Slashdot Mirror
Building an Open Source Nest
An anonymous reader writes "Google's recent acquisition of Nest, the maker of smart thermostats and smoke detectors, has sparked concerns of future plans for the devices, and how Google's omnipresent thirst for information will affect them. Thus, a team of engineers at Spark sat down and roughed out a prototype for an open source version of Nest. It looks surprisingly good for such a short development cycle, and they've posted their code on Github. The article has a number of short videos illustrating the technology they used, and how they used it. Quoting: 'All in, we spent about $70 on components to put this together (including $39 for the Spark Core); the wood and acrylic were free. We started working at 10am and finished at 3am, with 3.5 engineers involved (one went to bed early), and the only work we did in advance was order the electronic components. We're not saying that you can build a $3.2 billion company in a day. But we are saying that you can build a $3.2 billion company, and it's easier now than it's ever been before.'"
What is the Nest. Do they mean like a natural nest build by bees or what - it is not clear form the summary, is it just me who doesn't find Nest an obvious thing like Apache or Linux that doesn't need introduction?
OK, if this is the case, why in 2010 when we built an open source Android robot, the folks at Google literally told us that since they were trying to do the same thing, they would try to pretend we didn't exist? (They failed: It is hard to pretend something doesn't exist when it's humiliating you at Maker Faire, or making your hand bleed)
Liberty - Security - Laziness - Pick any two.
The hard part isn't building a smart thermostat. The hard part is finding somebody simultaneously dumb enough and rich enough to pay $3.2 billion for a thermostat company.
I actually only heard about Nest about a month ago and was VERY interested until I found out it was cloud based. I immediately typed a complaint to them about it. I'm very happy Google is heading this way with it. Even if Googles open solution is still cloud based, it should open API's and communication documentation so that people like me who are NOT interested in giving control of my house to a cloud app under someone elses control that can sometimes override by proxy. I'm a security concious guy and I simply do not want my homes firewall open to anyone but me. My phone or tablet should connect DIRECTLY to my in home equipment or server without anyone else having to be involved.
Digital is, by definition, imperfect. Analog is the way to go.
But we are saying that you can build a $3.2 billion company, and it's easier now than it's ever been before.
Were it not for patents...
If Pandora's box is destined to be opened, *I* want to be the one to open it.
"We're not saying that you can build a $3.2 billion company in a day. But we are saying that you can build a $3.2 billion company, and it's easier now than it's ever been before."
Doing the technical development is not the issue. It's the damn lawyers and patent trolls that will soak up millions and years...
Am I missing something here? Where's the link to the article referenced by Mr. AC.
This reminds me of an undergrad electronics project ~20 years ago. I suppose it was maybe 16 hours of work spread over a couple of weeks, and that was with no domain background. There's nothing new or difficult here - it's like those "build an AM radio in a minute" videos on Youtube, which demonstrate routine knowledge and good manual dexterity, but not groundbreaking achievement. I couldn't see a practical application then, and I don't see one now.
But the world's moved on, people are a lot stupider, and Google's a data-mining company who manage to sell any old nonsense to anyone (pro-tip: just because I read about expensive cars last week, it doesn't mean that I have $200k to drop on a car regularly over the next few weeks). So, I can see them convincing people to buy this crap.
You trust Google over Nest?
hahahahaha
It's pretty easy to build a version of most things once there's a working example in front of you - the real value is doing it first, not just copying.
We met in the backyard, where he took a rest from hauling a crate of beer to his hole in the ash tree. The woodpecker said "Hey dude, you are rambling on about open source and FOSS all the time. Could you get me an open source nest, by any chance ? Us woodpeckers are rather into the proprietary model, we all have our own beak. But mine has signs of wear, and the price of new ones is too high." "Sure", I replied, "Apache Nest might be something for you. Or otherwise, check out jNest on github." Last I heard about him, he was founding a business - together with some raven from the neighbourhood - for 3d-printing spare beaks.
Religous speak to God. Insane are spoken to by God. When all shut up, one can finally hear Shostakovich in peace
The hard part is getting Lennox, Traine, or Carrier to build heaters and or air conditioners that can be controlled by USB or ethernet. That is what I want. None of this smart refrigerator crap, or Nest smart thermostat stuff.
Throughout the years I have seen instances of precisely this kind of arrogance in various forms.
Everything always "seems easy" at first glance on the surface. This is more often than not a reflection of gaps in ones understanding or failure to consider the problem space with sufficient detail.
The other major issue is failure to understand the sometimes monumental difference between building something that "works for me" vs "works for everyone".
Anyone can hack together an arduino that flips a relay when temperature sensor reads outside of a certain threshold and package it up to look like a cheap version of the nest. This proves precisely NOTHING in my estimation.
From TFS: It looks surprisingly good for such a short development cycle
It's trivially easy to *look* good - being functional is somewhat harder.
And building a 3.5 billion dollar company is just a *little* bit harder than writing a few scraps of code and soldering some bits together.
...is an open source bird to live in it.
