Hi everyone
I am currently working on my PhD project looking at the feasibility of using a connected home to support independent living for adults with autism. I have been using Insteon to connect various sensors, with various amount of success, and I was wondering what people thought of Z-Wave as a possible alternanative?
I am looking at using room ocupancy monitors and window and door sensors to map how people use the house and use the rules capability of openHAB to create a monitoring environment that reacts to the users and provides a level of safety that can provide some feeling of confidence to the care network that the person is being ‘looked after’ to a certain extent.
I have zwave devices setup with OH2 on a Pi3. I am using the AEON GEN5 usb stick controller and ge(iris) zwave light switch, dimmer switch and wall outlet. Attaching devices to the controller is very easy and they are reliable and consistent. Openhab has a pretty extensive library for the zwave devices and great community support.
The one issue I have with zwave is the cost. As the protocol is proprietary, the devices tend to be more expensive than devices that use open protocols or wired solutions.
Outside of the zwave items, I have only used a few wifi based devices. These require more effort to setup and manage, and would not recommend as a reliable solution based upon my limited experience. They also require connected power, so that is a limiting factor that needs to be considered.
Another possible option would be to go for an alarm system based solution that has binding/connectivity options in Openhab. You would get the benefit of complete system solution with an automation tie in.
I believe your professional interest is on the application layer and you don’t want to spend too much work on the technology below.
Generally speaking, assuming that all of your control happens inside openHAB items and rules, you can combine any sort of sensor as long as the technology is supported in openHAB (Insteon, ZWave, others) .
So unless you require specific sensor capabilities (such as say IR, a very fast reaction, or a very long detection range), simply choose the cheapest available openHAB-supported sensor. Just don’t select too many different technologies because each (well, most) of them requires a physical gateway, resulting in additional cost (usually +/- 50-60 US$ or €).
That being said, ZWave is a good choice in many places. It’s reliable (better than Insteon), and there’s a broad range of devices available. While not really open, ZWave isn’t as proprietary as Insteon is. There’s at least a number of different vendors, competing on price.
For door/window sensor, have a look at the MAX! series of eQ3. While meant as an extension for thermostats (to get informed about open windows), you can use them as general purpose sensors for 20€ each.
Motion sensors - well. If you were able to deploy centralized wiring, you can get away with a central Pi or Arduino, but that’s quite a lot of DIY work in wiring and software. A ZWave battery-powered motion detector (+light +temperature) is 40-50 US$ or € (but don’t go with the Aeon multisensor on battery power, they suck battery like hell. If you want to use them, use mains power via microUSB).
I also use a WiFi Foscam camera (~60 €) with builtin movement detection. Might be useful for your application, too.
I recommend you PM @watou if he doesn’t chime in here. I know he uses HA to keep tabs on his elderly parent and is in my opinion one of the greatest home automation use cases around. He might have some good bits of advice in general.
I have zwave, mainly controllers (e.g. outlets, light switches) and a few smoke alarms. Ignoring the discussion on price:
It is very reliable in my experience, though sometimes something may go wrong and when it does you will need to physically go to the device and reinclude it into the network.
It does have a limited range (around 40 ft, 12 m between nodes) and a limited number hops (4 hops between device and controller) so if the space is large you will need repeaters or more devices to get good coverage.
Configuration of zwave devices in OH can be arcane and mysterious but there is lots of help on this forum.
If your controller dies you are likely looking at needing to readd each device to the new controller unless you happen to have one of the few that supports backup and restore.
But, as @mstormi has said, you don’t need to limit yourself to a single protocol. One of OH’s main jobs is to provide a layer of abstraction between the technology of the devices and the rules and logic of your automation.
Also, as @mstormi has said, wired will almost always be more reliable than battery powered wireless.
Thanks for your answers.
I agree about the cost @mstormi. The devices do seem quite expensive.
Thanks for the pointer to @watou@rlkoshak, I will pm him for some ideas. The range should not be a problem in my test environment, it is a 1960’s detatched UK house with reasonable internal walls. I am looking at using MySensors.org as another solution. I like the way OH can use many different communication protocols, this is probably the main reason I have chosen it for my PhD project; I didn’t want to be locked into one solution as the idea is to design a system that can have almost anything added to it, depending on the needs of the user. In my test bed, wired is not really a way forward, the Uni won’t let me rip stuff out and install new cabling…
