have sensors, and produce information that is not static. Think of a web cam, and we see that the web cam is a thing with a URL, but it produces a video feed that doesn’t fit into the WWW protocols very well, hence years of having to have flash plugins and stuff before we have arrived at html5 with its ‘add on’ standards for various media feeds.
Things then are similar to our web cam. They are resources on the internet, and so they can have URL’s and be treated as resources, but they also generate events, which is where new protocols such as MQTT and CoAP fit in. So the Web of Things is something different from the World Wide Web (Web of Information), and it is still in its infancy, without accepted standards, but with plenty of competing standards. It’s still wide open.November 27, 2014 at 4:17 pm #1523Simon MoffattParticipant
Agreed. There is a view of the little data to big data flow. So the little data is the web cams, sensors, PCL’s etc operating on CoAP or MQTT, pushing tiny pieces of data at stream like speeds, in isolation, into a local broker layer, that in turn can capture, aggregate and relay that stuff into the HTTP world.
I think when these disparate pieces of data then reach the cloud service, does the distributed storage and centralised querying take place, that the “Web of Things” can actually develop.
I see the WoT as sitting on top of this mass distributed data pond, using semantic analytics and intuitive non-app based interactions (NFC or beacons for example) to provide context and personalisation, without the need for decision making. Hence the people and process aspect. The WoT will certainly contain many different small pieces of a mass jigsaw, requiring differing communication, storage and access management protocols. Some which are getting there (MQTT, OAuth2, lower level encryption protocols etc) but certainly many aren’t.November 27, 2014 at 8:48 pm #1524Marius @ForgeRockKeymaster
Let’s re-coin the term “
IoI, Internet of Identities” while we’re at it. :)
Where they are – or what they are – or what they speak – are all subordinate to who they are right? Should be true even for a connected fridge. It has an identity that puts it in relation to other identities. It’s not just any old fridge, it’s my fridge. And give it a few years of AI development and it might even have it’s own attitude…November 27, 2014 at 9:02 pm #1525
Just to refine that further, IoAI the Internet of Authenticated Identities since, let’s face it, without strong authentication, IoT is going to be a mess. I certainly want to ensure that all of my connected devices — of which I have many — have strong authentication to go along with their identities.November 28, 2014 at 6:02 pm #1535
As an additional datapoint that this is not a settled matter, W3C has a charter proposal for a WoT working group. Now it’s not all that surprising that W3C’s working group would be Web- rather than Internet- focussed, but still.
But if you go and rad through the charter, it is tailor-made for ForgeRock to step into. They are calling for open standards and open source solutions for things like identity, authentications, etc. as a solution to unlocking the “business potential” of the IoT. And of course layering all of these within the well-known http and ReST communication mechanisms.
Getting involved with this W3C working group at an early stage would be an ideal move for ForgeRock’s IoT team.
You must be logged in to reply to this topic.