Coder's Stone: The programming language quick ref for iPad
Yes .. the raw materials cost $70.
But how much of the CNC machine, the laser cutter, their time and also the time needed to come up to speed to know how to combine these items into one product? (and thats not even allowing for the design and marketing time that should be credited to Nest of actually coming up with a concept that sells)
If anything I'd say the real cost of this prototype is in the range of $20k at the very least.
I am Slashdot. Are you Slashdot as well?
I love the idea of home automation, and have been involved in a couple significant DIY projects involving my own scripting (no custom electronics design, that's not my skill). I've focused on things like multi room audio and intelligent video surveillance. If somebody offered me a $250 thermostat (yes, I saw that they did it for $70 here), my response is "really? Isn't that kind of... boring?"
Don't get me wrong, I'm happy that they're innovating in the thermostat space. But I want much, much more than that.
I just bought a new thermostat. I really wanted a Nest because of it's cool factor however, I ended up buying a Honeywell. First, the Nest isn't as advanced; for example, the Honeywell has some features that allow me to run the fan periodically throughout the cycles. It also allows me to add an additional "slave" thermostat and average the temperature between my upper and lower levels. While the Nest allows you to view multiple thermostats in a single interface each stat required separate HVAC systems. The Honeywell also comes with a remote control that sense the temperature where you are sitting and will adjust the set point to make you comfortable. The bottom line is that sometimes new and cool isn't as good as tried and true when you actually do some research.
I can't see much use for USB, but Wifi would make a lot of sense. You could move your thermostat, or perhaps install an additional one, without having to run another low-voltage line down to the basement. It would also allow more sophisticated communication than ON and OFF.
Arduino-based thermostat projects have been around for some time, and some are networked. Easy to DIY and can be done for under $100. Google around.
BTW, I have had my Proliphix network thermostat for more than 5 years now and still very happy with it.
Prove anything by multiplying Huge Number times Tiny Number
In the summer you can use the AC to dehumidify, also, it handles humidifier/dehumidifier.
Wow, sent an e-mail as suggested when clicking on "use classic" banner, and got a fast response that addressed my msg
You set it. It controls the heater. Jesus fucking christ already, this is where we're at in 2014?
I wonder how long it will be before the Nest software is reverse-engineered and an open source replacement for it is available, as has been done with router firmware and almost done with the add-ons for Canon camera software (CHDK). From what I can tell, there's nothing wrong with Nest's hardware, but a lot of people would like a different cloud and remote access model.
Acquisitiona? Isn't that Portuguese for credit note or something?
Confucius say, "Find worm in apple - bad. Find half a worm - worse."
so is that what Google is thinking w/ their 3.2 Billion?
they get energy companies and local installers to push these things and that's how they make a return on their investment?
it seems to me the profitability horizon is farther than the point at which the competition will be able to...um...compete.
energy companies have been at this for a long time, they are like IBM or AT&T in that they manage to stick around using old-school capitialism US big-biz style...google is an ad serving company known for its search and email
i'm saying energy companies, like my local PGE will develop their own more sensitive monitoring along side this
not that i'm criticizing google either...i'd love for them to buy one of my companies for a billion dollars
Thank you Dave Raggett
What's with the run on, incomplete sentence?
Why do we call the people who put Slashdot stories up "editors"?
Google's recent acquisitiona
Four words in and you've already a word.
they've posted their code on Github. The article...
What article would that be? Oh, if only there was a way of providing easy access to it via some kind of clickable "link"!
systemd is Roko's Basilisk.
About ten years ago, even evil Honeywell offered programmable thermostats in the classic round form. Now? Can't find one. I'd like to replace the current thermostat, but it's mounted on a large, round escutcheon and placing a rectangular box in the center of that would look too stupid to tolerate.
True. The real growth is not in home owners, most of whom will never replace a thermostat, let alone spend $250 for one or more replacements. The market is the installers and manufacturers to include it with a unit or as an add on sale.
Before 2006 (and well in to 2007 in many cases) there were a lot of businesses running on the belief that most cell phone users would never buy an $850 MSRP cellphone and that the "real" market is the carriers to include a phone with service plans.
Apple's made a whale of a profit out of proving those assumptions wrong.
the Nest can use your AC to dehumidify the house. Not as good a way as using a real dehumidifier but it gets the job done.
Not that I use the feature, it's regularly below 30% humidity in the desert.
I *have* tried to use the feature -- it does not work.
All the Nest offers is a choice to run the A/C until the temp gets to 3 degrees below the temp target you set. If you set a temp target of 80 degrees during the summer days then the Nest will not run the A/C if the temp is 77 degrees even when it detects a humidity level of 65%.
The more I try to use the Nest the more obvious it becomes that the developers either live somewhere where the climate is always mild or they are accustomed to wasting more energy on their home HVAC than Al Gore does.
There's some very good HVAC control technology that hasn't yet made it to the home. Here's a way to build a product that does that.
