Hacker News Reader

Comment by paradox242 | Hacker News Reader

The idea of a washing machine, coffee maker, refrigerator, etc. including a network card strikes me as a solution in search of a problem. Unless you are the company marketing these devices and the problem is "how can start getting in on the user data collection game".

View on HN

If the machine had an open API that I could control and utilize entirely locally, I can see (minor) advantages to most of these. Really anything that is a somewhat highish power load and only needs to run very intermittently, having a way to control it with home automation (which requires some kind of wireless connection), could be of some use.

THe problem is that they are including this network connectivity in a completely closed down way where you, the user, do not actually have any control over what is sent to or from the device. that kind of connection I agree is worse than useless: actively malicious and bad, and I would pay extra to not have it.

My plan to solve this is buy an all in one like the new ge prestige unit when my washer or dryer breaks. That runs on a regular outlet so there are many, many cheap esphome based outlets that monitor power draw that will work. At that point its a trivial rule in home assisant to notify me when the thibg is done.

Do those even work with the type of power load a washing machine needs? I'm thinking of how you shouldn't connect the washing machine to a power bar

If it can handle the normal amperage a US outlet can supply, it should be fine. Washers don't use that much electricity; they run on standard 120V/15A outlets in the US.

Exactly, and they also provide the network stuff over and above an already complex local lcd/button/touchscreen UI.

They should just provide a local API, conforming to a standard, and then 3rd party remote control software could drive it. The physical UI on the machine itself is then just and on/off start/stop button.

Why paying extra? You just don't connect them to your local wifi, that's what I did with my Bosch devices. It's not like they can connect by themselves (can they?)

For now. I could see a lot of these devices including a cell modem in the future.

I don't particularly want any appliances that connect to the internet, but it would be quite fun to have a washing machine API that reported sensor data like water temperature, weight of load, drum rotation speed during the spin cycle, door lock/unlock status etc.

I can't imagine there's much value in LG knowing how often I wash my clothes.

On the other hand, it's nice to get a notification when the cycle is done if I'm on the other side of the house. Refrigerator sending a notification to my phone if the door isn't closed could come in handy some day too.

LG doesn't care - the advertising industry does.

data is always valuable to someone. if you collect it, then you can sell it to someone who desires it.

> I can't imagine there's much value in LG knowing how often I wash my clothes.

Depending on what their sensors can collect, usage data could certainly be helpful in building a more robust product.

Re: the refrigerator door. I’ve always wondered if just engineering a door that shuts itself is a better option here… like a servo that swings the door around enough to shut… what’s the overhead for either unnecessary piece of technology…

What if there's an obstruction?

Then it won't be able to close.

The main reason my fridge door doesn't close is if there's an obstruction. So a motor won't do much (unless its unreasonably strong). It's easy enough to adjust a fridge to tilt slightly backwards so gravity closes the door by itself.

what if the cat’s inside?

Depends whether the cat is dead or alive

It's both

Just incline your fridge a bit so the door closes on its own weight.

A no tech solution is better than a low tech solution.

This doesn't work; the door needs to swing to 140 degrees, once it swings past this range, it would never be able to round the corner to close.

I've had an idea for the longest time to hook up a smart outlet to my washer and just monitor when it starts using a bunch of electricity and when it stops doing that. That way I could DIY myself washer notifications, since I never hear when the machine goes off and its time estimates are near useless.

If you decide to do that, make sure you use something rated to handle that kind of load. The cheaper ones really aren't meant for more than switching on and off some lights.

I actually have a Tapo P110 ready for it, just haven't gotten to the "doing" part of "do it yourself". It has "MAX 3680W - MAX 16A" printed below the socket, so I assume it's fine.

Both use cases could be done with simple speakers.

It's frequently spoken about all kinds of devices that they are mining data.

I'm actually really curious, as someone far away, what does this actually look like?

