Zigbee / Matter Motion Switch | Project Linus (Blue Series)

[mbbush]

I ran into this exact problem, but fortunately the only consequence was some wasted gas and a very warm living room.

I have a gas fireplace that uses an RF remote control, that I had programmed into Bond a long time ago, was never satisfied with the result, and didn’t use. Then a few weeks ago, the actual remote was missing, and we wanted to turn it on, so I got bond out, plugged it in, and used it to turn the fireplace on. I have my bond integrated with home assistant.

The problem with the way my fireplace is configured in bond is that the remote sends the entire state of the fireplace (flame on, light off, flame height 3, back flame on, etc.) in each command, but bond thinks it’s recording individual commands (increase flame height, turn off light). I don’t actually know if this is something bond has solved in the several years since I set it up or not.

I have an automation that I run at night, when I push the config button on one of my Inovelli switches upstairs, that turns off all lights in any of the downstairs areas. The light inside the fireplace was registered as a “Light” in home assistant, in the “Living Room” area, so home assistant turned it “off”. But the “light off” command recorded by Bond was actually setting a state that included the fireplace on. The gas fireplace ran all night, and fortunately didn’t start anything on fire outside of the fireplace.

In the morning I unplugged Bond and put it back in a drawer.

[Rob]

Did you receive the test units yet? I’m anxious to see how they work!

[caduck]

FYI. My gas fireplace I bought this, Skytech 8001-TX Smart Home Compatible Fireplace Remote Control Transmitter Only

It is so simple, yet so perfect. I absolutely love it. Such a great solution.

[Rob]

Hey Eric, checking to see if there are any updates? I also have a question. Since the white series 2-1 should be coming in around the same time, and these switches contain the mg24, will there be a thread version of the mmWave soon (or at least a firmware update to convert)? I bought some of the blue series mmwave a while back on indiegogo and have pre-ordered a bunch of thread 2-1’s, so it would be nice for them to all be on thread network instead of splitting.

[Cortexian]

Very excited to have found this upcoming product! I’ll be moving to my first owned home (rather than rented) soon and very excited to upgrade the switching to Inovelli products. These mmWave switches will be perfect for several automations I already use with inferior WiFi switches.

[lnjustin]

@Eric_Inovelli would you be able to update the status for this product? Last I can tell, back in November you were expecting test units for beta testing? How’s that going? Looking forward to this! Thanks!

[Eric_Inovelli]

Yeah happy to give an update!

We received the pre-beta units a week or so ago and the goal of them is to lock down the look of them (hardware) so we can then open up tooling to get all the beta units sent out.

Here are the samples:

Z-Wave
We decided to go with the, “Glow in the Dark” color (kidding, idk why they chose that)

Zigbee

For the most part, the hardware was great. There are a few things that needed to be changed which I’ll outline below:

1. The height of the configuration button needs to match our current switch – right now, the depth of it is about half of what it should be

• The height should be inline with the top of the paddle height at the top
  

  

• Even Scott’s updated picture of the paddle doesn’t seem to reflect the updated feedback
  

  

2. The config button also appears to be rounded vs ours is more square – same as the LED Bar diffusor is rounded off (particularly at the bottom)
   

   
   
   The picture of ours looks a little rounded, but in reality it’s squared off, please refer to the samples we sent you

3. LED Diffuser shows a more pixelated view of the LED’s, whereas ours is smooth
   

   

4. The Ground Screw terminal is partially blocked by the metal plate that holds the ground screw into place making it impossible to insert a wire into the ground terminal
   

   

5. Screw plate slots are too big for the screw plates which is causing it to feel loose and it doesn’t get as much tension to spin the screw and will ultimately cause breakdown of the plastic since the edges are digging into the plastic (whereas with the current switch, there is no wiggle room which allows for more tension to turn the screw)
   

   

6. The screw plates are mismatched compared to our current ones (ours is on the right and yours is on the left). Our original supplier had this same issue (with rectangular plates) and they fixed it by making it square (the rectangle one caused stability issues). Please update to the square plate as per our samples given to you.
   

