Why Home Assistant is the FUTURE of Smart Homes
ฝัง
- เผยแพร่เมื่อ 28 พ.ค. 2024
- Home Assistant is the FUTURE! Let me explain...
Thanks to Paulus Schoutsen for letting me interview him at the Z-Wave summit!
For more Home Assistant info check out th-cam.com/users/liveoa__fLAr...
NVIDIA Jetson: developer.nvidia.com/embedded...
NVIDIA Jetson Orin Developer Kit: amzn.to/3ws6YqV
Jetson AI Lab: www.jetson-ai-lab.com/
My smart home favorites: www.amazon.com/shop/smarthome...
My smart home shirts!
USA: smarthomesolver.big3creative....
International: smart-home-solver-shop.fourth...
FOLLOW ME
Subscribe to Smart Home Solver: bit.ly/3joO1dZ
Subscribe to Reed's Smart Home: bit.ly/3EulO02
Twitter: / smarthomesolver
Instagram: / smarthomesolver
TikTok: / smarthomesolver
0:00 Intro
1:26 Home Assistant is growing
2:29 Getting easier
5:19 HA + AI?!
9:01 Z-Wave is not dead
10:19 HA is the future.
10:56 Number 1 feature?
---------
MUSIC
Digifunk by DivKid
Ending song is The End by EVA.
open.spotify.com/artist/6UIs5... - วิทยาศาสตร์และเทคโนโลยี
Do you agree Home Assistant is the future? Let me know what you think👇
no one cares
Absolutely, with so many different standards Home Assistant has already been the best solution for marrying multiple things together. Just need to get stubborn companies that lock down IoT devices from being open source or unwilling to provide any sort of API support to live in the year 2024 and realize people don't want to be locked into an individual application.
I think that eventually, Home Assistant will end up having to have a light version of itself with all the features they're adding to it. That way it can still be ran on a Raspberry Pi for the none advanced users. This is why I've recently migrated my home assistant over to a slim OptiPlex computer with an i7. That way I've feature proofed my integration, and it can run on a much more reliable NVME SSD!
Will, No one cares about whether you care or not, eather@@surprisednow
But it is so difficult to use…
I've been using Home Assistant for some time, and it continues to evolve, improve, and impress with every new version. I was initially hooked when I turned it on and it immediately began detecting smart items (from different manufacturers) on my network without user input.
Home Assistant is the future as long as they don't sell to Google.
this is Gold
They can't. That's the point of the Open Home Foundation.
I hope not Nvidia either
I also like the idea of being able to work on a local AI box integrated with Home Assistant to keep all of the data local.
Spouse approval is HUGE
Had no idea home assistant was created by a fellow Dutchie! That accent is a dead giveaway 😆
Python was as well, so it really is a Dutch product through and through 😁
Yay Netherlands! 🇳🇱 12 points for Home Assistant 😉
.
_Hallo mede-Nederlanders_ 👋 :)
Yes I am all in on Home Assistant. Needed improvements: assist voice listening accuracy.
Making automations easier... I know this will never happen, but personally I wish they include Node-Red by default and work on tighter integration, ditching their own automation workflow. Like maybe have a simplified front-end but it creates Node-Red flows behind-the-scenes that you can still access through the Node-Red interface to get all its glorious versatility.
I love Node-Red too, but this would be more a "tinkerer/maker" option rather than main stream. The Home Assistant automations can do virtually anything Node Red can do now, just in a different way. Node Red is a third party solution also. I don't think it will ever go away, but I don;t think it is the way forward for a main stream system either.
Just my opinion :)
@@EsotericArctos I agree Node red is easier to use but it's very much for power user.
Not sure if this gonna get mentioned (wanted to note it whilst I remembered) one thing I would love to see is the ability in dashboard creation is to be able to 1) state screen size (so maybe 8" phone/tab, 10", 12" etc) 2) lock colomn count so I want to always display 4 or 5 columns and have columns resize to fit the screen size rather than have colum jump to a new row 3) have cards and text in card scale to card size (okay this would be a pain if you want a title that is massively long but you just have to think of different name lol)
You can actually specify the maximum columns in the new dashboard
If they're taking suggestions, extending their first party radio USB sticks to network devices would be a nice feature. It is often the case that people run their HA host with their network stuff in some corner of the house, and having the ability to move the Zigbee/ZWave/etc. radios to a more optimal location is great. I have one now from TubesZB, really just a Zigbee radio attached to a microcontroller running ESPHome, and it works great, but I think first party offerings would help for newer people.