The basic kit consists of the cool-looking "thermostat" controller, and a window fan. The window fan unit has sensors and an RF link to the controller. The sensors include inside and outside temperature, humidity, CO2 level, noise level, and light level. The controller has the same set of sensors. The controller can turn on heat, A/C, or HVAC only, and has full control over the window fan unit.
Now we're ready to apply some smarts to HVAC control. The basic idea is to use outside air and recirculation when possible. Big building systems have done this automatically for decades, but somehow it never made it to the home. The problem isn't component cost - it's the difficulty of configuring such systems. That has to be fully automatic for the home.
So the controller spends the first few days of its operating life learning the thermodynamics of the house. When heat or A/C is started, how long does it take for the effects to show up at the thermostat? What's the rate of rise and fall? If it starts the fan blowing air inward when it's colder or hotter outside, how long does it take before the effect shows up at the controller's sensors? In a few days, the controller should have all that calibrated.
Why all the sensors? The outside temperature sensor tells the controller what the fans will do to inside temperature. The CO2 sensor tells the system how crowded the house is. When CO2 is higher than normal inside, but not outside, it's time to crank up the fans in exhaust mode. When CO2 is very low, the house is nearly empty and air can be recirculated. The noise level sensor is used to decide how high the fans can go before they become annoying. If someone is having a loud party and the CO2 level is up, the fans can be run at max. The outside temperature sensor tells the controller what the fans will do to inside temperature. The outside light level sensor is used to figure out the day/night cycle and length of day. After some time, the system will know the approximate date and latitude.
That's the base system. Add-ons include more fans, controls for built-in fans such as attic fans, humidifiers, and so forth. For larger houses, multiple controllers can be used, and will coordinate their operation. (Coordination is essential in multiple-zone HVAC, or you spend money cooling air you just heated, or vice versa.) For new houses, more of this could be built in, but the base unit plus a window fan is enough to get going. So it can be sold as a consumer product.
Note that this doesn't need to connect to the "cloud". It doesn't need to "phone home". It doesn't need your ZIP code so it can connect to some place and get outside air temp. (It might offer a WiFi interface so you can talk to it via a browser, but that's not essential.) It has a lot more smarts tha the "Nest", and will make a house more comfortable.
It sure makes me wonder why Google felt it had to spend $3 billion on a thermostat. There's nothing magical about Nest, nothing that a handful of reasonably competent engineers couldn't make an equivalent of, at a remote fraction of the amount.
They did a great job and they clearly demonstrated that Nest thermostats are highly overpriced and Google done goofed, but...
1. They used a CNC router. Who has these? They made it sound like any and everybody has a CNC router on hand.
2. They used a laser cutter to cut their acrylic parts. Who has theses? I've never seen one in real life. Yet, as before, they made it sound like everyone has one of these.
3. Did anyone else's browser free out when they hit the blog article? Firefox CPU utilization spiked and my CPU fan started racing, I presume due to the multiple looped HTML5 videos all playing at once on the same page. Holy crap!
Seriously, you wonder all those things? 3.5 people spent 17 hours, a total of ~60 man-hours.
There are lots of internet based CNC and laser cutting fabricators out there. Upload your drawing, enter your credit card and wait. Five days later your stuff shows up at the door. Cost? ~$100 for this project.
They stole the design, so that's free. As for marketing, they did it as a lark and put up a blog post. There are now literally thousands of people looking at it and talking about it. Many of them perfectly willing to drop $100 on it.
Your $20K number has been demonstrated to be WAY off base. The true cost of this prototype is in the ~$2,500 range and, that is fucking awesome!
What this summary needs is a good old fashioned link to at least one article about the project.
'The tyrant will always find pretext for his tyranny.' - Aesop's Fables
We've already got one
https://earth.esa.int/web/nest/home/
I'm positive Google is unable to resist the urge to start harvesting data from the Nest thermostats.
Crank up the heat, and you'll start seeing ads for blankets and pillows in your browser (if not on the thermostat itself!). Crank up the A/C, and ads for cold beer and swimwear appear.
They didn't build a Nest, they built a remote controlled thermostat. The Nest has a lot more smarts and adaptive logic in it to learn cycle times, self tune the control loops, etc.
Oh, and a user manual...
What they did is assemble the hardware needed to provide a development platform for a Nest-like device (without some of the sensors). They're several weeks/months worth of work away from duplicating Nest functionality. Now, it's true that someone with an open source control algorithm itch to scratch might decide to pick this up and run with it.
But this is kind of like saying: here's a keyboard, monitor, CPU, memory and hard disk all wired up. Look it's a PC or Mac, all you gotta do is reprogram it.
Just practicing the headline for a relevant piece future news.
My dad has the non-wifi version, and having just got my own place I'm planning on ordering the wifi version soon.
http://www.heatmisershop.co.uk/room-thermostats-c1/wifi-thermostats-c12
Not really sure how it compares to the Nest model, but it isn't cloud based so the NSA won't know how warm your house is. There are various related projects on github for alternate interfaces to it as well.
So 3 engineers are free too? You do not value your time and skills?