Tony Fadell: The Nest Thermostat Disrupted My Life
spectrum.ieee.org
spectrum.ieee.org
For those unaware, the majority of homes in the UK have a gas boiler central heating system with TRVs on each radiator. This means that you end up with two competing temperature control systems in your home, which result in some rooms regularly being too cold/hot, it literally worse than have no central thermostat.
We now have the Nest set to about 5deg higher than we want, then have all the TRVs set to what we want each room to be.
In our last house we had the Tado system with “smart” electric TRVs, you would think that would solve the problems, but it was flaky, noisy and very expensive.
If I was doing it again I would get whatever the cheapest boiler controller with remote (internet) control I could find. But then I would probably not be putting in a gas boiler again, I'm hoping that by next time we need to overhaul a heating system heat pump systems have dropped in price in the UK.
I'm sure that in countries where people tend to have forced air HVAC systems these thermostats make a lot more sense. And I do love the industrial design, it is a "beautiful" thermostat.
Ever since work from home the “away” hasn’t mattered much and the house has so much thermal mass that letting it cool down a bit when we are actually away doesn’t seem to do much at all.
When I moved into my current house we had painters working on one floor. They cranked the newly programmed Nest that was in learning mode to 90 degrees for one night to keep the paint warm while it dried. I, of course, had no idea they had done something that silly, but for weeks afterwards this smart thermostat will crank the heat to 90 and I have to manually turn it down. I'm convinced that it will not be able to unlearn and I'll have to delete the profile and recreate it. I should set it on a schedule as you mention.
My heating/cooling desires are straightforward: unless I'm gone, stay in this range. I am not gone on a predictable schedule, any pattern it picks up will be incorrect.
The only thing the "ugly beige box" doesn't do that I want is remote access, and that would only have been handy a couple of times in the last 11 years I've been in this place. And that is not worth the surveillance or the freakishly buggy behavior some people report.
I'm increasingly convinced the major effect of "AI" will be to ensure that instead of bugs not being fixed because ultimately they aren't considered worth fixing, bugs will not be fixed because nobody understands them.
It works flawlessly. It's been months since I manually tweaked the thermostat. When I do need to override it, it keeps that override in place until the next leave/return event.
Very much this. Either the pattern is predictable which makes it easy to program into a simple thermostat or it is not in which it'll always be wrong in any "smart" thermostat.
Nest and similar are great case studies on the ills of overcomplicating something that should be super simple and reliable.
My mid-90s thermostat is much more reliable. I programmed my schedule into it decades ago and it does its thing with zero confusion ever since then. When I need to override it's just a button press and it automatically resets back to the schedule next cycle. Nothing to go wrong.
It's simply:
- Has WiFi (update from phone)
Then also:
- Being able to link with Alexa or similar to do things like turning off when you leave automatically
- Custom programming modes. So instead of setting a temperature at a single time during the day (home/away), you could do (morning/home/home2/afternoon/night/away)
- Can run my fan once an hour automatically for just 5 minutes
- Looks cool
Chinese branded and produced ones like Tuya and MoesHouse are really the only option.
Even Honeywell aren’t producing them.
Ours, somehow, has gotten into a state where it thinks we like to bounce between 68 - 70 degrees all day, and doesn't "learn" when I try to adjust it back to the same temperature it was. The weekly schedule has over 70 different temperatures on it and there is no (obvious) way to reset it, so I haven't bothered to go clean it up. I think it could do better there.
Then to edit it you have the worst possible UI and end up, like you said, turning all off the smart features.
The only way to reliably set temperature is by reading it elsewhere in the room. With radiators that’s really inconsistent because they’re slow to heat up and slow to cool down if you “over heat”.
Also the window detection feature on any TRV should be disabled. It’ll only cause you issues.
I simply have a “fully open” or “off” control from a Zigbee gateway per room. I’d consider adding sustained motion sensors to each room but to be honest just having a timer to turn them on in the morning, off at bed time and the ability to adjust and view what rooms are on manually from my desk or phone is more than enough.
