kurtmccaslin
Active Member
Chris, this is getting interesting with your planned support for additional protocols. I would like to prewire for this in my new house. Do you have any recommendations on the best way to prewire for CastleOS?
CastleOS - new home automation software with Kinect voice control
- Thread starter ChrisCicc
- Start date
ChrisCicc
Active Member
rockinarmadillo said:
With the protocols we support, no pre-wiring is necessary for things like light switches and thermostats and sensors. But if you want a tablet in the wall, like an iPad mini, you should prewire each run with at Cat6 cable which will actually be used for low voltage power to run the tablet, that way no wires are exposed. See this link for more information about tablet mounts.
kurtmccaslin
Active Member
Any prewire required for the Kinect box? Best location for the box?
ChrisCicc
Active Member
rockinarmadillo said:
That depends, it's not necessary, but depending on what you wanted to do you could run USB extensions back to a central location. What we recommend though is to pair our Mini Computers (really Intel NUCs) with a Kinect, and plug it in to each TV in the home, providing dual purpose voice control and cable TV, while offering a full computer to do things like Netflix, Hulu, pictures, etc. That's what I have in my home, and what's used in the video that Discovery shot.
Longhorn84
Member
What's the word on Z-Wave? All devices in your store are insteon?
NeverDie
Senior Member
I'm don't know whether it's possible, but if leveraging open z-wave would allow Castle OS (or any vendor for that matter) to offer better transparency into what's happening inside z-wave than the usual black box approach to z-wave that the major HA vendors provide, I think it could potentially win you a lot of z-wave customers. The usual excuse that the established players give as to why they don't disclose more about what's going on inside is that they are bound by an Zensys NDA and aren't allowed to say more. Whether that's true or not, I have no idea. All I know is that the lack of transparency can sometimes turn z-wave into a bit of a guessing game as to exactly *what* is going wrong when things aren't working right, and that can lead to time consuming trial-by-error attempts at resolution. I would think that leveraging open z-wave would mean that you're not bound by an NDA, and so you're in a good position to give maximum transparency onto z-wave's sometimes opaque behavior.
At the same time, I also wonder how quickly open z-wave will track actual z-wave. It seems there's about to be new behaviors available from the next gen z-wave chips, and so how will anything based on open z-wave be assured of keeping up with the changes? In that respect, open z-wave may be a double edged sword.
At the same time, I also wonder how quickly open z-wave will track actual z-wave. It seems there's about to be new behaviors available from the next gen z-wave chips, and so how will anything based on open z-wave be assured of keeping up with the changes? In that respect, open z-wave may be a double edged sword.
ChrisCicc
Active Member
NeverDie said:
Hello! We actually had used Open Z-Wave to start, but we had to swap it out for the official Z-Wave SDK due to the fact that it does not (and did not appear it ever would) support secure Z-Wave devices like locks. In addition, there have been improvements to Z-Wave (i.e ZIP and others) that the Open Z-Wave project is not currently tracking. We'll see how it all evolves!
NeverDie said:
The VR functionality is an optional component of CastleOS. It's actually a separate app too, called the "Kinect Service". The central controller piece is the "Core Service". If you don't want to use the VR, you can skip installing the Kinect Service and use the web and mobile apps only.
Also, we currently sell the CastleOS Mini Computer which can support up to one Kinect, we're working on an upgraded version that will be less expensive, and support as many as four Kinects. We also plan to release the Kinect Service for free use with media players like MediaBrowser, XBMC, Plex, and DirecTV (use for automation will require the Core Service, which is the licensed piece). That's coming very soon!
NeverDie
Senior Member
Kudos to CastleOS for the rapid response! That beats the response time for the HA package I actually own by huge factor. 
For what little it's worth, I would vote for gesture recognition over VR, unless it's VR that can be whispered below human perception or somehow subvocalized. If I were living alone I might feel differently, but living as I do in a typical family, the thought of family members unexpectedly shouting "Computer! Do This!.... Computer! Do That!!" at random times day and night :nutz: would seriously undermine our zen.
If it weren't for that, I admit it would be a very desirable feature to have if I could say, for example, "Decrease temperature by 1 degree" or "Activate Lighting scene Tango" or the like and have high confidence it would actually happen.
For what little it's worth, I would vote for gesture recognition over VR, unless it's VR that can be whispered below human perception or somehow subvocalized. If I were living alone I might feel differently, but living as I do in a typical family, the thought of family members unexpectedly shouting "Computer! Do This!.... Computer! Do That!!" at random times day and night :nutz: would seriously undermine our zen.
If it weren't for that, I admit it would be a very desirable feature to have if I could say, for example, "Decrease temperature by 1 degree" or "Activate Lighting scene Tango" or the like and have high confidence it would actually happen.
ChrisCicc
Active Member
NeverDie said:Kudos to CastleOS for the rapid response! That beats the response time for the HA package I actually own by huge factor.
For what little it's worth, I would vote for gesture recognition over VR, unless it's VR that can be whispered below human perception or somehow subvocalized. If I were living alone I might feel differently, but living as I do in a typical family, the thought of family members unexpectedly shouting "Computer! Do This!.... Computer! Do That!!" at random times day and night :nutz: would seriously undermine our zen.
Believe it or not, you're far more likely to generate a false positive from gestures than from voice commands. Our false positive rate is virtually zreo. And you also avoid the complication of having to have end users program custom gestures. All in all, if you're worried about a loud family tripping the system, don't be! Also, it can easily detect normal conversation voices - but that is background noise dependent