After living with the Samsung SmartThings Wi-Fi hub functionality for a few weeks, it’s time to move on. Well, sort of. I like how Samsung’s product is running my smart home currently, so I’m not abandoning it.
Instead, I’ll be looking at a more DIY approach on a limited basis because it’s been a long time since I’ve done that. And by “DIY”, I mean installing open source software on a Raspberry Pi, then configuring and maintaining it myself, which is very different from a retail solution such as SmartThings, Wink, or Hubitat and the like.
There are a few options in this space but I’ll be going with Home Assistant, which, if its forums are any indication, has a fairly robust community. And this isn’t my first rodeo when it comes to implementing a very hands-on smart home system. In 2010, I installed a headless server (read: a server without a monitor that could only be accessed through a network connection) and set my smart home up with Insteon products. I remember the amount of effort and technical knowledge it took back then, so I’m curious about the more modern experience of Home Assistant.
But I’m also coming into this project with several questions because the landscape today is far different than it was back in 2010.
Today we see more products gravitating towards near-ubiquitous radio and protocol solutions such as Wi-Fi and Bluetooth. Apple’s HomeKit platform has standardized on those two, for example. And some lighting products have moved away from using Zigbee, which requires a supporting hub or bridge device, turning instead to Wi-Fi or Bluetooth.
Take the C by GE Wi-Fi bulbs and switches or Philips Hue choosing Bluetooth for its next-generation bulbs as two recent examples. These choices are likely due to the proliferating of voice assistants that can easily support devices built with these radios.
That’s very different from a decade ago when X10 and other proprietary protocols were used. So I’m wondering if a DIY smart home is really even worth it for most people when a simpler solution may just be to buy a retail hub, Google Home or Amazon Echo product and move on.
One could easily argue that Home Assistant or a similar open source product gives you more control over your smart home, keeps your data local and supports a wider range of products. And there’s merit to each of these advantages. But another question I have is: How much longer will that be the case?
Google, Amazon, and others are already working on local control for smart home devices, so I suspect in the not too distant future, that advantage begins to diminish in a DIY smart home.
Hubitat already keeps your data in house, for example; there’s no cloud component so your data stays with you and your smart home continues to function if there’s an internet outage.
In terms of supported products, when you have open source contributors that are passionate about the smart home, you often find some new code or script that will let your new smart bulb, switch or lock work.
But I haven’t been pondering the purchase of a new smart product that didn’t seem to have SmartThings support. And when I have in the past, I was able to find someone in the SmartThings community that built a script in the Groovy language used by SmartThings for custom device handler software. So, at least when it comes to Samsung, there’s a robust community there.
This test of Home Assistant over the coming weeks should be eye-opening. Not just for the actual installation, configuration, and use, but also for what key benefits it brings to my smart home. My hope is to get a better feel for the DIY approach in 2020, sharing my thoughts along the way.
So while I dust off my trusty old Raspberry Pi, I’m curious to hear from current DIY smart home folks. Is the experience still better for your smart home as compared to the retail, off-the-shelf products or have those plug-and-play solutions caught up in your opinion?