There’s value in electric TRVs but not in the automating part beyond scheduling.
Thermostatic Radiator Valves for anyone wondering.
In a while I realized that to make the experience smart would not be to add a different thermostat but to change the windows and have some proper insulation. To get that I had to ditch the whole country tho. It's just abysmal how much gas is fired to heat the streets.
Insulating an old house in the city is expensive but it had been there for 150ish years .. the breakeven from adding insulation would be what? 10 years? 20?
My thoughts: 1. Lots of Steve Jobs talk. There's a whole chapter on the distinction between real assholes, and assholes that just really care about the product quality / customer. The distinction drawn was in motivation - but I wonder if it might just be a winner-writes-the-history situation. 2. Some advice goes heavily against current startup orthodoxy. He rejects fail-fast / figure it out later mentality, and argues for a lengthy product design process (for both atoms and bits!). 3. Lots of good details about how to think about managing people, managing and scaling team, and issues with scaling, etc, taken from his days building Nest.
Both (2) and (3) stand out to me as functions of his specific background. Building a company like Nest is obviously wildly (!) hard, but doing it after building the iPhone is a different game. I don't think his advice about lengthy design processes make a ton of sense for my startup [1], for example.
Generally, I always try and remind myself when I'm reading company-building advice books: this is probably good advice for someone in the same context as Tony Fadell (e.g. someone who is launching a second company after building... the iPhone). For me, it may or may not be relevant.
The good thing about failing fast is that you can use it at a strategy-level, aka fail fast _at_ failing fast, and switch to longer product cycles if you think that might work better. Starting with a long-product cycle doesn't give you much of a chance to try again if your first swing is a miss (which Mito's first attempt was...).
There's many good gems here and his philosophy hits close to home.
I don't think you need to be of his fame to apply it. Many of these tips are applicable. I find lots of his product & marketing advice to be very useful for new PMs to tech too.
He's been doing a lot of history rewriting.
Don't get me wrong: Building and shipping Nest was a great accomplishment. However, he was a famously terrible leader in many regards and drove a number of great employees away.
When Nest acquired Dropcam, a large number of the Dropcam employees left. Tony Fadell then went on to publicly disparage the Dropcam employees as being "not as good as we hoped" saying "unfortunately it wasn't a very experienceded team" ( https://www.businessinsider.com/nest-ceo-tony-fadell-has-dro... ) Predictably, that's a great way to drive away the rest of your knowledgeable employees and make the company toxic for hiring.
Nest didn't go on to revolutionize the space after Tony Fadell arrived with his hard-hitting management styles. Nest cameras haven't really evolved much and the Nest security system was abandoned.
I don't doubt that there are some lessons to be learned in his book, but in practice Tony Fadell hasn't been a great leader in the past decade. I'd take his book with a grain of salt.
^^ This.
Aamir Virani, who was the co-founder of Dropcam, has his take on this here: https://aamusings.substack.com/p/who-lives-who-dies-who-tell... (also https://threadreaderapp.com/thread/1522280960984707072.html)
We had a git repo that /only/ contained design docs and recorded decisions as PRs. Nothing was discussed “offline” so everything was recorded. In meetings we would make the final decision and actually merge the said PRs.
It was so different from working on any other project, and actually more fun.
Seriously give a team at your startup to try that route. I think you’d be surprised after a few months to a year. The quality of the work we output was light years beyond any other team at the company, IMHO.
