Slashdot Mirror
OpenMotics Offers Open Source (and Open Hardware) Home Automation
Home automation is a recurring topic around here; we've had stories about X-10-based home-brewed systems, a protocol designed for automation, and more than a few Ask Slashdots. Now, an anonymous reader writes OpenMotics is an open source home automation hardware and software system that offers features like switching lights and outputs, multi-zone heating and cooling, power measurements, and automated actions. The system encompasses both open source software and hardware. For interoperability with other systems, the OpenMotics Gateway provides an API through which various actions can be executed. The project was open sourced 2 years ago and was started about 10 years. The choice to open source the project was very conscious: we want to offer a system where users are in full control over their home automation system.
thanks!
I've thought about home automation for a while, and seeing an opened system such as this one is an immediate temptation! ...Except for the cloud management.
I noticed right away that the gateway controller 'comes with a 1 year subscription'. Sure, I get that they're a business that needs to make money, but what if I want the system without cloud support? Is that even an option? For all they're touting openness, I couldn't find that obviously posted on their site. That's a pretty big deal-breaker for me, if I cannot disable their cloud integration. So what if I can't run it from a mobile phone? I'd rather use something like SSH and write my own interface, following my own desired rules for network security. And I sure as hell don't want it reporting anything back to them, or giving them the option in any way, shape, or form of sending remote commands when I have elected to not use their service. This is my home we're talking about.
Guess I'll keep thinking about rolling my own, someday.
You can't even compile the firmware without paying for a compiler.
"The firmware of our modules is written in Pic Basic Pro (except for the power measurement module which is written in C). Pic Basic Pro is an easy to use programming language that can be learnt very easily. We provide all source code free of charge (GPLv2 licence). The Pic Basic Pro compiler however is not for free so we ask everyone to play fair and to purchase the compiler (PBP Gold edition that supports the Microchip 18F) from www.melabs.com." - http://wiki.openmotics.com/ind...
With 50 euro for a power supply, 100 for a sensor conditioning module (without the sensors!), 300 for a base station and 800 for a complete starter pack, I don't care if it is open source or not, it is way out of budget for the casual hobbyist. There are already enough different alternatives, most of which appear to be vapourware. Home automation seems easy enough that many people who follow the IoT hype start their own project. But we don't need more standards, we need less. The best would be if one of the existing protocols (not necessary that one) would win, so that people could mix and match their own components, which don't have to be more fancy than some arduinos and RPis thrown together.
karma police: arrest this man, he talks in maths; he buzzes like a fridge, he's like a detuned radio. [radiohead]
How do you get past the certification problem in an open hardware design? I don't know about the rest of the world, but in Canada, you're on very shaky legal ground if you go plugging in equipment that isn't CSA certified.
Source: am building automation engineer.
To gain access to the thousands of apps being independently written for HomeKit, these devices need to support HomeKit natively or work with a hub/gateway that can translate. That doesn't mean they can't also support some other protocols (even open source protocols), just that they won't be relevant for much longer to the hundreds of millions of people who will be doing home automation around the globe without support for HomeKit. If they aren't economically relevant then there is no point in investing in their other protocols.
"OpenHAB - developed in Java" Yeah, no thanks!
Your thin skin doesn't make me a troll
... because there's a group advertising that it's launching an open-source home automation control platform/protocol/widget. And naturally this is going to be the One Protocol To Bind Them All, there will be a grand unification (about as likely as North and South Korea unifying tomorrow, by the way). The world does not lack "universal" or "open source" HA control protocols and products, it lacks manufacturers willing to support them, or even open their own control systems to support them. If there was such a thing as a truly universal protocol implemented by everybody, then manufacturers of hardware would be totally commoditized and competing solely on price/feature points. As things stand today, manufacturers profit by locking their hardware and software to work only with overpriced accessories and licensed "friend companies", and also by selling compulsory monthly service fees. Oh, and that's not even counting the fact that approximately five people (and all hipsters, at that) actually care about home automation. Home automation is, always has been, and always shall be just five to ten years before going truly mainstream.
... As OpenMotif? And wondered what an ugly old X toolkit had to do with home automation?
For home automation (or industrial control), it's not the software that's a challenge, or even the hardware schematics. It's *building* the hardware that is the hard part, especially at low cost. 30 years ago, I was dismantling X10 modules to retrofit them into track lighting cans, and I got a real appreciation for how hard it is to make good *inexpensive* controls (X10 is not good, but they were cheap, esp through DAK mailorder). Those X10 units dissipate a lot of heat, and are not tolerant of being hot: the little brown module plugged into a wall socket is isolated and has good ventilation.
And then there's the whole safety and reliability thing. A hobbyist living alone might be ok with poor grounding and isolation, and live with the occasional unexpected power on/off event, or losing control or sensor devices from transients. But I think the average software schmo with a family is going to find a lot of problems.
And you really want a wireless solution, either RF (Zigbee, for example) or carrier current (X10). Carrier current, these days, is probably no cheaper than RF, and RF doesn't have anywhere near the interference and propagation issues that carrier current does.
I'm interested in power measurement (rolled my own), so I took a look.
The ready-to-buy pricing is interesting.
A power consumption measuring module for 8 lines is £324.00; a common breaker box with a capacity of 40 circuits will require five of these, for a total of £1620.00. Then you'll need 25- or 50-ampere current sensors, 40 of them at £12.00, adding £480.00. Now, if you want remote control through their cloud, add another £295.00 for the gateway module. The power supply module is £50.00. Then you get to subscribe to their cloud solution, I think, an idea I got by the "free one year of cloud subscription" you get with a small bundle of components they sell -- though I didn't find a price for the cloud itself.
So a one-breaker panel solution seems to be about £2445.00, or at today's exchange rate, $3,843.13.
That's not horrible for what it does in terms of commercial solutions, but it certainly isn't in the low-end zone, either. You can make a calibrated current sensor for under a dollar if you dig up some surplus ferrite, which I've not found to be particularly difficult (though ferrite isn't the only workable way to go. An optically isolated op amp configured balanced over a tiny resistance also works great.) So roundly, $40 for the ferrite based solution. An op amp and an A/d channel together don't amount to a dollar per either, so another $42 for those (I use a final pair of channels to watch AC voltage and phase at the breaker box, comes in all kinds of handy. Power consumption's not just about current!) Add about $10 worth of digital logic, a $40 Raspberry Pi [there's your computer and wired web server, add $5 to put it all on wifi], roll your own software and PCB or hardwiring, throw in a tiny power supply, and for about $150 US, you've got equally capable -- or better -- measurement capabilities. If you want to be fancy and uber-safe and avoid the whole ferrite space and cost and availability issues, you can add $5/line for another $200 cost for optically isolated op amps would would put you at about $350. And of course there is no need whatsoever for a "cloud." Just a webserver, which the Pi or similar can neatly provide. The Pi is a good choice because it's low power, well supplied with features, and capable and sufficient to the task. You can toss a monitor, keyboard and mouse on there permanently too if you want a fancy at-the-breaker-box position, but you don't actually need to, so I don't count that.
I did wonder what it'd cost to build from their PCBs, but there doesn't seem to be any way to really figure that out other than doing it. Pretty much has to be less than $3840, though.
I've fallen off your lawn, and I can't get up.
It seems they're missing home security in their use case. Is it me or do half the home automation solutions seem to forget to integrate hardwired and wireless door/window security. At least with openHAB this is allowed even if a binding needs to be written to interface with the security system.