There's a whole bunch of Ethernet ZigBee coordinators out there
Private local AI box was just amazing
I’ve been in the technology business for nearly 40 years. I like Home Assistant because it reminds me of the early days when smart guys would figure out something and then be happy to share it, the days before Bill Gates decided to entrap users and charge for everything, the days before every vendor made his product proprietary because he was obviously smarter than everybody else. There are countless brilliant people who prefer to work independently, outside the constraints of big business, and Home Assistant lets them shine.
I’m excited about the potential of local AI, shareable dashboards, and appreciate the emphasis on making a smart home that works for everyone in the home. Thanks for this preview of what’s to come!
I’m using AI in my home to control things, but only if I ask it to.
The beauty is being able to say to your voice assistant “hey I wish it was brighter in here” or “I’m feeling kinda down, can you adjust the lights accordingly?” and it goes and does what it thinks is best. Obviously “turn on the living room light” is straight to the point and it does that too.
I would absolutely want a local AI box in my home. I'm working on DIY'ing one right now with some leftover stuff, and donated hardware just for giggles. It had old Intel and NVIDIA gear but I think it'll at least be something. I would go for $2k, but only if I could condense everything I'm running elsewhere into a single box - NAS, NVR, containers, etc. which I need to look into these NVIDIA boxes to see what they can do vs. using my current PC desktop when I do my next upgrade (i.e. desktop class). Yeah, it'll take up way more power, but might be 1/2 the cost. I have plenty of room, and while I'd like to save power, that'd be a long ROI to justify. Interesting stuff, and thanks as always for the video, Reed!
I actually agree with you, I’ve been looking at the Orion AGX, but I’m actually comparing it to like a Miniforums MS01 with an RTX A4000 SFF and 64GB ram which has a total cost of about 3k in the end but has capabilities to install promox, which I don’t think the AGX is locked into the running os.
A very important roadmap! I agree about dashboards... Sharing them should be easier!
I'd like to see a set of "Home Automation Kits"... Where those of us who are newbies, could pick from starter kits thst included everything needed (equipment and installation) to setup the Automated home. The kits could include basic, intermediate, and advanced kits. Meaning a basic kit might be like.. "so you have a few Wyze Cams, sensors, ect.. Here is a basic kit to get started with Home Assistant, how to setup the automations in Home Assistant". Then move up from there.
And some video instructions on setting up different options would be fantastic. Like individual smaller instructional videos for each type of automation, what it does, how it's implemented, ways to use it... Etc.
One feature I like to see is a online UI based card creator. where you can easily create cards within cards. trough a set of template/grid designs.
Similar to how the sections works but then on CARD level [minecraft principle], where you have a large grid and are able to put in cards/blocks like lego, and design them based on what you want, with sliders above to change the PX distance between cards and it giving a error if the card is bigger than the default.
I have seen many card-in-card things I like, but have no idea on how to properly set it up. or change things like instead of the template with 3 cards in a card, I want 2. removing one breaks everything.
or a card with a dropdown menu. that automatically closes. things like this are things I want, but have no idea on how to add
This tech does sound great, if not for my low income and my learning curve. After I had strokes, leaning has been a challenge. One thing I wish I could set up is all 60 smart lights in my home to be able to set dimming specific to time of day. I too started my smart home journey seven years ago. I now have 112 smart home devices, one main router and four access points (19 rooms - 3 floors). I use a different smart home assistant which has limitations, but works for my needs. Being disabled, the two things I am glad to have is the ability to reset Alexa, Google Home, and all cameras via app or voice, and everything on battery backup.
My first Mr. Coffee I installed a Sonoff Mini R2 inside, which lasted three years. Now I use TP-Link Kasa 6 outlet smart WiFi power strip.
Favorite routine - "Alexa, good morning" turns on Mr. Coffee, turns off dining room light, turns off kitchen stove vent light, turns off porch light, raises the east kitchen blinds, turns on bed light, turns off foot warmer (heating pad under my pillowtop), sets thermostat to awake mode and tells me the weather for the day.