We got a Nest thermostat for free as part of a SolarCity solar panel installation, and I spent months fighting with it to try and make it do what I told it. The Nest has a motion sensor on it, which is how it determines whether to turn on the energy saving mode - if it senses that you're home it will run at the set temperature, otherwise it will disobey your commands and go to the configured energy saving temperature. There were 2 problems with this, one universal and one application-specific: first, if you live in anything other than a small apartment your thermostat may be in a different room from you. In our 2-story house, the Nest thinks nobody is home unless you're specifically in the living room. The 2 2nd-floor bedrooms, the 1st-floor bedroom, the kitchen, the basement, the dining room, and the 1st-floor office all don't count. In our house specifically the thermostat was installed behind a wall-mounted TV, which only makes things worse - even if you're in the living room, the Nest thinks nobody is home unless you're actively playing with it.
Trying to make the Nest ignore whether you're home and disable the eco mode is an exercise in futility. I found a setting to disable eco mode, but the house was still never the temperature we set it to. When I checked the thermostat I saw the green leaf icon, indicating that the Nest completely ignores the setting that disables eco mode. When I called Nest support, they couldn't figure out what was going on either. The only way I could get it to function semi-reasonably was to manually set the eco mode temperature to the temperature we actually wanted the house at, which means that you have to set the temperature twice to actually set the temperature.
The idea of a thermostat that tries to save power automatically is great, but it falls apart when you rely on a single motion sensor to determine if an entire house is occupied. This bad assumption was a letdown, but the fake setting to disable eco mode turned me off of Nest completely.
And it works just like a normal thermostat, which makes me happy. I’d never buy a smart thermostat now, I keep using Nest because previous owner of the house installed it, and after lobotomizing it, it works good enough.
At one point, I had a cheap WiFi thermostat. No smart features, just allowed remote control via app or web page. At one point, I had reverse-engineered the web app API and wrote a script to implement "away" detection and shut the thermostat off if both phones were away.
The problem was two-fold. First, it detected the phones being away by simply pinging the phone's local IP address. If the phone was in a power saving mode that disabled WiFi, the script would detect the phone as away. Second, I ran into the problem the article author had: If everyone was away, it got cold in the winter, and an absolute oven in the summer, and my aging A/C unit couldn't get the temperature back down until the sun went down.
The only nice thing about Nests is the overall look, but they're really poor devices. The OS on them constantly lags, causing dial inputs to suddenly "catch up". Each one of our Nests has its own personality: one lags, one turns on when it wants to (prior to factory reset, probably the third or fourth time we've had to do so), and the other one seems to actually just work.
I have 2x Nests in my house and I don't know if I'm missing something, but I only get a usage history for the past week. I feel like a "smart" thermostat should tell me a lot more about my usage and it should be pretty trivial to develop.
I must be in the minority then. I end up with a ton of these product-specific screwdrivers and get a little frustrated that I have to throw it away, contributing ever-so-slightly to the pile of consumer waste.
It'd be cool if you could select a "need tools" option sort of like how take out now has a "need utensils" option.
The rest go into the landfill unfortunately, and I agree the waste isn't really worth the 3 random sizes of allen key I once needed.
As I go through my stash of direct-from-the-Shenzhen-market-at-10x-the-price fasteners from Amazon I'm replacing them with equivalents from McMaster. I have to say the increase in quality is noticeable. I half expect to break an Amazon screw and find that it's hollow inside to save a few cents.
The wrench definitely made the product feel higher quality, and I have a tool I will keep for a long time, if I don't wear it out.
And that's just here in the US. Our working poor are the world's one percent. If our definition of consumer waste is things no one would like to hold on to, then even what we consider the most banal tools are not consumer waste, and it will be a very long time until they are.
I bought three Nests and while the thermostats are pretty annoying, I really love the screwdrivers.
They are really pleasant to hold.
For the past four and a half years it's been on a shelf in the basement, in its box. Gonna throw it out someday.
What crappy tech. Its problem: its backing plate heats up. Causing the house to be perpetually cold.
Three Nest backing-plate replacements later, we gave up.
We presently have a much dumber thermostat that can be remotely controlled. We're happy now, but we'll never purchase another Nest product again.
Either way - she hasn’t installed it and it’s worked fine. It’s possible you had a dud unit?
