Sorry, this is probably the one place where I completely disagree with you. Amazon's current version of a smart home is voice commands for every little thing. To me, that makes it dumber than a 50 cent light switch on the wall. My approach has always been heavy on the automation side though.
2 Likes
Have you tried out the recent iterations of routines, they are by no means perfect but they are getting there, and IMO are even exceeding the capabilities of IFTTT. The biggest feature as of late is the ability to trigger ANY command you would use with your voice in the routine, for instance, at my secondary home which is much less smart, I have an alarm system that hooks into Alarm.com, but they only have voice functions. So in my goodnight routine, I typed out the voice command to arm my alarm and it arms my house among other things.
They also support motion sensors and contact sensors. Yes there is a way to go, but it is a good system for a basic setup IMO. The key features missing are virtual devices, branching logic, and general organization.
1 Like
Voice Control in the marketplace is a very "hot" item at the moment.
As soon as I explain to a client that Alexa/Google/Apple can control all their lights, automations, devices, etc. they want it.
It's the most requested item, because people who have the resources want simplicity.
For a user of Home Automation devices, there is nothing simpler than voice control.
In my humble opinion, people are willing to overlook the loss of privacy for the control that voice enables. It's perhaps the #1 requested feature.
On a personal note.
I installed Alexa, Google Home, and an IPad in my home to try things out. We operate here primarily on the basis of automations, not even a dashboard. As soon as my wife found out that she could control stuff with her voice on the Alexa, she was "all in". It's her favourite feature, bar none.
2 Likes
Yeah, as I see it, even after becoming Home Automation focused versus Home Control focused, Alexa, or really any voice assistant, is super good at being a catch all for devices you can't find, do not know if you need, or do not want to spend money on right now. Yes you can replace every voice command with another sensor or button or switch, but voice assistants are great for one, showing off 
, or two, as a halfway measure until you know what you want.
But see, this is my point. IF (humongous if), you were able to tie all these things together with a single standard, maybe we wouldn't have to wait for someone to create Alexa skills, Google deep links, Apple magics to control all the things (voice or not). A consortium like this is the first step towards that.
The skills are just for the WiFi/Cloud side of the house though, AFAIK the zigbee side of the house works like any other with Alexa. I don't think a perfectly unified system will ever happen, all we can hope for is a unified central hub that each of these services talk to and can talk to each other through, which is what I think Zigbee and Z-Wave not to mention local WiFi provide.
Which brings up another point, what does CHIP do that Tasmota or Shelly devices don't already do with base WiFi? First thing that comes to mind is battery life, but iirc the shelly motion sensors have years of battery life and they are WiFi.
It is just so crazy to me that they are putting this much work into a new protocol to seemingly solve nothing. I would think FAQ #1 would be Why? and it would list out all of the current problems with the current systems and what/how they plan to remediate these issues. The FAQ page for CHIP just screams red flags to me.
Have you tried to tasmotize a light bulb recently? I've heard that it's getting harder and harder to find devices with the requisite hardware, with companies trying to save money as well as force their cloud adoption. I have been eyeing some shelly devices out of sheer curiosity though.
How about this? If nothing else, I'd love see the tags for "Works with <Alexa, Google Assistant, Homekit>" go away completely and just get replaced by "CHiP certified." Still a big dream, but smaller than the original ask. haha
Relevant side note: Battling a Leviton Z-Wave Plus switch that can't seem to make up it's mind about being on or not. ARGHHHH
This hobby ain't for the faint of heart. I'll probably just give in and replace with another Caseta switch. $$$
4 Likes
No I have not but I have heard much the same. For instance Feit products from costco, used to be a great entry into cheap local smart home, but they have since transitioned away from ESP8266 chips, thus breaking tasmota. If you ask me, this is super shitty and super anti-consumer if it was an anti-tasmota move, but I doubt we will ever know for certain.
Shelly looks pretty good as of late, it is not cheap, but it seems solid and everyone I have talked to that has them has liked them. I just have no need as I have Zigbee and Z-Wave. But for a new setup, it 100% would be in contention.
I think this is where we are going, but at that point it's just a third "universal" protocol, so like again I ask "why?". But I get what you are saying. Conspiracy theorist inside me thinks that this is purely a gateway drug towards phasing out the other "broken" protocols just due to development times and "new thing shiny".
I have learned as of late that issues like this are sometimes as related to the devices as they are to the host controller, in this case I assume HE, not just the protocol. Who knew that implementation matters as much as the base spec? 
My situation as well. Although, I probably wouldn't go the full wifi route until I'm fully Wifi 6 converted (which I haven't even started). It definitely seems more IoT friendly.
But we digress
1 Like
Lutron too
https://www.reddit.com/r/Lutron/comments/lebp5q/new_ra2_select_processor_with_clear_connect_type/
There's obviously a lot yet to know. CHIP is the "language" that is being spoken, which I believe is based on development work done by the Zigbee group and Google. Initially CHIP will be spoken on Thread, WiFi, and Bluetooth networks. Zigbee is not being sunset, but if Thread is successful then who knows what its future is?
