Connect multiple wireless mics to control from different rooms

project_voice

(Michael) #1

Hello guys,

Before i go any further i want to say that i have decent programming experience and not afraid of writing some code if i have to, to implement this. At the same time i don’t need this to become a full time job)))

I have this idea where i would place multiple microphones, one in each room and depending on which mic picks up the voice command, a different action would occur. (assume we are using ST hub). Also each room would have a speaker and it would be more of a conversation.

For instance i say “lights” and only lights in the room where i’m standing turn on.

I was curious if anyone tried something similar and have the following questions:

  1. Would u use any existing home automation software that would be compatible with SmartThings? if so which one?
  2. I’m trying to go wireless, any particular mic+speaker set in mind? I think bluetooth may not work for this application.
  3. Is it possible to integrate this way with Siri Cortana or other ready solutions or may be use some APIs? if so any pointers would be appreciated.

#2

I think most people are just using Amazon echo for voice control. It works great with smartthings, and it has amazing far field voice recognition hardware, but it doesn’t give you the context information (which room you’re in) that you’re interested in. “Lights” would always be the same thing unless you put a different Amazon account on each Amazon Dot.

There are two other ways that people have done it. The easiest is probably just using a combination of Tasker and sharp tools on any android device. Then each device would have a meaning for “lights” which was specific to that room. So you would need one android device in each room, but they’re pretty cheap. @joshua_lyon can say more about that.

There’s also a community member who built himself a whole house system using multiple hardware and software pieces.


(Joshua Lyon (SharpTools.io Dashboard)) #3

Let me know if you want any additional information on the Tasker/Android approach. It takes advantage of the standard “OK Google” wake word used on modern Android devices and enables you to control your house with your voice.

Since it’s a custom setup, you can also setup custom profiles for each device or even for a single device based on context (eg. Home vs work, depending on which room you are in, etc).


(Eric Wright) #4

I think people constantly misunderstand my pi in all this. I’m trying to build the system with the cheapest possible solutions. I’m not really working on this project at the moment, but I really don’t like the Amazon Echo. I’d rather use a system that’s free and isn’t locked down by a mega-corporation. So is nice for those people who really don’t want control over their environment. They’re comfortable spending a lot of money to solve a small problem. Most people just don’t have the money to throw at stiff like that. So yeah, Echo might might work well for some of your userbase, but for the actual community who wants a smart home, let’s face it, they looked at smartthings as an affordable solution because they couldn’t afford anything more expensive. Sorry for the small rant, Amazon Echo just isn’t for me.


(Michael) #5

Thank you for all responses. Shawn I agree with you that Echo is not the best option. it is more of a generalized mainstream oriented product which doesn’t scale for me…
Is it a hub? no… can i use it as a central brain power of my home automation? no… can i connect multiple mics to this thing to set around house ? no…

It is too bulky to carry around. So to me it is a large speaker with microphone that can do a subset of things that my cell phone already can. And anything extra that i may do - i don’t care about that functionality :wink:

Joshua, i’d like to learn more about multiple profiles for a single device. What is the trigger to switch context? Is there a guide? After reading some about your app i think it can do most of what i need with an exception of connecting multiple mics. I read there is an ability to pair with multiple bluetooth devices: https://groups.google.com/forum/#!topic/android-developers/adeBD275u30 In my opinion ability to identify location by getting a uuid of a device that sends you a signal would be a big step forward. Especially if it could select a device with stronger signal when 2 mics pick up signal at the same time.

Thank you


(Chris Means) #6

I doubt you’ll be able to beat the cost of the Amazon Echo. I am a bit of a fan-boy for the Echo, so I am clearly biased, so feel free to ignore everything here.

  1. You can build a Raspberry Pi unit that has a mic & speaker…you can also connect it to Alexa (the Amazon Echo voice service), if you want. Amazon has even “endorsed” directions on how to build one.

  2. I doubt you can find a better price for a unit with the distant-field mics that come with the Echo & Dot, but I’m also thinking about it’s professional finish. Maybe you can get the hardware together cheaper but I doubt it would look “quite” as nice. The Dot is a lot smaller than the Echo, and you can pair it with a BT speaker to augment it’s weaker internal speaker.

  3. It will be a lot more work (and maybe that’s the challenge you’re really looking for) to roll your own solution. I completely understand that. I personally have a chess set with a single missing piece…but rather than just buy a replacement piece (or even a whole new set, which would only cost $19), I wanted to be able to replace that piece by 3D printing it…I’ve spent a lot more time and money on getting the 3D printer, scanning etc. going than the $19 to replace the whole set…however, I’ve learned a ton, enjoyed the experience, and can now do a lot more things than I could before.


#7

a small problem

A small problem for one person can be a large problem for someone else. Different things work for different people.

I myself am quadriparetic with limited hand control. So voice options solve a very big problem for me. :wink:

I have been using voice technology of one kind or another since I first started using a wheelchair. There is huge variation in the quality of the voice recognition hardware, let alone the software.

The Amazon echo is amazing for its far field technology. It can recognize commands from across the room while the television is on. And it has natural language processing. I’m never seen anything like it in a system that cost less then $20,000. There is some serious engineering in that device. :rocket:

That doesn’t mean I think everyone should get it. Many people don’t want to be dependent on a cloud service, other people want to try to spend less money, some people just get satisfaction out of building their own solutions. And, as mentioned, to get context awareness with echo requires a lot more money.

But don’t put down echo as a potential solution for some people. “My use case is not your use case.” There are real hardware advantages to the echo solution, they just aren’t advantages that everyone will need.

So as always, even for voice solutions, “all home automation is local.” There’s no one right answer, because different people have different requirements. Choice is good. :sunglasses: