After thinking about home automation on an on-and-off basis for several years, I have concluded that the only thing that's needed is automation of electrical sockets. Being able to flip a physical switch and/or check its current state from an iPhone covers good 90% of automation needs. The 9.9% is a thermostat, a temperature sensor and a flood sensor for basement. The remaining 0.1% is exotic stuff to measure the magnetic field flow and a wind speed in a bathroom.
And I am convinced that the first company to produce a no-nonsense WiFi-enabled drop-in replacements for conventional wall switches, junction boxes and sockets will wipe the competition clean.
Ahem... now, back to your regular programming of raising $1M off geeks who get all excited about a random sensor having its own Twitter feed :)
I agree completely ... it's not really interesting when it just goes to twitter. That's why it's designed to talk REST, EventedAPI, and MQTT to other software. The real value begins to grow when more complex rules systems can be applied to larger numbers of sensors, and then integrated with sensor data from your mobile phone, etc.
You leave home ... the security is enabled ... home environment reflects you being "away".
It's when all of this begins to reflect a more "neural" pattern that things get really interesting!
Really interesting in a geeky kind of way (to which I can relate completely), but practically speaking though - home automation has been around for a long time, and it didn't really go farther than thermostats, motion sensors + lights and security systems (that still need a hook up to the monitoring services to be useful). I remember tinkering with X.10 almost 10 years ago. Nothing much has changed, it's still engineers designing for engineers. Admittedly I haven't closely looked at the scene for a couple of years, but I researched the subject back then and remember seeing no home automation system that was designed with a notion of usability in mind. Lots of options - sure, plenty of buttons - you bet, perhaps a hierarchical menu with the depth of Cameron's dive - oh, that's a deluxe version :) It seemed that whoever designed those systems didn't actually use any of them.
(Not to hijack the thread)
There shouldn't be a remote, an existing iPhone should suffice. It should know which room I am in, and it should have relevant options based on that. Default to the electrical/heat/blinds and show all outlets and sockets. Allow to switch to other rooms, ideally in form of a floor plan. Make the remote work when I am not at home.
Sell this as a kit of sockets, pass-through junctions, wall switches and a small form-factor controller box. Put it all on WiFi, and give me an iPhone app to control the box. Make it easy to pair the box with devices (show a list of available devices and let me claim them). Allow hooking up the box to the Internet service for out-of-house access (charge me for that too, that's a fair deal).
Finally, put an effort into designing the UI and all visible physical parts. Not to be Nest-slick, but designed nonetheless. [1] is an example of what is not designed, but engineered. Feel the difference. Go to Europe to get inspired, if needs be.
Basically build a system only if you actually have a personal need for it, and not because you have a handful of assorted sensors. And "they will come" :)
I recently discovered this http://www.pwrusb.com/powerUSB-digitaI0.html and apparently it supports standalone operation but is considerably more expensive--although it also offers more controllable sockets. I've not played with one yet but it looks like a possibility. (Apparently also supports 240V operation with connector adapters...)
In the past I've rewired the remote control of RF-controlled sockets but the signal's not very reliable.
Getting an SMS if one of my doors opened during the day might be nice.
Drop in on light switches would be huge... But how do you manage drop in? If the consumer has to actually replace the switch with a "smart" switch, that's DOA (requires wiring, messing with the breaker box, etc).
Replacing a switch is one way, a pass through on the socket is another--you plug the smart device into the socket and your actual stuff into the smart device. That won't hit all devices (not everything is plugged into a socket), but would be a good way to get the more casual user involved.
If I got an SMS that my front door was opened in the middle of the day, I'd call my wife to see if she went home early. If she didn't, I'd call my neighbors and/or the cops.
My point on light switches is that (unlike outlets, which could just have a pass-through unit) they require someone actually un-attach and re-attach wires, which many/most home owners aren't super comfortable with. Even if they are, it's a meaningful time investment in comparison.
I think you're right about home automation. There is also potential for home analytics/intelligence as well. For example, water sub-meters on each water faucet could help conscientious water users conserve.
Couldn't you calculate this simply by monitoring the "run time" of the devices using the gas? If we can tell how long the furnace and/or water heater run, then you can do the math vs. the gas bill?
Now that's a good idea! I could tell whether our central heating boiler is running by its electricity consumption (which will either be on/off) and measure what it actually consumes when it is on.
Kind of obvious once you mention it - as all the best ideas are!
I've thought about it quite a bit, it wasn't off the top of my head; I would also like some sort of usage monitoring, but I'm not eager to tinker with gas lines or the insides of a furnace.
How are these little startups dealing with the FCC regulation of consumer products that are classified "intentional radiators"? I'd really like to get back into this so this is a question I really would like to find an answer to.
The last time I looked into adding wireless to a device we were building, I was advised that any little nit that burps a frequency above 9Khz intentionally should expect an 8 week $30,000 trip through independent testing and regulation purgatory to get an FCC ID # and permission to sell to the general public. Not doing so and selling anyway was serious federal crime territory.
Granted this was years ago, but nowadays it seems that every other technical kickstarter is a whole bag full of different (which would be different models) "ninja boxes"(1) that tweet wireless data. How are they clearing (or sidestepping) this regulatory hurdle?
A module from Microchip that looks relatively easy for someone who already monkeys with micros to add low speed wireless to a project for about $8/unit. Very Nice.
I think it was more like a month, their distributor/manufactures did it for them and most especially, pis are not intentional radiators. They left off wifi partly for this very reason.
I'm sure I've seen 5 or 6 of essentially this same thing on kickstarter over the last couple weeks. It seems to be a popular idea, but until the price drops enough for them to be disposable I don't really see the point.
That's what I thought of as well. Twine's, though not as robust, seems a more simple and elegant design. The question I'd like to know is who will take these systems and implement them in applications that have a greater "wow" factor for the consumer. Nest is a great example.
There are so many applications, especially when one considers location sensitivity relative to a phone.
I actually work as a programmer for home automation gear. You'd be surprised the extent to which (non-geeky) people want to control their homes. You'd also be surprised at the crazy number of automation startups.
This one looks pretty interesting, but so few actually go anywhere at all.
"We also have our board designs complete, and have already been through several initial board revisions. Once funded we'll submit our boards to production..."
They claim to have some board designs of their own (I guess in addition to the off the shelf sensors). They should emphasize what they're bringing to the table that's new and exciting and not spend quite so much time describing the hardware you can already buy elsewhere.
We do have our own boards in our gateways. Yes, in our initial platform release we are in OEM partnerships with several companies - one of them Monnit for the sensor modules.
Although you can go purchase the Monnit sensors themselves, our IP is in the multiprotocol sensor/control gateway/router technology, and eliminating the need for software developers to understand the lower levels of hardware operation.
I love tech as much as the next person but I think the Internet of Things is just going too far. I've read tech news on a daily basis for years and years and I find it all really interesting, to a point, and that point is coming.
I don't like GPS, ebooks, smartphones or even video games, though I love a good Apple laptop and I love keeping up with the news. The fact that I even stopped learning my friend's phone numbers by heart was the tipping point. A seemingly small change, sure, but it reflects a wider tendency. We are slowly replacing our memory with computer memory. There have been articles, books and many comments on this effect but I get the feeling that 'new and shiny' is too attractive to most and I worry about there being no choice in the near future. I hope the losing side, in the end, is not the same side that, from the start, yearned for everything to be 'smart'.
How are these sensors different from these ones on this site: http://www.monnit.com/products/ I think is the same amount of money and I can get them faster.
Monnit is our first OEM partner on the sensors, but our core focus in the sensor/control integration aspect. If you want to do your own low-level development for these specific sensors for a vertical solution, you could. We're providing a much broader solution and doing a lot of the software implementation work for you.
As for the ship date, Kickstarter is working hard to ensure that we give dates (that we are pushing to beat) that are realistic and do not slip like some projects have.
This is exactly what we are doing with Wovyn. The issue is that there are numerous standards being used, and we're implementing a wide range of protocols to support the key platforms people are wanting to integrate with. This is exactly the space our IP exists in.
After thinking about home automation on an on-and-off basis for several years, I have concluded that the only thing that's needed is automation of electrical sockets. Being able to flip a physical switch and/or check its current state from an iPhone covers good 90% of automation needs. The 9.9% is a thermostat, a temperature sensor and a flood sensor for basement. The remaining 0.1% is exotic stuff to measure the magnetic field flow and a wind speed in a bathroom.
And I am convinced that the first company to produce a no-nonsense WiFi-enabled drop-in replacements for conventional wall switches, junction boxes and sockets will wipe the competition clean.
Ahem... now, back to your regular programming of raising $1M off geeks who get all excited about a random sensor having its own Twitter feed :)