barrygordon

Yes, mine also has the same three wake up words, and yes there is no way to assign skills to a particular echo at this time.

You can have multiple echos in a home today. The only problem is that all the echos that recognize their name will respond to what is spoken. In the connected home all Echos run off the same data set that is provided by the ISY. There is no fencing. Amazon claims to someday add additional recognition words besides Alexa and Amazon, but still there will be no fencing except through a skill. You can assign a different skill to each Echo, but then UDI would have to be able to have multiple skills. It would be nice if Amazon provided fencing along with multiple wake up words. They would probably sell more Echos. You have to keep in mind that the "connected home" was developed strictly for lighting. It will be interesting to see if Amazon expands that to handle other "Smart Things". I handle my theater using a Pronto PRO as I play it at significant volume and was concerned about using an Echo. The theater is controlled by a PC so I can have Alexa/ISY send any command I want to the Theater's control PC. I do this with a network resource and a simple program with only a then section. I control my kitchen TV in that manner. I currently can turn on the theater by telling Alexa to do so: "Alexa, turn on the theater"; Lights dim, shades drop, music plays, components turn on and warm up . . . I have a program I wrote some time ago that handles all my non-Theater IR requirements based upon commands it gets over raw TCP which is no problem for ISY network resources, or my Pronto PRO's. It deals with the Global Cache line, but mainly with a GC 100 which has 6 IR ports. The program holds all the Global cache timing information for each IR driven device I own, there is a companion program that will develop all the needed timing information form a standard description (know as IRP files) that I co-developed many many years ago with John Fine. A typical input string to that program looks like "DOIR, 1, STB, Channel UP" or "Macro, 1, CNN". It should be obvious what each string does. The named device (e.g. STB) denotes which IR protocol to use. The integer is the GC port number. With a Macro, the macro name (e.g. CNN, Tivo-CNN) denotes the protocol to be used which can be mapped to any GC IR transmitter. I am currently driving three TV locations (Guest bedroom, kitchen, Master bedroom). If I give house guests the Amazon Echo remote, they can control the Guest Bedroom TV by voice via the Echo in the kitchen. I am considering converting that IR system, to an RPi. The IR generator is already on my web site as the IRGEN program. If I do that I will probably integrate it as a node server for the ISY. But that is a ways off.

Probably the best approach would be for the connected home to associate a priority with an entry and search in priority order. I bet there are just a few entries in each home that get 80-90 percent of the action. It is just that when I was using the Emulator it was definitely much faster from spoken phrase to the beginning o the device action.

Jon ... I expect the speed of response is related to the number of entries that Amazon has to search through. That would make sense especially if they are doing a linear search.

I am not having the issue with the word "lights". I was careful when selecting the spoken words to always use "Lights" and never "Light". How the Echo does a match against the speech (what you say) vs its table of spoken names is interesting. If there is a spoken word ending in Lights and the beginning of that spoken word is never repeated in any other spoken word, then my experience is that Alexa will accept either light or lights as the speech ending, but nothing else. If however that beginning speech part does appear in some other spoken name the Echo will reply with "I found several devices matching. . . " . To clarify what I am saying; I have a spoken name that is "Cafe Picture", Which is a picture with an art light. The Echo operates that light just fine if I say "Cafe", Cafe Picture", "Cafe Light", "Cafe Lights", but does not find a device named "Cafe Dingbat". I have several spoken names that begin with "Office" as in "Office lights", "Office Fan" Office Fan Low", "Office Fan Medium", and Office Fan High. In this case the speech must exactly match the Spoken name I wish to operate. To me that makes sense in a best match sense and ambiguity resolution. The only issue I have is how long it takes from the time I finish what I am saying until the device actually changes state. I am seeing times ranging from 3-5 seconds followed by another 2-3 seconds before I hear "Okay". IIRC my experience with the Hue Emulator was an almost instantaneous state change after I finished speaking with a 2-3 second additional delay till I heard "Okay". I have 91 entries in the Portal and the emulator had 93 entries spread across 4 instances of the emulator with the largest on any one instance of the emulator being 35. What type of timing are others seeing?

The preferred phrases for Alea are : Alexa Turn on the <spoken name> and Alex Turn Off The <spoken name> I have had general success with Alexa Shut Off the <spoken name> The word "the" seems to be taken as noise and it works without it. I suspect that your problem may be in the phrasing with the command at the end as in "Alexa turn <spoken name> off

Your PLM has gone to the PLM Hereafter. I always keep a spare in the house. They last about a year.

How about "Alexa, Turn on the Front Door Lock" and "Alexa Turn Off the front door lock" where "front door lock" is a spoken name and have it the spoken name of a program. That should work and is not too confusing, in fact if we are treating Alexa as a house servant, then it actually makes a little sense.