Mass integrations over multiple protocols, Drag and drop automation workflows ala homey and really good voice affects that can integrate with Alexa for shopping related tasks will kill every other platform
Glad to see some API integrations to leverage existing local LLM capabilities. I've been training a local LLM on my Home Assistant database for almost a year now and it's very useful to ask the local model to explain some events/incidents and/or to troubleshoot some "non-working" scenarios/automations.
To quote your buddy Paul Hibbert, “it just works!” 😅
Really? Mine doesn't...
th-cam.com/video/FnGUw1XK5hI/w-d-xo.html
'It just works' is reserved for products he is paid to review.
Unfortunately, it doesn’t always “just work”. I’m a techie through and through (professionally in both software development and electronics design and implementation for 4 decades), and my straightforward implementation of HA just doesn’t work sometimes, for no apparent reason. I have a dedicated NUC with a Conbee II. For example, my Zigbee light switches work 95% of the time, but sometimes some of them just don’t. One minute they work, the next they don’t. Rebooting HA usually gets them working again. 95% reliability is pretty good, but not good enough.
@@Innesb It was actually just a joke. If you watch Hibbert Home Tech, it’s part of his schtick.
I’m just starting my Home Assistant journey but even though the learning curve is steep the benefits are clear and the effort and thought going into it’s continued development is inspiring. Keep up the great work everyone!!! 😃
i just noticed that you have lost a quite a bit since i started subscribing, you looking good my man, keep doing great
Great video! To answer your question about local AI: yes, I could envision having a local AI "box" in the network closet. We tech people are used to paying a premium to have local hardware instead of a subscription (perfect example NAS vs cloud storage). As with any tech, the more mainstream it becomes, the cheaper it should get. Exciting things to think about for the future of the smart home!
HA has progressed tremendously over the past 2-3 years. It's great how it works out of the box, however dashboards are still a weak point and require a lot of configuration. I have 2 HA instances which both work well, but only have basic dashboards because I don't have the time to really get into them. Having a focus on this vs AI would be my vote.
I finally got into Home Assistant because of Home Assistant Green, which came out late last year. This is exactly what they need to grow their market. It has to be plug and play, no buying your own PC or assembling different parts and then installing the OS yourself. If you make it very easy to get into the world, people will stick around once they try it out and learn more about it. Especially when they learn that it's not hard at all to use. I'm just doing the basics and it's more than good enough for me and will be for the vast majority of users.
Nice to hear that Z-Wave is not dead. Big glitch last year after an update to ZWaveJS but all good now and rock solid on series 900. I still have Insteon running on ISY, but that technology died because they kept proprietary too long (at least that's my theory).
I've been using Home Assistant since 2016 nd I have seen it come such a long way. Home Assistant today is like a completely different system to what it was back then, yet in some ways you csn still see the roots. Home Assistant did what no other did. Completely local, no cloud needed. This is what set it apart for me.
In many ways, even as it is right now, Home Assistant is easier to use than most other platforms on the market, and the way so many devices can integrate with Home Assistant just makes it ideal as that one hub fits all platform, even though it is much more than just a hub.
Thanks for sharing Reed. Yes, Home Assistant is the future as it was in the past. 😊
Very good stuff! after going going thru the MyQ debacle I'm really only interested in local only or local capable devices. i don't need my data constantly leaving my house. Im currently working thru the Odin Project to learn front end and backend to be able to do more of my own developing.
Fantastic video. Thanks for brining the information from the summit here!
Been on HA since 2017. I looked at two other system, but chose HA and never had any regrets in my choice. HA and local control rules.
I've been using home assistant for almost 3 years now and have seen it change a LOT. It is a lot easier to use and it's functionality has grown. I currently make my own sensors with MCU's and sensor modules along with ESPHome add-on for home assistant. I agree with you that this is the future of the smart home.
I'm not the first to say this by any means, but if I were on the Home Assistant product team, I would be working on onboarding/migration tools. In other words, how to make it easier to switch to HA. I have to believe one of, if not the biggest obstacles to adoption is customers that don't want to rebuild their set up in a new environment. If I could just copy over my set up from any other platform, I would switch.
