> reducing ewaste
The solution is to stop installing advanced touchscreen computer systems where they don't really belong.
I've never seen a Honeywell thermostat fail, get hacked, or "lose support". I dont have to worry about what will happen if there is a weird software update or power outage. There are physical switches that configure the current mode of the system. Only the temperature setting is volatile, but it defaults to reasonable settings depending on the current mode (60F for heat, 80F for cool). Because of this simple design, there is no option to install a battery. There's nothing to remember.
The Nest Thermostats listed here are falling back to "Honeywell local thermostat" functionality.
They will still work for turning the dial locally to control the temp. They just won't be IoT connected.
The 3rd gen Nest thermostat was released in 2015. The 2nd gen and 1st gen were released in 2012 and 2011 respectively. A decade lifespan for a device that gives me metrics on my climate control and lets me tune it from my phone is pretty solid in my book.
> A decade lifespan for a device that gives me metrics on my climate control and lets me tune it from my phone is pretty solid in my book.
That's the attitude I take issue with. The device should work as it did at the time of purchase in perpetuity. If it can't, because of the architecture of the product (i.e. needs third party servers on the Internet) it shouldn't be marketed as being sold, but rather rented.
Acceptance that products we buy will just stop working (or, in the case of this product, losing significant functionality) is only encouraging companies to ratchet down lifetimes and hurt consumers more.
This feels like a distinction without a difference.
If all the shelves at Lowes that sell network-connected IoT thermostats said "10 year rental price" instead of "purchase price", the average non-nerd consumer wouldn't change their shopping behavior. The nerds... maybe they would? but they're the ones who already realize the impact of a device needing cloud connectivity to function.
The reality is that vendors cannot and will not offer "in perpetuity" commitments to run the cloud portion of their gear. For strategic reasons like "they want to deprecate legacy endpoints and functionality on their end" but also unintentional reasons like "the company went out of business".
Everything you buy has a lifespan. Physical objects degrade over time, objects with cloud connectivity rely on external functionality, etc. Picking between them is part of the decision making process: do you buy the cheap furniture that falls apart the first time you move? Or the more expensive furniture that lasts decades?
Should smoke detectors be listed as "rentals" because they degrade and have to be replaced?
People will often choose to not have a countdown.
I buy a smoke alarm with a fixed 10 year life to avoid battery replacement. Others buy models with replaceable batteries, still others install wired alarms.
You should disclose the lifecycle at sale so consumers can make a decision. I’ve been in the computing infrastructure business for many years - I can model out the expected life and cost structure for all of my gear.
Deviating from your argument, but 10y battery life is a EU regulation. In some EU countries, you cannot rent out a living space without smoke alarm that has a 10y battery in it, even if it’s wired in. If you’re looking for a high quality smoke alarm (ability to trigger on different types of fires and certified for EU) I failed to find one that came with a replacement battery, and only a few that are wired in, all of which will have a 10y lasting battery. The more expensive units do connect with internet/vendor and can do smart stuff like call fire department etc, you’re right that there’s a choice. But, the choice is very limited.
I’m not sure that an average consumer is even aware that some smoke alarm sensors degrade with time, or not trigger at all types of fires, there is probably some logic in forced unit replacement every decade.
That’s interesting. There is a certain logic to a regular maintenance cycle for high impact items.
It may be an example of an unintended consequence - to be fair to the Nest people, when you’re releasing a new product, you may not be thinking of end of life. Today we should be stamping a support date on them - Google is pretty good about this with Chromebooks, for example.
The forced upgrade assumes that sensor technology will improve between purchases.
> Everything you buy has a lifespan. Physical objects degrade over time, objects with cloud connectivity rely on external functionality, etc. Picking between them is part of the decision making process...
Exactly. Consumers have no way to know, right now, if something they "buy" is truly a purchase or a rental. They can't make informed decisions without that knowledge. A technical person can tell the difference, but the average consumer cannot. That's why the distinction needs to be explicit.
> The reality is that vendors cannot and will not offer "in perpetuity" commitments to run the cloud portion of their gear.
Right. That is precisely why nothing that is meant to last should have any kind of dependency on "cloud". It must run locally only, so there is nothing that can be deprecated or dropped.
My mid 90s thermostat will continue to work independently of whether the original company that made it is out of business. Everything needs to be like that.
this is where the FTC should step in and say "FOSS it" so at least consumers can have a choice and companies compete. Google could argue that it costs some nonzero amount (which is infinitesimal, I know), therefore they can't offer it perpetuity, but they could be required to unlock it so a 3rd party can compete and offer you the same service or let FOSS users control it.
A decade? Wow. If 1.5 years is a computer generation, the. These are more than 12 years or 8 generations behind. They should warn people about this. 12 car years = 8 computer years = 200 human years. I hope they still keep track of time.
Or make the device as SIMPLE as possible.
The thermostat should be a remote display and a sensor. The real brains should happen on a controller near the heating system, that can be swapped in and out as time goes on.
As is the thermostat is a set of brains that can be swapped in and out as time goes on. I just swapped two thermostats in two homes a few weeks back and it took <20 minutes each with no prior knowledge, even though in both cases I had to swap the backplate.
What is the benefit you see of moving the swappable brains closer to the heating system? Doesn't that actually make the whole system more complex because now you have two separate devices where there is currently one?
The controller for the HVAC system costs a heck of a lot more than the thermostat.
The thermostat is the controller is most simple HVAC systems. The evaporator and air handler will have some "brains" re-compressor cycling and potentially fan speed, but the thermostat is calling for heat/cool.
You mean condenser instead of evaporator, the evaporator is inside the furnace and has no controls as it’s just a coil. The outdoor a/c unit contains the condenser and compressor and has the contactor (and controller if it’s an EC motor) for the compressor motor. The contactor control coil is 24v and the contact is closed when a call for cold is received from the thermostat.
Newer A/C units have EC motors with variable speed motor controllers and newer furnaces will have an EC motor blower fan and lots of factory control points wired up to a controller, mostly for the burner (and related gas/burner safeties).
Still, even in newer furnaces and a/c units, the thermostat is the main controller as like you said, the calls for heat/cool are the only input to the system.
Commercial HVAC controls for air handlers, boilers, and chillers have a lot more control points, an AHU can have 40 or more control inputs and outputs tied to it. Discharge and return air temp, multiple differential pressure sensors, static pressure, outdoor air temp, outdoor air pressure, duct avg temp, etc. These systems are much more complex than any home HVAC setup and usually have multiple controllers networked back to a master controller which orchestrates everything.
Did you read my last point about "a way for Home Assistant based HVAC control setups to work with energy company savings programs where appropriate" ?
Unless you're living off-grid, I think there's still some benefit to having one of the largest energy loads (building HVAC) coordinate a bit with utilities to ensure there aren't blackouts and such. Even better if there's an approach that works with open-source solutions like Home Assistant.
Where I live staring at the grid load and available energy is a thing during heatwaves and other inclement weather, and the utility generally issues "energy rush hour" limits to smart thermostats to help ensure grid reliability. The utility also incentivizes this I think by providing a small discount on the utility bill.