Are PMs at these companies putting data collection stuff into their roadmaps "just in case," or is this a directive that comes from the higher ups with a specific business plan for how they are going to monetize the data? How does someone actually buy washing machine data? Are they selling the data? If so, where, how, to whom? Do I approach LG or does LG approach me? Are the numbers of how much money these companies are making on data actually real? Is there an eBay for washing machine data? Is it an API that streams it out in real time that they expose or do they just give you a 100GB CSV?

So many questions. So many.

A washing machine that is networked can be scheduled to run automatically when electricity prices are low, or when your local solar panels are at full production.

Unfortunately, it can also be used to pipe data from your TV that you intentionally have not hooked up to the internet because you don’t trust the shady things certain TV manufacturers do with your data.

Seriously? It can't be scheduled if it's not networked?

Can't is a really strong word. Customer convenience drives sales.

You need to think about the average consumer. You first have to time sync. Then you have to program it. Do you remember VCRs? 9/10 people could not program their VCR. Oh yeah, and when the power goes out and you forget to reset the clock your schedule is no longer correct.

And with that, you still have no dynamic control.

These days you can put in a touch screen UI that more people would be able to understand. You can spend those funds you wasted on the cloud servers and maintenance to build it.

> the power goes out and you forget to reset the clock

Btw what's with that crap in 2024? Alarm clock manufacturers have mostly learned to put in a slot for a backup battery for the time, why can't anyone else learn that?

My coffee maker can be programmed to start at a certain hour - with some difficulty :) - but if the power goes out it loses the clock. It wasn't the cheapest, they had enough BoM to put in something that detects when it needs descaling for example. So why the hell not a battery slot too?

The de scale is just a timer.

After x hours of operation, the light will go on.

> You need to think about the average consumer. You first have to time sync. Then you have to program it.

Yes. If you want to save money you have to learn something, even something so incredibly complicated as setting the clock and adjusting the schedule.

You can wire a relay in parallel to the start button and control it with your own controller.

Ripple control (low-frequency digital tones superimposed over AC) worked before the Internet.

I wish that I could feed my washer and dryer into my home automation system, but I can’t. I want to do this so I know whether the laundry will get stinky or not. Right now I measure the power consumption and have a sensor on the door of my washer and dryer. That works, but imagine if I could just get an mqtt stream of events! I will never connect an appliance like that to my network in the real world, but I can dream!

I have a "dumb" washer and dryer and I would like to make them smart as well. Mainly I want alerts when the cycle stops. Can you share more details about your solution?

I do this with a Zigbee smart plug plus a simple rule in home assistant: notify me if power use drops below 2w for more than 5 mins.

I use home assistant, with the same logic as the other poster mentioned. Home assistant allows you to create automations based on events.

Just mount a small accelerometer on them to tell when they're operating. There are plenty of cheap options on the market nowadays.

I did the accelerometer at first. It was way easier to measure power consumption. :)

Far from it, for different reasons. All appliances will be network connected in the future as the world electricity supply shifts to renewables. It simply doesn't matter if my e.g. dishwasher switches on at 10PM when I press the button, just that it's ready the next morning. But it matters if it runs when electricity is expensive because of low wind or cheap because of high wind. Or if there's a lot of sun during the day or not. This will not need gigabytes of data though.

edit: should've read more, someone already mentioned it in the thread

but while I am here, another good example is for instance heating. I visited friends in their small mountain house recently and was wondering how they got it warm so quickly because they just got there. They said, oh we have a new oven, we switched it on remotely when we were on our way driving.

I have an app install on my phone for my Siemens dish washing machine.

Things that are genuinely useful about it:

- It tells me when a cycle has finished.

- It keeps count of used tablets and I can customize the count at which it reminds me that I need to get new ones.

- It tells me when I should do a "cleaning cycle", i.e. that I should run an empty load in "Machine Care" mode to avoid it starting to get odours. However, a bit annoying about this it only tells me as I turn on the dishwasher, i.e. just as I've finished loading it and am about to run it, it would be much better if it told me at the end of the previous cycle.

- I can schedule a cycle to begin/end at a particular time.

Mildly useful:

- I can see how much time is left of the cycle (it's a built-in unit meaning I can't see any display without opening it).