I work from home so turning up/down the temp when I'm away is rarely worth it other than when I'm on vacation for a week or a long weekend. I'm happy enough to be able to control it from an app and not have to get up and go to the other side of the house but there is one big thing that is missing for me: it being smart about the outside temp.
Let me elaborate. I want my house to be 65 during the day and 60 at night but there isn't a good way to accomplish that without micromanaging it when the seasons change (or freak weather which happens more and more often).
Depending on the outside temp (and humidity) I have to manually tweak the temp ranges. It would be nice if I could set heat to 65 and cool to 65 but that would kill my HVAC system flipping between the two modes (if the apps even let you do that), I just want it to be a little smarter. Something like "If it's >60 outside then set the cool to 65 and heat to 55" and "If it's <60 outside then set the heat to 65 and cool to 75" (though humidity, inside and out, probably need to be taken into account as well).
I'll probably attempt to tackle this using Home Assistant at some point but it's frustrating that these devices are in some ways stuck in the past of "Set a cool or heat or cool+heat settings", how about a little more intelligence? I want the inside of my house to always "feel like" 65F.
All of the "smart" thermostats are really just the same as before with a way to remotely change the settings (which is an improvement) but I'd never call them smart.
I think humidity is playing a part in it but I don't want my house to get up to 70 so setting 60-70 isn't going to work and likewise, no matter the season, I don't want it to get down to 60 unless it's at night.
I've dialed it in so I know I love 65F, that's what I want my "smart" thermostat to facilitate.
She had issues with these two particular tenants who seemed to have an obsession with fighting over the temperature dial.
It cost her a whole ton in either heating or A/C every month, and would leave the whole place either ridiculously freezing or way too hot all the time.
(She lives in the basement herself, so this affected her as well.)
She bought a Nest, put a password on it, left it there, and has likely already saved around what the Nest cost within a year or so just from stopping these idiots and future potential idiots from constantly messing with the thing.
She’s never had an issue with it.
Essentially - it really depends on what you want a product for, and how you intend to use it.
I’m aware there are other, cheaper; less convenient solutions such as a lockbox, but honestly the Nest has been a wonderful solution for her.
That is first world problems hall of fame material right there!
But a great origin story for the Nest regardless.
Granted our thermostat isn't really "smart" it just has an API exposed that integrates into Home Assistant.
I wouldn't put in one of those cloud based ones like Nest or Ecobee
1) The thermostat drives most home energy usage. It's probably above $1000/yr for most households.
2) The existing programmable thermostats were awful and no one used them effectively. There's a ton of waste as a result.
3) The smart thermostat can easily pay for itself in a year or two just by having a better UX that you'll actually use. That's an amazing value proposition.
- it didn't look crap;
- I could figure out how to work it
I'm not so bothered about it being smart, least of all for it's 'learning', it just looks so much better on the wall (visually prominent when you walk in) than the white plastic sharp cornered box the developers put up.
If I spent more time/energy on it I'd do away with a wall-mounted thermostat entirely, and just have a HomeAssistant-connected relay right at the boiler.
That said, I like the Nest, though I'm partial to the Ecobee because of the remote sensor it includes.
I wish Google had sold it, then I could buy a Nest. I do not trust Google and do not want them to have more of a foothold in my house, my networks.
The first iPhone made connectors and electronic components easier to get to? Sure, it indirectly helped, as any other new, recent electronic device did. But please don't try to rewrite history.
He couldn't make things work over the phone, although he invested large sums of money and dedicated hardware? Come on, I rigged an old analog (rotary) phone to do that using a relay when I was 14 or 15 years old. After all, in that case all he needed to do was to switch on the heating and nothing more.
The whole thing just reads like a marketing fluff piece.
I'm happy for Tony. But I also miss my grandfather and his kind of engineering.
Thanks for sharing.
We need cooler, game-changing hardware. Not a janky drone (looking at you Snapchat)