Dancing at Two Weddings: WebRTC & iOT – not

25/08/2014

Here’s my official opinion: WebRTC & IOT aren’t related. Not really.

WebRTC and IOTThese two? mix the best on a logo and nowhere else.

I’ve been noticing recently a lot of discussion around WebRTC and IOT. IOT here stands for Internet of Things and not Interpoerability. I’ve been thinking about this topic on and off for the last year or so, and my solid opinion is that there’s nothing really related between these two.

The main reason for me to take this stance now is several incidents where I’ve seen the two being placed together:

  1. There’s the Vidyo raising some more money… apparently for “video” IOT
  2. @ThingsExpo’s full track with the title “WebRTC – Backbone of the Internet of Things”
  3. A few vendors I have talked to who tried working the angle of IOT for WebRTC

The critique hasn’t been out there to state that this is just trying to hitch on the bandwagon of IOT.

I’ve seen it in the past in the video market. Green technology? video conferencing.  Eyjafjallajökull erupts in Iceland halting all European flights? video conferencing. Travel cost reduction? video conferencing.

For some reason, this type of thinking is now also plaguing WebRTC, where we’re trying to stick excuses to using it instead of focusing on benefits.

Why IOT and WebRTC don’t mix

So here’s the problem I have with this approach of fuzing WebRTC and IOT:

  • WebRTC is for people. IOT is for… Things
  • While WebRTC has a great utility called a Data Channel, it is used only in the browser. Anywhere else it is rather useless
  • IOT on the other hand, runs browserless. Because it gets embedded into… Things
  • I once thought that it would be great to place WebRTC’s data channel on a Raspberry Pi and run it, but since then left the idea – simply because you can just put Node.js on a Raspberry Pi and P2P with it to your heart’s content sans WebRTC. There are already good open source mesh frameworks for use on devices – no need to add WebRTC
  • The place where WebRTC really fits into IOT? When you want to go low latency between people who are interacting with each other and there’s some Things they need to handle for that purpose. Not that I have a specific use case for it at the moment

IOT is now talked a lot about and it garners huge investments in startups and acquisitions – a lot more than anything in the WebRTC space. It is no wonder that communication vendors are trying to put an IOT sticker on their products. But for me it makes no real sense.

Back to my reasons

Vidyo touting IOT? Probably just a way to raise yet another 20M USD, after raising more than a 100M USD in the last decade.

@ThingsExpo? Events… attract audience by placing enough buzzwords in headlines.

Vendors? Trying to attract customers.

My suggestion? Say things as they are and don’t force yourself on areas you don’t really play (or even understand).

WebRTC vendors should stick to improving user experience of communications. There’s nothing wrong with doing just this one thing and doing it best.

Responses

Dean Bubley says:
August 25, 2014

I’m thinking through this at the moment as well.

My take is that devices fall into a few different categories:

– Ones with screen + camera/mic, eg a PC or phone
– Screen-only, eg digital advertising hoarding
– Camera/mic-only, eg CCTV camera or drone
– Neither camera nor mic, eg a sensor

The easiest wins for WebRTC are in the first category, as they are used by people – and they “watch” or “listen” to people as well.

The second group are interesting, because it’s really easy to add a camera to a screen-only device in future. ATM machines might fit here.

Non-display devices might use bits of WebRTC – eg Opus codecs & RTCWeb protocols in a deskphone

But non-visual/non-display devices? Can’t really see the point, unless perhaps they federate to other local devices for display or camera. (eg maybe a Nest + Android combo?)

Reply
    Tsahi Levent-Levi says:
    August 25, 2014

    Dean,

    That’s a good way to segment the various types of devices. Chromecast can be seen as a type of a screen-only device if you want to take a stab at it (it also uses WebRTC).

    Reply
Martin Barclay says:
August 25, 2014

IoT is for things and people…people (consumers) connect to and use IoT via smartphone apps and browsers (monitoring cameras, getting alerts, controlling devices). Along the lines of what Ubley said (federation), IFTTT enables consumers to mash up IoT sensors and devices into recipes that they’re connected to via smartphone app. If door opens, take a picture and send it to me.

Reply
Tony de Sousa says:
August 26, 2014

Seems that in the internet era lots of people and companies jump on these hype cycles to garner some attention and hopefully some easy money.

It’s all about marketing and “fake it until you make it” ethos. I fear this attention deficit disorder of chasing the latest buzz is diverting resources from building real products and sustainable companies.

Reply
David Alozie says:
August 26, 2014

Hi Tsahi, this is a good discussion, I also wrote a bit about this in my last research paper.
well, it is very true WebRTC has to do with people, and IoT has to do with things.
but the fact is that people always deal with things.
Technologies might not need to relate for them to work together, it depends on what a developer/designer/company/organization goal and objective, if they feel they could combine this to solve a problem, then it is all good.
Its possible data channel might be strong point of WebRTC to be explored by IoT.

Reply
    Tsahi Levent-Levi says:
    August 26, 2014

    David,

    Never said the two can’t be used – just that it doesn’t make sense for a 100% of the vendors I’ve seen announcing IOT support in their WebRTC/video products; and I don’t see the real appeal of having both in long tracks of events.

    You will see this connection coming from those who deal with VoIP and WebRTC and never from those dealing with IOT. That being the case, I question the need.

    Reply
What’s In HTTP/2 For The Internet Of Things? 1/2 | A Bunch Of Data says:
April 2, 2016

[…] What’s interesting is that a significant number of the improvements CoAP brings are also covered by HTTP/2 (e.g., binary format, header compression, etc.). However, CoAP goes an extra mile and provides further optimizations such as the use of less-resource heavy security (using DTLS), resource discovery mechanisms or the use of UDP instead of TCP. UDP is interesting for very constrained Things as the protocol is less resource hungry than TCP. UDP is not only interesting for Things but also for IoT Clouds as it is easier to scale because it does not require an actively open connection like TCP. As a consequence UDP will probably play a role in the future of the Internet and the Web of Things and it could also be through technologies like CoAP or other UDP based protocols such as Web RTC (even if there is yet to be seen if there is value beyond the hype!). […]

Reply

Comment