I'm just looking into all of this, so forgive me if this is a newbie/idiot question/statement - What I'd like is to integrate my smart home (which to be honest I'm initially looking mostly to replace my existing home alarm system) with my video surveillance system. Where I see AI is with an integrated system that handles both my smart home and AI, and using that AI to identify people, capture license plates that go past my house, etc., from the video feeds so the home security system can act on it. AI could also quickly, after a few days or weeks determine which areas in the video feed to ignore because of random movement (trees moving in the wind, cars driving on a street on the other side of one of my neighbors, drops off of the gutter, etc.) My existing video surveillance system (Lorex) sucks so incredibly badly at this. A system that integrates both smart home (alarm) and video security, that uses AI would be amazing. Though, for now, I'd just like to get my home automation system to tell my video security system to turn on interior cameras when it arms when nobody is in the house (at this point in time I'm hoping to link that state to neither my or my wife's cell phone is in the house).
Just set up my home assistant server up this morning! 🙌
47 seconds in and gas prices go up when Reed comes into town. look at the Citco sign
Wow! Your eye for detail! It's true!
I already liked HA, been using it for a couple years now, but after this interview with Paulus it made me like it even more!
That's great to hear! Paulus is such a nice guy and we hung out for hours after this. It was so much fun. He is the real deal!
I like the idea of an “AI box” and have kinda already been experimenting with it using Ollama on my MacBook (which never leaves the house so it might as well be a desktop) so the idea of being able to further integrate it with home assistant would be cool.
I installed HA on an old laptop about 6 months ago more or less to mess around. While the install wasn't Installshield easy, it wasn't super horrible. However, what really surprised me was how easy it was to get things up and running. Sure, there are a couple devices I can't use, but most integrations are pretty seamless. Automations just make sense, and are even better with some recent updates. No where near as scary as a lot of people make it out to be.
I like the idea of using local AI in this instance. and buying an AI box could be cheaper than building it on your own
Loving the videos Reed! Was super surprised when I saw your networking setup and it was nearly identical to mine. Definitely made me feel more validated that I'm doing it well lol. Please keep your amazing automation ideas coming. Can't wait to see more :D
Thanks for the nice comment! Yes more automation idea videos soon.
I have Homey and HA - Automation in HA .. not so friendly as Advanced Flows in Homey. Yes I can add Node Red, but "Advanced Flow/Node Red" should be core way to do automation in HA and not an HACS add on.
Home assistant is likely the best platform to end up on. Starting on a Google/Alexa ecosystem is usually more attractive as it's built in to everything we buy these days. These automations are a lot easier now than even 6 months ago bridging the gap between the "starter" platforms and home assistant.
I feel like I’m in the rare camp where I don’t actually like HA… yet. I purchased the HA Green and everything functioned. But the UI was pretty horrendous and the items I was attempting to integrate had issues. The amount of time I needed to spend to get these attractive dashboards and automations to work, just weren’t worth the effort needed. It feels like the product is for die hard only at this point. Until that changes, not yet for me. I just don’t have the time to invest in getting things working for my needs.
If Home Assistant had easy to setup automations like Homey Pro's Flows, I would be back on Home Assistant. Having flows, and having them just work, and intuitive, is what is keeping me from running Home Assistant.
Hi, quick question, what do you prefer and easy to use? Home assistant or Homey Pro?
@@anilkumar1971 Home Assistant has more options with more devices, but Homey Pro definitely is WAAAAAAAAAAY easier to setup automations with their "flows." And it isn't even close.
@@TheRiehlThing42 Thank you so much! Just curious: what are the main "flows"?
@@anilkumar1971 Flows are basically how Homey Pro does automations. It is drag and drop. If Home Assistant started using them, I would probably be using Home Assistant.
An "edge AI" that runs locally close to your devices, we think it's a great thing when, for example, you have about 120 sensors in your local network and you can't monitor them all, or react when you have a problem or urgency and need to shut down the systems. At first I think these AIs don't have to be so powerful, but well suited to concurrent monitoring and basic automation. Imagine you have several municipal swimming pools and a small child leaves you an unforeseen gift, with an AI starting all the "urgent" debugging processes should be able to simplify the process.
Honestly, I am not too worried about a 600 usd AI box.
Especially if you have a larger home with more complex automations, it will be worth it.
It only becomes expensive when you buy it unnecessarily for a 1 bedroom apartment without a garden to speak off.
Just! Ordered a home assistant green a few days ago. Really looking forward to it
Sweet!