The biggest point of contention is the lux sensor. As you can see in the pic above, it’s basically a hollowed out section. To us, it looks unfinished and we’re afraid that when someone gets this switch, they’ll think the same thing. We asked if we could put some sort of cover over it that matches the LED Bar, but the problem then becomes that the lux reading would be inaccurate. In addition, it would cause complexity with production as it’s such a small part.

What we’ve considered is to either make the LED Bar longer so it’s just one piece, but that ended up looking a little strange and also they had concerns over the light from the LED bar causing inaccurate readings.

What we’ve settled on are a couple of options which are shown below:

We’re leaning towards the option on the left, but we haven’t heard if the option on the right is feasible yet. It’s more of a pin-hole size so it looks like it blends in more.

–

We should be wrapping up all the hardware changes this week and then I’ll have a more accurate timeline on when beta units will be in production.

The good news is the firmware has mostly been built so we didn’t really lose any time for building the firmware.

My best guess at this point is an April/May production, but I’ll have more info next week when the timeline is updated.

Hope that helps and I’m happy to answer any questions!

[lnjustin]

@Eric_Inovelli

We’re leaning towards the option on the left, but we haven’t heard if the option on the right is feasible yet. It’s more of a pin-hole size so it looks like it blends in more.

Great, nice to see you’re putting so much thought into important details! For what it’s worth, I like the option on the right too, if it would be feasible, since it does seem to blend more and would provide a larger target for a button that is getting more functionality with the fan module. I wonder though what having the lux sensor there would do to readings upon button press when a person’s finger/hand goes in front of the sensor - maybe in the typical case it wouldn’t be impacted by a quick button press…I’m sure you’ve thought about that though!

[MasterDevwi]

Thanks for the update! I love how many details you guys tend to share.

FWIW, I really like the LUX sensor option on the right—assuming it’s feasible—since the gap around the hole seems very big on the left. I also agree with @lnjustin about how the bigger button would be easier to press. But both options are better than the gap the current sample has.

I’ve been spending a lot of time getting into mmWave sensors over the last few months. The two most popular ones seem to be the Tuya ZY-M100 and Aquara FP2:

• The ZY-M100 would be great… if it worked. I’ve seen a lot of people rave about them, but I’ve purchased half a dozen and couldn’t get any of them to pair well with ZHA. I’ve also heard more than a couple can wreak havoc on your Zigbee network.
• The Aquara FP2 does an amazing job of telling you where in the room the user is, it’s a bit slower than I’d like and very bulky. And I typically don’t need that degree of presence detection.

I’m so excited about Project Linus because it should hopefully be the perfect combination of discreteness and accuracy. I don’t expect it to be able to know exactly where in the room the user is (although that would be nice), and thankfully a binary occupied/unoccupied sensor is sufficient in most cases. Hence why I preordered 18 of these.

When you get a chance, I had a few questions about the final mmWave sensor:

1. Can we set a maximum range for presence detection?
2. Can we block off areas of the periphery in case the sensor faces another room? I seem to recall discussion about allowing the sensor to ignore specific quadrants in a grid?
3. I believe you settled on 24GHz, is that correct? I was thrilled to hear that the FOV would be 150° too.
4. In testing, have you seen any issues when the sensor is facing a device that could cause interference like a computer screen or a fan? Hopefully #1 and #2 above can help solve any issues with interference at the edges of the room.
5. How fast has the sensor been responding to presence so far? I’m sure this may continue to improve, but I’m just curious.

[Dunlapjc]

Some thoughts on the lux sensor in the button vs. below.

1. Lux sensor In the config button, especially in that pinhole size may create some long-term repeatability/reliability issues with dirt on a button touch surface slowly collecting enough fingerprint grime to block or inhibit the lux readings.

2. Lux sensor below the config button. Probably cost-prohibitive, but the covers could be two-shot or over molded to give you the indicator diffuser and a cover over the lux sensor without the need for handling assemblying small optical parts.

Or this would need to be tested/vetted, but you could possibly mold a single-piece indicator cover diffuser / lux sensor cover with a small back-frame between the indicator diffuser and the lux sensor cover. Keeping the diffusor portion from acting like a light pipe and eliminating total internal reflection (TIR) with some design bends would keep it from acting like a light pipe and interfering with the lux sensor readings.