One of the big things about Thread is the no-hub requirement. Devices can talk to each other. Nest Protects have used early Thread technology for their interconnect features for years. According to the project, response times are faster than Zigbee. Here's Nanoleaf's explanation.
https://nanoleaf.me/en-US/integration-hub/thread/
I think that is part of the issue, how does anyone set out and start to make a new language before they even know what the issues are with the current languages, how would they make sure they do not walk into the same pitfalls? And if they do know, which I suspect they do, why are they not being more transparent about it? I think they are trying to hedge their bets so that if they do not solve anything, they have some cover against claims of DOA by pleading ignorance which imo, is never a good defense.
And these kind of questions are not particularly complex, I feel like "Why?" is basically the first question any well-grounded project should answer before doing anything, and if they cannot answer that simple question then what is their project built on?
I feel like that is much more marketing than engineering. If "no hub" is true, then what holds the commands or core automations for the devices on the network? If all of them store it, that seems very inefficient, if the cloud stores it, then the cloud is the hub, if a single device stores it, then that single device is basically a hub. I see a slight distinction, but do not really see any benefit. Often times, "There is more than one way to skin a cat" but that does not mean that everyone should switch to the new method just because, especially without explanation.
CHIP will help a lot for hobbiyst/maker of device like me. I am not interested to build hubs, platform or ecosystem. I make devices/sensors which talk to hubs.
Today, with nrf52840, I can support Zigbee, Homekit, Open Weave with 3 sets of firmware. Although I do not need to design 3 different boards, I will need to test the 3 firmware independently. For small maker, this is a huge burden even the possibility is there.
CHIP help me one step further. I will need to build and test one CHIP firmware for my device. I would be able to make my device available for Hubitat, HA, ST, Echo, Google Home, and Apple home kit natively.
For small maker like me, I am now can plan to make 50 devices in a batch vs 10 devices knowing that my devices can run in more platform. The cost to buy a single nrf52840 module at quantity of 10 is about $10 each. At quantity of 50, one module of nrf52840 could be somewhere around $7. That has not count the saving for supporting components to make the final products such as power management IC, passive component such as inductor, capacitor and resistor. If you extrapolate the quantity for medium or larger makers, the cost saving should be huge. This will translate cheaper devices/sensor to be used with our hubs. I hope this is good enough benefit for us.
3 Likes
That is fair and understandable and I love that you give a unique perspective to this conversation, but would this outcome of simplifying your workflow and making the product be cheaper be also achieved by simply having Google Home and Apple Homekit natively support Zigbee? I get there is a benefit to CHIP in a vacuum, just like Zigbee or Z-Wave are great in a vacuum, and I do not think I have never stated the contrary.
My issue lies with the benefits of CHIP in the context of the currently operating protocols, namely Zigbee and Z-Wave. It seems to me that Google and Apple have fabricated a problem that only they are uniquely suited to solving, and IMO they are throwing the baby out with the bath water while they are at it.
The market will dictate the need and speed of implementation - not the companies creating CHIP. Just because they "make it" doesn't mean the customers will come.
Both Google and Apple are historic for abandoning 'good ideas' as soon as they lose interest. Their involvement is more of a negative than a positive to me. So many people have been burned by those two that many now 'wait and see' before spending engineering time and money.
That said, if the market embraces CHIP then other vendors will likely add support to their products. But I don't expect widespread 3rd party adoption any time soon.
2 Likes
That is my worry, as of late Apple has been a sort of trendsetter, just look at the headphone jack or notches in phones. I just worry that the market will try it out because Apple is Apple and it will gain a foothold based on that, not based on its merits, which lowers the requirement of success from actually being good to just not being bad. And then at that point, we have even more a segmented market than we have now.
Also, IIRC with the latest announcement for Inovelli's Zigbee switches, they said that they plan to support CHIP straight off the line.
Dunno. I'm not worried about it though.
If CHIP takes over the market and yields crappy devices I'll make my own, stay with Zwave/zigbee for the next 10 years (which equals infinity in technology), or do something else entirely.
Zigbee and Zwave are definitely not perfect. I have so many cheap/poorly made Zigbee devices that only 'kind of' work it is ridiculous.
So maybe there is room for another player. Dunno.
And I'll leave this here. Not sure I agree 100%, but what else are they supposed to say?
1 Like
You too? I only bought the Leviton Z-Wave switches because they came with the light almond paddles for no extra charge. Should have just paid the extra money in the first place, but $55 for a light switch is painful!
Yeah, I mean, I'm 97% on Caseta anyways, with a few Leviton stragglers in a basement and in lesser used applications. The plan is to switch them all out eventually, but sometimes things like this push the timeline. The real problem with replacing them with non-Z-wave hardware is wondering what it does to the remaining mesh... oh well! We'll see what happens!
Once I got them all up-to-date with firmware, they were fairly solid. But they were never Lutron level reliable.