The only AI I'll consider using is a local-only one. I like the idea of having a local, central AI that hooks into things like jellyfin to make suggestions or create playlists. I don't necessarily want that to be baked into jellyfin because I'd prefer to have a centralized "brain" that can be used by any media application or similar and not have one per-application that's tied to only that application. I'm excited for the improvements to the UI and dashboards. I've been using HA for probably 7 or 8 years and my dashboards are a hellscape of "unavailable" things that I can't be bothered to fix. It would be great if something would solve that for me without me having to spend a bunch of time on it. I feel like at some point I'll just give in and do a fresh install since I think a lot of my problems are very old problems.
Also - shared dashboards and dashboard blueprints would be amazing. Especially if there was something in there that checked whether devices exist for each category and only show tabs and panels where devices are present instead of creating everything in the blueprint when only 75% of it is applicable to your instance. Example: don't show a garage tab at all for people who don't have a garage.
Local, exactly. I don't need google knowing how many times I passed by a thermostat or other sensor to go to the bathroom each night. Because you know your data will be sold.
Incredible!! 😮 I can’t wait to see what has been revealed
Unless integrators or home builders start selling it, it will still be niche. Matter integrated with smart phones and micro hubs would be the likely thing that would be spur adoption more than home assistant at this stage.
Tbh, even though I play with tech full-time for my job, even I have steered away from Home Assistant due to it's lack of easability, setup, etc. I look forward to seeing what they do in the future
I think HA has huge potential. I've been using it myself for the past year but spouse approval is a key factor here. In my journey I have been conscious to always retain manual control ability for all smart devices. I primarily use Zigbee and also have a small number of cloud-reliant devices where I couldn't find zigbee alternatives.I don't really have any automations setup so far, I mainly use it as a dashboard for home, being able to pull up an app and control any device in the house is very useful, I'll add more to it in the future but for now I'm content. My next big jump will be to add voice assistants. I've been tied to Google Home for too long and it just hasn't kept up...at all, if anything it is far less useful now than it was at launch
I do however want to see improvements around HA updates, it seems every time I open the dashboard there's an update available. I want updates to be automatic, and to only notify me if there is a failure (it would be cool to see a auto-rollback in event of failure too).
Home Assistant is just way too good when it comes to all the options out there with how modular it is to pick anything else and the fact that almost everything can be controled without the need for cloud api calls that don't need to happen just to turn a switch on or off is perfect.
I have been using it for about a year now in a virtualbox instance on my nuc so I can keep rotating snapshots and backup the entire VM folder when I make major changes. I use the trigger by list with the choose nesting to organise the actions for the devices that I need to control which lets me keep things nice and clean for the amount of automations I have in the list.
The skyconnect zigbee stick in multipan mode has been working great and don't have issues so far, so looking forward to grab the zwave stick when its ready from them to support.
I would really like to create automations in the same way as Homey. Just select a device and say what you want to do with it. So turn it on or off or give it a different colour. Now that's still sometimes tricky.
With regard to using or sharing one's data to train an "artificial intelligence" so that it improves, I wouldn't be too worried about issues related to privacy. This data can be anonymized and pooled together, and I can't imagine too many people would be interested in what I turn on or off and at what time. If no one is willing to share their information, then it's going to take a lot longer to properly train these systems to improve them for everyone.
I've barley scratched the surface of HA, and I see so much potential. I'll have to work on making a nice dashboard.
This is extremely exciting can wait to see what happens next , big yes for blue prints 🙌🏻
From 150 to 600 is quite steep just to add local AI. Let's wait for the first results here on youtube! By the way, amazing that the guys come all the way to Austin from Europe and after the conference they continue doing actual hardware testing in the hotel. Commitment!
Depending on the AI task you want to run they can totally work on a Pi. An LLM, no, but lots of classification, regression, and optimization models work great on small embedded platforms.
Tell him to make a remote control /hub combo to make it a fully universal remote for controlling your whole home without the need for pulling out your phone. That might help make it more mainstream and compete with control4, urc, creston etc.. and it will help generate revenue from selling a product!
I caught the home automation bug when I build my house in 98 and installed a Stargate w/ lots of X-10, which I'm slowly moving away from. I've gone down the HomeSeer path before Home Assistant was created, as HS had a Stargate plugin. I've since upgraded to HS4, but now they're starting to charge an annual fee for some connectivity. Because I abhor paying fees, I've now started researching Home Assistant again as I've had my eye on it for years now but just not made the jump. I've a fair investment into Z-Wave equipment, so this video really piqued my interest, as I don't really want to replace stuff with Zigbee.