-Jonathan

[mharris1984]

I really hope Inovelli put a design patent in on the use of MMwave inside a switch, to prevent other companies from suddenly running with this, or start seeing knockoff’s on the market. Honestly, if Inovelli doesnt have design patents on the LED bar and the customizations that can be done, I think they really need to jump on that to prevent other manufactures from running with your ideas (again). I think thats the best way they can position themselves in this space… If suddenly Leviton puts a design patent in for having a MMwave sensor under in the switch to detect human/animal/motion/breathing/etc… then Inovelli would be forced to stop producing them.

Also, I wonder if the firmware could be setup to read spacific data off the MMwave sensor. Like ammount of objects (moving/breathing) detected (up to like 7 or something like that?), distance for each object, mass of each object, etc… This way you can setup automations based on mmwave data for differnet things… i.e. kids mass is xx and setup automations for just them when entering room, like turn on TV and set to Netflix and chose their fav show. If mass is xx and = dog, turn on just a night light low to give them some light to walk through to the doggy door at night. or if everyone is gone and a full grown mass is detected, send alert of possible intruder, etc… I can’t wait for this release. It is truely a first of its kind switch and is a HUGE breakthru in switch automation… Thank you @Eric_Inovelli

[pyrodex]

I love this project and thank you for all the hard work!

I’d like to explore a few of these for bathroom lights and also usage to toggle on smart switches controlling exhaust fans, via another switch of course, and can’t wait.

One question I did have though was around limiting the 150 degree field. I was planning to replace a very old dumb PIR light switch I have in our closet and I had to “hack” the old PIR sensor by putting a piece of electrical tape over nearly half of it to cut down the range of the of the sensor. The closet is a long rectangle with the door/entry way on a long side of the rectangle and the light switch is on the shorter wall near the door way. With the closet door open it would toggle on when someone walked past it and the “hack” helped me prevent that. So is it going to be possible via software settings, I use Zigbee2MQTT today, to limit the field of detection to prevent such situations?

Thanks!

[Eric_Inovelli]

Hey thanks for the comment – we do have a patent attorney we’ve been chatting with, but it’s a bit more complicated. Design patents are easily circumvented by slight changes and to put a patent on mmWave in general is a tough one because I’ve already announced the features of this product more than a year ago so it’s public knowledge.

However, one of the things we rely on is our firmware which is very hard to replicate as it would take a ton of man-hours and cost and even then, they don’t have the community or the beta testers that we do that continuously help us improve the product.

Finally, even if the best case scenario happens where we did get a patent, trying to drag out a legal battle with the big companies would be tough for us as we’re a small company. Even the random Chinese companies that tried to copy us are difficult to fight as they just close up shop and try under a different name over and over and over.

Sucks, but it’s the world we live in unfortunately.

[Eric_Inovelli]

Project Update:

We are still working on the hardware around the lux sensor and I think we’ve finally gotten to a place where we can move forward with tooling. We should be receiving the final version next week for approval with all the changes we requested.

Simultaneously, they will be releasing the first round of firmware so we can get a head start on testing it while they produce the beta units, so things should really start picking up here (thank God).

Here is the latest (and hopefully final) look at the switch:

Here it is mocked up next to the Blue 2-1:

[jsb]

That’s looking VERY nice. Cannot wait to get my hands on some of these.

[Eric_Inovelli]

Thanks! Me too! Really looking forward to testing the mmWave feature. As of now, the manufacturer has been having all the fun testing out the mmWave sensor…

[MasterDevwi]

Love it! I’m so excited for this switch.

Also on my dream wishlist? A mmWave power outlet. It could be ideal for adding a discrete mmWave sensor in locations where a light switch isn’t available.

[ibrewster]

Same here. I have two locations that could really benefit from something like this. Though I might prefer a z-wave version…

[sdholden]

Simultaneous development. Z-Wave 800 Series Motion Switch | Project Linus (Red Series) - #64 by Eric_Inovelli

[ddeflyer]

I’m not sure if the height of most outlets would be optimal for mmwave but the number of outlets in a room might brute force a good presence detection!