Really? Another light switch?

[Timoh]

Interesting, but not so interesting project over at KickStarter...
http://www.kickstart...n-smart-home-te

Do we really need yet another light switch? And really I don't get how this is different from z-wave or RF insteon.

What I would like to see, and maybe I should fire up my own Kickstarter project, is embedded 802.11b/g/n in a light switch, give every light switch+ in your home an IP, and be done with all this "which protocol is better" nonsense!

Tim

 

[tmbrown97]

This cropped up briefly when one of the project members posted about it here: http://cocoontech.com/forums/topic/21737-recommendations-for-a-newb/#entry176925

It'll be interesting to see where it goes - it looks to be a full mesh network that communicates more info than just standard on/off, including the ability to track the power consumption of the load behind the switch - that part is pretty interesting. They also appear to want to be as open as humanly possible with their protocol - I doubt there'll be any licensing issues or other such barriers if others want to experiment with them. That said, they probably won't have the backing that comes from having a bunch of large players getting together and forming an alliance.

Time will tell; I can't say any one switch on the market today is perfect - so there's still room for people to experiment.

As for all devices being IP based, there's a reason that doesn't happen; 1) the power requirements are higher; 2) the overhead of tcp/ip would be much greater than the packet sizes of what's being transmitted; 3) Believe it or not, there's a limit to how many wifi devices a single access point can handle (the radio time has to be sliced up and shared between devices) - even the best commercial AP's on the market don't do well past 30-50 connections.

To go with that last point - I already in my home have about 20 devices... I have about 65 UPB switches today; and I'm starting to experiment with Ubiquiti's mFi products which is upping the number of IP devices for their mPort and mPower - if I were to add all these together, I'd easily have 100+ wifi devices in my house. That means a whole lot of network traffic, and means I'd have to turn down the power on my existing AP and instead place 3 AP's throughout the house.

So many people think that IP is the answer to everything - but it's clearly not - it's a relatively bloated protocol when you think about the small size of the data packets needed for things like this. Also, the required network knowledge to support such systems just isn't there with most home users - and honestly, a lot of IT people don't know as much as they think they do about it.

 

[jaymartin]

One interesting feature (though of questionable usefulness) is the switch's ability to not control the local load when physically activated - that lets the HA system jump in between the press and the activation. You can't do that on INSTEON devices (it's really too slow anyway) and I don't believe you can do it on Z-Wave or ZigBee devices either.

I'm interested to see how it works out for these guys. We'd certainly welcome a plugin for Indigo for them.

 

[JonW]

Hey techiedavid - you have 3 posts today and they are all about the kickstarter project you linked. Are you a shill for them?

 

[damage]

I am more interested in this kickstarter project:

What is your "expert" option on this kickstarter project? Is it worth it?

the name - uConTroll - doesn't inspire much confidence

 

[Timoh]

To your point work2play, I think the only show stopper for IP based would be the number of devices on an 802.11 network, as you say.

I doubt that power consumption is going to vary and/or be a deciding factor on which HA light switch people use. Protocol could be a factor, I totally agree that it is overkill (aka bloated) for some applications... But if you're using some lightweight/zigbee/802.15.4 protocol, and then sticking bloated XML onto it, you're not going to be much further ahead. (Unless in this case, the Luminode dongle is doing some magic and translating xml into something proprietary & more compact.)

Those points aside, it will be interesting to see where it goes... Power monitoring + the ability to inject HA between press & activate, are not really selling points for me. Now the ability to control the colour of the LED might be!

Tim

PS: I'm concerned with the Troll in the ucontroll... I've got enough gremlins in my house I don't need trolls too.
PPS: In reading up on 802.15.4, I'll back peddle away from suggesting 802.11+IP could be used.

 

[tmbrown97]

I sure hope they have something more compact than bloated XML being transmitted OTA!

Power monitoring is a nice feature for some of us - I have electric rates up to $.41/kWH which is astronomical, so being able to see where every penny is going is a big interest to me. That's why I'm also looking a the newly announced Ubiquiti mPower products, as they allow you to monitor power consumption theoretically per-outlet on a power strip, and control outlets based on other factors.

UPB today could interject other automation in between if you want to - I'm not sure how bad the delays would be, but you can have it so the rocker doesn't control the switch but triggers a phantom link which your HA software watches for; once it sees the phantom link, it could then perform a series of actions. Using this you can set it so that an "on" press does different things depending on the time of day.

 

[tmbrown97]

As for techiedavid above, I think Troll is the key word here!

 

[techiedavid]

Hey techiedavid - you have 3 posts today and they are all about the kickstarter project you linked. Are you a shill for them?

I did pledge $100, but it is having a slow start I'm starting to wonder if it will meet it's funding goal, so I started to look around for alternatives and was wondering if I was missing sometime and people knew of something better.

 

[wolfdown]

XML .. you have got to be kidding me. Totally useless on an embedded system.

 

[az1324]

What is really needed is for someone to go all out on economies of scale so we can get X10 pricing with new technology ($15 per switch). It could happen when some lutron patents expire in 2016 but it seems more likely that the mass market will come through LED smart bulbs than switches. Although smart bulbs with dumb switches has its own set of problems.

Also for those calling for IP https://en.wikipedia.org/wiki/6lowpan is seeing some backing by the SoC manufacturers.

 

[Timoh]

TechieDavid, I don't think you are missing anything... IMHO I just don't think the demand is there for the extra features of those switches compared to Insteon, Z-wave, etc. To entice people to try something new, I think they'll need a much better price point.$30-$40 would be interesting, and sub $30, they would probably get a lot more pledges.
Tim

 

[jhdale]

Putting home automation devices directly on the IP network seems like a security mistake. Every device is exposed if someone outside the house wants to monitor you or control your devices. E.g. turn on your light and see if anyone is there to respond. At the early stages it has security through obscurity but I wouldn't want it to catch on.

If you have a separate network (e.g. Insteon, UPB, Z-wave) with an internet-enabled controller, you only have to worry about security at that one point. And if there is a problem, you only need to update/upgrade that one single device, not every single light switch.

This becomes even more of an issue if you use it for functions other than lighting, of course.

 

[tmbrown97]

While I don't think IP-enabling every device is the answer, I do love to play Devil's Advocate...

One can essentially get the same level of separation by having their automation network on its own subnet and using firewalls within the house to control what traffic passes and from what source - and by using strong authentication mechanisms... then again, regular IP networks are above most homeowner's heads - this is getting into a realm that a lot of lower level network administrators wouldn't even be comfortable with.

 

[Teken]

I wish the inventor all the very best and look forward to seeing them in the open market.

Teken . . .