A home-based, upgradable AI makes perfect sense especially if it can integrate across multiple platforms e.g. your smart home, computers/network and home entertainment. Information you don’t always want to share with the world!
Question - recently installed HA on a thin client and have it connected to a dedicated network I set up for this. Is my HA device supposed to be on this network or my main network that my computer is connected to? How do I access it to set up dashboards and what not since my thin client is sitting in a closet with no monitor. I was able to access via the URL prior to moving to a dedicated network.
i already had a linux system always-on in my basement so i run home assistant through docker on that. are those usb connectivity devices compatible with that setup? i have an ezlo 100 plugged into the wall to control my somfy zwave shades that i'm looking to replace since its connectivity is bad. was planning to find a smartthings hub that still has zwave support but since i plan to put it in home assistant anyway that might be simpler . . . though no mention of when that might be available. i'll see what i can find on the home assistant site.
i was using Zwave and HA as my primary setup for smart home 9 years ago. HA has really come a long way. i ditched zwave for wifi and zigbee because it was just too unaffordable and less choices. I hope zwave will have more devices and lower their cost.
I don't know anything about HA other than what you've presented over time but I do think, generally speaking, the "best" platform is usually not the one that wins. It's the best marketed platform and HA will need a major corporate backer to achieve mainstream adoption.
I’ve been on the fence with HA for a while running it in Docker, but I think I’ll be ordering a Yellow and CM5 once they are available.
It's definitely got an excellent core operating system and libraries. Though I still think there are some major pain-points in the architecture which will need a bit of a rethink.
I struggle with the way that Helpers, Templates, Automations, and Actions are stored separately, all in a flat global scope, often defined using different syntaxes (YAML vs Jinja), and impossible to encapsulate together into reusable modules.
Modelling a network of devices is one application where an Object Oriented paradigm is ideal. But Home Assistant seems to have both an Imperative system, and a Functional Reactive system, but which don't always integrate neatly.
I have never used this before and I’m very basic in my home automation and “smart“ integration, but I can see how things like this would be beneficial for someone like me to “upgrade“ a smart home without having to go down the rabbit hole…or am I totally off on this? 🤔
Hi,
Great information:)
I'm currently building a house, and it's in the rough construction stage. What minimal work should I do to prepare the building for smart solutions? Does every device need access to grid power? How can I optimally plan the layout of electrical cables in the rooms?
I would suggest some central location for WiFi router and home automation computer for fanning out radio signals to all parts of the house.
Typically these pieces of equipment end up maybe in a basement or attic, so the spread of the signal isn't optimal for coverage.
Cabling (Ethernet, video, power) isn't as important as it was before the advent of wireless technologies.
Now most devices can be button battery powered, which is kind of a pain, but beats having to deal with powering stuff.
Some devices are now using PoE, so you may want to think about that.
Otherwise, for certain devices having an AC outlet nearby what you think you might want to automate might be a good idea.
@@indiggioman thank you. I think about installation in all rooms Aqara FP2 sensor. After that I will be able to create a lot of automation using this sensor.
I don't use voice activated services with Home Assistant because I don't want cloud services to have my voice date- any more than they already do.
Buying a Jetson or Xavier is a small price to pay to get voice services running locally.
Hell yes I want Home Assistant with a local AI accelerator.
Just came here to say you're looking great! Not sure if you've been doing some fitness regime or some diet but whatever it is, it's working! Keep going!
LOVE Home assistant! You can pretty much do anything your heart desires. I'm concerned though...and I'm sure others are too..how long will this last as open source; free? I'm worried this will change and break a lot of hearts...
Great video per usual Reed. Home Assistant is definitely the future and I agree that Zigbee and Z-wave aren't going anywhere anytime soon!
How did you make the "Room Select" dropdown on your dashboard? (3:18)
Being able to offload the AI work to desktop graphics card would be the best solution for me. I'm already hosting it on my desktop box with anything LLM used to access it from other computers.
Time to save for a Nvidia Jerson! $2000 is expensive! But if it can accelerate local TTS, STT, LLM, object detection! I am excited!
If home assistant is the future. There is alot of work to make the learning curve more accessible to ordinary people
HA is great but it’s nowhere near being “let my elderly mom buy it and use it”. Long long way away. They’re in rapid development still with nearly zero regression testing and it shows. The UI also doesn’t consistently follow a design document. It’ll be a while - assuming they get a higher level of discipline.