I think dim and all may be bad choices. the Echo takes special pains for the word dim in the connected home as in" Alexa, dim the kitchen lights to 50 %". I suspect special pains may also be taken with all as in "Alexa, Dim All The bedroom lights". I never use those two words in a spoken phrase. I use some thing like "Alexa, turn off the kitchen area", where area is a scene having exactly what I want off.

rhodges, See post #90 in this thread

I am now set up the way I want to be with the ISY and the Echo (I think, but I always end up changing things. I have my own skill named Sarah that I am not using very much. It is just a foundation for later work if and when I want to play with skill development. I am using the ISY Connected Home (ISY-CH) to handle all of my ISY needs at this time. It is working as well if not better than the HUE emulator solution with one caveat. The amount of time it takes the ISY to actually respond (take the appropriate action, e.g. turn a light on or off) is significantly longer with the ISY-CH than it was with the HUE emulator. With the HUE emulator the action was just about instantaneous with the completion of what I spoke. With the ISY-CH it takes 1-3 seconds before the action takes place. Also the time it takes for the Echo to reply with "Okay" also seems quite a bit longer. I hope that this is just a reflection of the development state of the ISY-CH. I am using the HUE Emulator and a small proxy program to handle all non ISY controlled things in my home. At the current time I am only dealing with non-ISY controlled things that are of an on-off nature, mostly associated with my pool system. On the plus side the Discovery process with the ISY-CH and a single instance of the HUE emulator has yet to fail. All my "devices" are discovered in a single pass. The one feature request I have is non-echo related, but rather ISY related. I would like to be able to send a scene a brightness value such that the various dimmer switches that are respondents in the scene are set to that brightness. I understand that this would only be operational through the RESTful interface, but I could use it in several instances. I believe if implemented it would also allow the Echo to dynamically set a "Scene Brightness".

Bgrubb1, I believe the ISY connected home works the same way. You assign a spoken name to a program. turn on <spoken name> will run the then clause of the program and turn off <spoken name> the else clause. I think Benoit stated they were go to try and have run <spokenname> run the if clause of the program and obviously the appropriate then or else.

As I stated in a prior post: Re Fan lincs. In the emulator I would set up programs for all my fans with the following names (using my office fan as an example). All my fans are controlled by fanlincs Friendly Name Say to Alexa Office Fan Low Turn on office fan low Office Fan Medium Turn on office fan medium Office Fan High Turn on office fan high Office Fan Faster Turn on office fan faster Office Fan Slower Turn on office fan slower Office Fan Turn off office fan All of the above work as expected in the ISY connected Home. I have similar entries for the fans in the Guest bedroom ("Guest fan"), the Master bedroom ("Master Fan") and the Lanai ("Lanai Fan").

Re Fan lincs. In the emulator I would set up programs for all my fans with the following names (using my office fan as an example). All y fans are controlled by fanlincs Friendly Name Say to Alexa Office Fan Low Turn on office fan low Office Fan Medium Turn on office fan medium Office Fan High Turn on office fan high Office Fan Faster Turn on office fan faster Office Fan Slower Turn on office fan slower Office Fan Turn off office fan I will be trying these exact program names tomorrow when I completely set up the Eco with the new ISY connected home skill adapter. For now I am just playing with simple on/off switches In general I assume I can cause the system to use only spoken names and not the names of devices? If so, could someone tell me how (I am lazy) as I don't want to have twice as many devices as I need to. For what I have set up (basic lights e.g. "Kitchen lights") when I speak the command "Alexa Turn On The Kitchen Lights" ( or the similar off command) I see a significant delay with the top of the Echo spinning before the action actually occurs and then an additional significant delay before the Echo responds with "Okay". The emulator would perform the desired action almost immediately after I stated what I wanted (<1 or 2 seconds) with a long delay until the top stopped spinning turned off and the Echo responded with "Okay". I assume this will improve as things progressed in the ISY Skill adapter.

At the current time (last 3-4 days) I am not hearing any of the "That command doesn't work on that device" from either of my Echo's. After reading the links to the other forums where issues are being reported, I do not have any groups defined in the Alexa app. I do have "lighting groups" but I set them up as scenes in the ISY. For example in my home there are six lights that make up the kitchen area lights: Kitchen Lights, Peninsula lights, Laundry lights, Pantry Lights, Counter Lights, and Room Divider Lights. There is an ISY scene that has all of the switches for these lights as respondents. Alexa runs the scene when I say Turn on (or Turn off) the Kitchen Area. With the capabilities of the ISY I have found no reason to use Alexa Groups. Also if you define an Alexa group, that group is deleted when you "Forget all devices", which I was doing quite a bit when developing/testing the configuration application for the HUE Emulator. My intuition/experience indicates to me that it is an Amazon cloud problem exasperated by cloud/Internet loading. The least Amazon could do was state which command and which device are involved in the spoken message. If I understand how Amazon operates there might be multiple data centers that make up the cloud, perhaps chosen by geographic area where the Alexa is installed. If this is so it will greatly affect who gets the error message and when.

I do not believe that is possible without a full skill. The skill adapter developed by Amazon is quite limited on what it can understand which is why it works as well as it does when the cloud is not stormy.

