Diso - also for real-world objects?

0 views
Skip to first unread message

Vlad Trifa

unread,
Nov 18, 2009, 12:24:09 PM11/18/09
to diso-p...@googlegroups.com
Hi folks,

I've been passively observing this list for some weeks and found excellent the ideas flowing here, but never really jumped it. I just wanted to say hi and introduce myself, and hopefully take this riding train and start some great discussions here.

Briefly, my name is Vlad Trifa and I'm researcher at ETH Zurich and SAP, and my work is very much about building the Web of Things. Unlike the über-used internet of things, our approach fosters to reuse as much as possible existing Web standards (so REST, as opposed to the zilions of existing protocols for home automation, devices, machinery, etc...) to build interactive applications on top of heterogeneous embedded devices (mobile phones, chumbys, sensor networks, wireless routers, and so on). Our goal is to leverage the existing Web infrastructure and simple access barrier as a solid backbone that can be extended beyond humans with browsers and web sites/api, towards physical objects that contain tiny web servers onboard, thus support http(s)/xml/atom, and hopefully other emerging web patterns (social, real-time, programmable web). Ultimately, our vision is a future Web-based ecosystem where physical things can be directly reused, shared, linked, hacked to build "physical" mashups, where you go beyond "just real-time data".

In any case, I foresee a huge potential to think about beyond "just people" in the future Web towards a simpler integration of the real-world and build a physical web, and for that I think the efforts I saw here are great and I would love to explore them with this machine-machine aspect, or machine "users" of future social nets.

Although this is not the core topic for diso, I think it's an important part of the future web and I would love to know what you guys know about ongoing efforts in the direction of "social device networks", and how you see devices fitting within the diso vision. I am totally looking forward to discuss that with you and would love to share more the things we've been trying/playing with. You'll find more infos about our work on our (quite sleepy) blog webofthings.com.

Looking forward to read your thoughts and reactions on this topic.

phfeww, we can breathe now after this intense mail (and congrats for those that made it up to here)!

Cheers,

Vlad

Steve Ivy

unread,
Nov 18, 2009, 12:41:05 PM11/18/09
to diso-p...@googlegroups.com
Hi Vlad, and welcome!

I think that a lot of folks on this list and the activity streams list
[1] would be interested in the work you're doing. The Activity Streams
work specifically ties in closely with this idea of connected devices
-- we've discussed before the idea of these devices producing activity
streams detailing relevant events.

Feel free to join the activity streams list [1], check out the specs
[2] and chime in!

--Steve

[1] http://groups.google.com/group/activity-streams/
[2] http://activitystrea.ms
> --
>
> You received this message because you are subscribed to the Google Groups
> "Diso Project" group.
> To post to this group, send email to diso-p...@googlegroups.com.
> To unsubscribe from this group, send email to
> diso-project...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/diso-project?hl=.
>

--
Steve Ivy
http://redmonk.net // http://diso-project.org
This email is: [ ] bloggable [x] ask first [ ] private

Chris Messina

unread,
Nov 18, 2009, 7:20:51 PM11/18/09
to diso-p...@googlegroups.com
Welcome Vlad.

+1 to what Steve said. You'll find that the activity streams model and concept, while "actor"-centric, does not preclude actors from being [social] objects. So yes, your toaster could, for example, emit an activity stream about the toast that it just popped out.

As well, there is ThingLink (connecting to real objects):


And I wouldn't be surprised if this isn't the direction that Opera Unite wants to go:


...but, they're a LONG way from getting there.

They do, at least, support activity streams.

Chris
Chris Messina
Open Web Advocate

Personal: http://factoryjoe.com
Follow me on Twitter: http://twitter.com/chrismessina

Citizen Agency: http://citizenagency.com
Diso Project: http://diso-project.org
OpenID Foundation: http://openid.net

This email is:   [ ] shareable    [X] ask first   [ ] private

Vlad Trifa

unread,
Nov 19, 2009, 7:32:45 AM11/19/09
to diso-p...@googlegroups.com
Great, thanks!

I've been exploring notifications mechanism and web push solutions around rsscloud, pubsubhubbub and webhooks, and recently started reading about activity streams. I definitely agree that as make much sense when it comes to social objects, or even more to enable social objects to use AS, but when we (okay, I) think more about more sensor web scenarios (pollution monitoring, traffic status - i.e. real-time sensed data vs. events), it seems to me that AS are not optimal for this task. Although open and configurable, they were made for people life-streams of all kinds, and I'm not sure how this model maps to simple devices that report their sensed data periodically, am I wrong?

The thing with browser servers is what you do when you're behind nat/firewalls? You need some hack to have it running (like comet-based streams or something like that)... of course a clean bidirectionnal solution for http messaging would rock, but well.... LONG way from here as you said.

I will investigate that anyway more in detail and come back with questions and suggestions/ideas.

Cool!

Vlad

Steve Ivy

unread,
Nov 19, 2009, 10:53:45 AM11/19/09
to diso-p...@googlegroups.com
Victor,

Perhaps one way to approach this would be to imagine it from the
client side - how do you imagine this sensed data being consumed? If
you're looking for a machine-to-machine protocol, then yeah, Activity
Streams probably aren't the way to go. But if you're looking to
populate, say, a user's dashboard with up-to-date information, then I
think Activity Streams has a lot too offer, with understandable verbs
and social objects. Another question to ask is - is the data from
these devicdes going to be of interest to a larger population of
connected users? Then, again, AS is a good fit.

--Steve

On Thu, Nov 19, 2009 at 5:32 AM, Vlad Trifa <vlad...@gmail.com> wrote:
> I'm not sure how this model maps to simple devices that report their sensed
> data periodically, am I wrong?



Steve Ivy

unread,
Nov 19, 2009, 10:56:05 AM11/19/09
to diso-p...@googlegroups.com
Oh my goodness I flaked on your name, Vlad. I can't believe it. SO sorry.

Vlad Trifa

unread,
Nov 20, 2009, 5:46:45 AM11/20/09
to diso-p...@googlegroups.com
honestly, I'd rather hear me called Victor than hearing my name
pronounced by my German colleagues, they all call me Flat...

Anyway,

I think you have an excellent point in that depends how/where the
humans enters in the loop. AS are excellent to notify human users of
real-world events based on high-level transformation/aggregation/
filtering of raw data such as "peter came home" as sensed by different
sensor in his house or "your amazon shipping will be down your house
in 24 minutes" or "michael, your dad will come pick u 10 minutes late
coz he's stuck in traffic", etc...

What's nice behind pushing http in all parts of your system (as
opposed to use custom protocols for the sensing networks and only
offer the data they sense through a web-based API) is that you can
leverage the features offered by proxies such as caching, but also
transparent data processing hidden behind a restful interface. It's a
nice way to introduce processing units to transform high-frequency raw
data into high-level composite/mashupped information using a sweet
layered processing approach with different level of granularity
different clients can choose from. This is done easily and one can
nicely map between directly between AS and other web push/pubsub
systems without complex protocol transformations.

I'll definitely be checking out this in more detail!

Thanks for your thoughts!

Vlad
Reply all
Reply to author
Forward
0 new messages