HA has certainly come a LONG WAY in ease of use. I looked at it in 2020 as a possible replacement for SmartThings but it was waaay too geeky and complex. Everything then involved editing TEXT (YAML) files. No Z-wave or Zigbee. Nope. Fast forward to recent months. I am on Hubitat and love it ... BUT ... there are things that do not integrate with Hubitat. So out came the old Raspberry PI and I reloaded lastest HA. WHAT? This is different -new - easy - changed! Intuitive even. I have been using it more and more I'm not ready to move away from Hubitat but it it now part of my Home Automation arsenal. I am very impressed by their recent announcements and am following the development of HA closely.
I've tried HA twice, both times have given up after hitting problems early on. Sitting here right now looking at sonoff temp sensors that I'm using in dumb mode because when i add them they add just fine, but they stop communicating within minutes. 😢 No clues as to why they stop working.
The day Home Assistant doesn't ever expect you to use yaml, I'll consider it for our family home. :( So much of the coolest stuff in HA is locked behind that and I cannot begin to imagine troubleshooting a yaml error for my folks. Likewise, the fast updates can also mean breaking changes, which is not cool for consumers. So, YAML and breaking changes are my biggest concerns atm. Until then, HA can't be "simplest / easiest to use" home automation platform.
I used homeassistant appliance on my VMware, it filled the disks with logs and now it has been broken for 6 months and I do not want to reinstall it since it will probably happen again. Logrotation is a thing and if you make an appliance please have your stuff in order.
im on habitat now I did it because it seemed easier and more stable than home assistant between updates. I want to change to HA for two reasons. Graphing data without a ton of hacks and setting up a separate SQL server. and I want to get Alexa out of my home. I want a locally ran voice assistant that can do intercom, announcements, make shopping lists, timers, of course can control smart home things but streaming music and telling me the weather is important to me. I can live without it answering questions but I would like that ability too. if a local voice assistant that did these thing well, worked with HA fairly easily, Id make the switch today.
Yeah, they definitely need to overhaul Lovelace dashboard when it comes to theming. More settings (like card transparency or tap to open a stack that pops up a room or entity grouping) via UI rather than editing the theme.yaml or getting specialty entities/cards from HACS. Also? Significantly easier way to change wallpapers via UI too!
This is very exciting. Cannot wait to see what the future brings. And yes I would use an AI box in my home
I have Govee and Tuya lights and devices and an Ecobee thermostat. Would everything be private and local if I switched to Home Assistant?
I really want to get onboard with HA, but I can't walk away from the voice control offered by my Alexa devices.
The only time I want to look at a dashboard is when configuring devices.
Question as I am looking for a new smart home system for my new house build. I have been told home assistant or homey pro. Thoughts as you’re a big smart home guy. Just want to use my HomePods still.
HA , is a huge improvement of our life. Not a joking.
I would like to start using Home Assistance, but I already have several smart products with their own apps, can they be used anyway or you need products specifically
When will HA come to cars? So, you can monitor and protect your car the way we monitor and protect our home?
If it’s truly local and private, I’ll consider it.
If I could have an AI box in home, that could run my lights/music etc. without needing internet access it would be heaven. I'm not sure that is what is being suggested however.
Yep, I'd put in an AI box. For now I'm happy to wait for homeassistant on a normal GPU to bring enough over the table to be worth getting one of these for the homeserver.
I really liked home assistant, but its UX is pretty atrocious. I switched to HomeBridge, not because it's any less technical or complex in the backend but because only I ever need to see that. All it does is expose devices to HomeKit, and bridge to Apple Home which is what the rest of my family then interact with.
I would have loved to replace Apple Home with Home assistant, because it's so much more powerful and customizable, but it's a mess. The moment I showed the UI to my family, they almost gave up on the idea of home automation entirely. If interacting with it it is more difficult than standing up to use a physical switch then you've already lost a lot of people. Apple Home does this really well, in my opinion. Automations are largely straightforward and streamlined, the layout is logical and simple, and it's just a very clean UX for the end user.
For HomeAssistant to truly become the future, there needs to be a much clearer line of delineation between the front end, which people interact with, and the backend where things are set up.
Definitely some interesting things here, I'll definitely be keeping tabs on it!
Can you please review tapo devices? How is their protocol? How is their integration to home assistant?