If you are using my AWS_Config program to set up the bridge, then Harmony activities and devices can be setup. You need to turn on the Harmony capability by setting specific entries in the AWS_Config ini file. The ini entries look like [Harmony Hub] Emulators = 192.168.1.81:8081 Antonyms = Up/Down, Left/Right, On/Off, Forward/Backward, ToStart/ToEnd, BackToTick/ForwardToTick, FastForward/Rewind, Play/Pause, 16:9/4:3 The antonyms are so that if you pick the first of a pair, e.g.Up, for the on command, the second, e.g. Down, will be automatically selected as the command for the off sequence. With regard to your earlier post (I know very little about the Harmony system); If the Harmony system in an activity lets you set up a macro, and one of the things in the macro can be a TCP/IP command then you can directly send a REST command to the ISY to do any lighting action you require.

Not using Netflix, but you might want to try changing the friendly name to "Netflix System". Also try "Shut" or "Shut the" as the voice command

"Run" was never a supported command for the original Alexa native skill. I suspect they are playing with changing the native skill as they begin to expand the original native skill to handle more "hubs" beyond a hub for just lights. You are right; there is no trail of what gets changed in the cloud. In my opinion Amazon is taking after Apple with an arrogance beyond belief. Amazon is a great on-line retail service provider, they are not so great in the software development/deployment arena. If I could figure out how to do it without the cloud, I would go there. I know others are working on a "sunny sky" (cloudless) solution but I do not believe the technology is there yet to support it. IMHO, the Echo is the best game in town but not yet ready for those expecting an appliance solution. In the native skill the most reliable commands (and the only ones documented AFAIK) are Turn On, Turn Off, and Dim. I generally have good recognition of Shut as in "Shut the office lights", but I almost always say Turn Off. I assume the hour you spent was with "Customer Service" and not developers. I suspect that Amazon knows exactly what is wrong, but just doesn't know how to fix it. Probably hard to reproduce from a test bed adjacent to the cloud.

I assume the response you got was "That command doesn't work on that device." I just tried my system and it is working perfectly., and it has not failed in that manner (nor any other) for the past week or so I suggest that ayone with this problem immediately report it to Amazon Help by phone call and request to have Amazon Alexa development call you back to discuss this problem. Remember the old adage "The squeaky wheel gets the oil" Lets start being squeaky.

I started with a smile, but then got really annoyed. It sounds like some person in a call center reading from a script. i especially like the sentence: . . ." I'd like to inform you that there is been no defect from our end, Please be assured of this." . . . That is not even a syntactically correct English sentence, i.e. bad grammar. I did not know that Amazon development wrote absolutely bug free code that has been fully tested and certified and works under all load conditions and all Network idiosyncrasies. My suspicion is an intermittent problem under certain high load situations. Any time I tried my system today it has worked correctly. Amazon has stated in many of its documents that they welcome D-I-Yers to experiment with the Echo and have provided special situations for them to do so. The Hue Emulator is an instance of a D-I-Y development shared to a significant community. Do not hesitate to contact Amazon and discuss this with them in detail. It can not be the emulator doing this as the emulator does not provide any verbal responses via the Echo, Only Amazon does that or a full skill is able to do that. The Echo/Emulator is driven by a skill adapter, as are the Insteaon. SmartThings, and Wink, and not a full skill such as IZZY. The Echo uses the standard skill adapter for the Philips HUE bridge. That spoken error message is coming from the Amazon Cloud. I would like to know from Amazon the exact situation that causes that message to be returned, since when the Echo is spoken to several times with the same speech fragment some times it will work and sometimes return the "That command doesn't work with that device" response. The Amazon cloud is fully capable of diagnosing a misunderstood speech fragment in a sensible way, and does so many times. Stating a "meaningless" response with no information as to what was bad with the speech fragment (e.g. what Command, what Device) is just bad design. I have never seen a card in my Alexa App account for this instance of a response. I have never received a call back from development and I have asked for it. The people I speak to in customer support really do try to help but clearly indicate that they need to get the "Development team" involved.

bfish, NO. the network module is made to communicate with a simple IP based device. The Emulator is a lot more complicated than that. It needs to be able to deal with the Echo over a secure link. The emulator does communicate with the ISY but via the ISY's Restful interface.

Benoit, I have not yet looked at the Portal. When I have a UI with a "Button Set" I generally put a frame around the button set and then put a title in the frame like "Add" which then pertains to the entire set in the frame. Users seem to like that approach.

IIRC a Hue can only handle 16 lights. With early versions of the emulator there were issues with Amazon discovering all the devices with a limit of some 40+. I have no issue on the amazon handling the emulator with up to 37 devices, I just never tried more. I have 93 devices spread out on the instances. It is not a RPi performance issue as only one instance is ever active in my home and each instance is a negligible load. So the setup is just one executable with four start commands each with different values for some of the parameters (port number, Database name, Logfile name). It is discussed a couple of pages back in the thread.

I know, I have 5 RPi's running at this time. One doing the Emulator, one doing Kodi, and the other three just hanging around. They all run headless using tightVNC to provide a remote keyboard/mouse and either Samba on the RPi or WINSCP running on my widows macine