Moving from specification to specification. I recently finished the notes on INSTEON, it is definitely an interesting technology and solid enough a specification to make it interesting for OpenRemote to integrate with. It should be much more robust compared to X10 and on the plus side INSTEON still maintains backwards compatibility with X10 so we're still covered there.
On the downside, INSTEON seems to be pretty non-existent in Europe. Something to do with that 110V vs 220V power I guess (I wouldn't know, I'm clueless on hardware). But the other reason may be there's already a strong contestant in Europe, KNX.
Going through the KNX website it gives me a pretty convincing story. You've got an open standard (ISO, ANSI, EU, even Chinese), a comprehensive membership roster and partnerships with installers, training and certification program and tools to help installers and integrators. On a quick glance it looks like a full ecosystem, maybe bigger than INSTEON? Something worth investigating.
However getting into KNX does require a bit of an investment. We are looking into that. So far despite the cost the technology looks interesting. There's a bus architecture, independent of manufacturers and set of proprietary tools to get installers, systems integrators and software developers on to the platform.
On the software side the good news is that we can integrate via IP stack -- using UDP -- that we will get from Linux on our controller. On top of that we'll need to implement the KNX's telegram. Some products that do this are already out there. The catch: implemented as MS Windows DCOM components and costing a pretty penny.
We'll see if we can do something about that :-)
I've started a page on KNX where I'm currently collecting notes. Feel free to add (please turn on the diffs if you do) and stay tuned. More to come on that page.
KNX definitely sounds right.
Also now that the telegram is on UDP/IP over RF/low-voltage etc there is no reason for it to not work in the US. I was reading about their panel support and you have some pretty slick in-wall design coming from a variety of vendors in Europe. The fact that AMX went ahead and interacts with KNX certified devices is a plus for them. If I can buy KNX compatible hardware, and have it work in the US that would be a huge plus. Shading, lighting, HV, Security, Metrics.
JL. I will be in Europe and would love to play with lighting control at least. I was reading that the reference handbook you talk about contained a book on where the expected behaviors in applications are specified. I could take a look at that. Do you have it?
I'm not sure that is the case. The spec will have the final answer (sent you an email about this) but the impression I got so far is that Ethernet is one of the supported media (RF, TP, PL being the others) so I would expect they use the KNX telegram on PL or TP, maybe even RF -- don't know. They mention the RF protocol is in the spec. So I suspect its the 'telegram' that goes over all media but not necessarily IP (so it's not routed).
But we will see once we get our hands on the spec CD.
I've no idea if we can program to RF directly (what is the interface there?) but obviously RF works in the US just the same as Europe, and probably twisted pair too (ethernet obviously yes), so it might be just the powerline that is an issue in the US (KNX is selling circuitry for PL so obviously that's EU specific, unless you can get circuitry for US from AMX maybe?)
Yeah noticed the panels too. Didn't really know if it's difficult to get good ones for other systems. Obviously the standard and vendor ecosystem helps with number of choices.
Don't have yet, send me your new address in Spain.
And install JDK, you'll need to code! :-D
Juha Lindfors, OpenRemote
Join OpenRemote Chat
Well everything that is protocol is not tied to the AC specs. The only AC media specific is powerline and that would apply to light switches and endpoints that derive power from the line, which is a lot of the devices that work in the KNX ecosystem but nothing too complicated. There will be vendors in EU or US that will adapt for the US market if isn't done already. As Neil said
yes, I think that is the big point of KNX and why for us, as a software vendor it would make sense to integrate with it ASAP. It is the closest thing to a EE ecosystem as we knew it and played it back in the days. Except it costs $1000 not 500k :)
You got it! sign me up as lead on KNX module implementation. I am ready to spend some time on coding actually. Haven't done it in years, so I am rusty, but it is the only way for me to get down to business.
Well, we can certainly try to give it a push.
Yes, my thoughts exactly. That's what makes it interesting and somewhat familiar. And that hopefully helps.
Alrite ;-)
Juha Lindfors, OpenRemote
Join OpenRemote Chat