- From the app I can choose to "pause" the cycle for a little, this could be useful when watching TV or doing some other activity where you want quiet until later.

- I can use the app to select the desired mode and begin a cycle, however, this is only barely useful as I can't use the app until I press the power button on the dishwasher (however, not that relevant since if there's no tablet insert, there is no way it could possibly be started remotely anyway).

If there is no significant extra cost, I would like my next clothes washing machine to be able to send me a push notification when its cycle is done.

A smart oven would also potentially be nice in that you can remotely turn on the oven from a place like your office or at a particular time so that it's already heated by the time you get home, or you can tell it turn off after a timer.

I would definitely prefer if it was an open platform, but in truth, as a founder in small startup who also has young kids, I don't have the time right now to work out how to set up custom automations.

> - It tells me when I should do a "cleaning cycle", i.e. that I should run an empty load in "Machine Care" mode to avoid it starting to get odours.

Leave the door cracked so the drum can completely dry between uses and the machine will never smell.

Same goes for washing machines.

My new Samsung induction oven has an app. That app doesn’t do the only functionality I could ever care about - turn it on to preheat! Best it does is allow you to set the temp and bake/broil etc, but then you have to physically press start on the oven. What a silly waste of an app.

Ultimately, though, that's a safety feature. You don't know what's happening with the oven, what's on it, or etc.

I have a Smart Load Center (main electrical panel) from Leviton (https://www.leviton.com/en/products/residential/load-centers...). It does a lot of thing, can show me events of power loss etc., current draw per circuit, etc., etc.

What you CANNOT do is reset a breaker via the app. And you never will, per my electrician, for safety reasons - is there a fire hazard? is someone working on the cabling? And so on.

Ok, I can see that, but shouldn’t it be my risk to take in this case (which seems quite a bit less risky than flicking a breaker on remotely when someone other than the app user might be working on it)?

i surely hope that safety feature is designed in such a way that you can't ever actually turn on the oven without standing in front of it, lest someone is able to break into the system and turn it on without you knowing. the risk is just to great.

sure, the benefit of the app at that point is questionable, but the alternative should then be to simply have no app.

> lest someone is able to break into the system and turn it on without you knowing

Allow users to set their own unique PINs that they have to type in to unlock the “oven on” command every time they use it.

Surely there has to be a tradeoff here where a user can be allowed to be a grown up and make their own decision. I mean, I have to click “I promise not to look at the giant screen on my dashboard while driving” every time I start my car.

And what’s the worst case? That my oven is on and wasting money? If it could be set to get so hot that it causes a fire then that’s a totally different issue.

the most likely scenario is that something is inside the oven that should not be there. and that causes the fire. say you bake something, and then leave it inside to cool off, and then turn the oven back on forgetting that there is still something inside. so turning on should only be possible while you can actually see what is inside.

Or what's in it when you turn it on.

Likely, during the development process of the product, an App was desired. And these ideas about remote preheating were included. But later removed when the conceptual process ended and the Regulatory Compliance phase was started.

> The idea of a washing machine, coffee maker, refrigerator, etc. including a network card strikes me as a solution in search of a problem.

There are tons of problems that networking these solves. Maybe minor to most people, but convenience is key.

A common example for me: If I am packing for a trip last minute and doing laundry, I don’t want to have to keep going down two flights to check if my washer or dryer is done, disrupting what I am doing.

But I don’t like the idea of everything being connected, so I will live with the problem.

The cheapest washer/dryer of the past 3 decades came with a audible timer; most annoyingly loud.

Mine is not loud enough, even at its loudest setting (mine can be controlled with a knob) for me to hear it in by bedroom. But it will wake up people sleeping right above it.

My washing machine shows how many minutes it will take based on the combination of options chosen. I mostly just remember the number and walk down after so many minutes. Or I could put a timer for that many minutes on my watch. No need for a network stack and spyware on the washing machine.

I'm wondering - are there developers who think that this functionality is something any user actually needs? Or is it the higher-up part of the company pushing on "we need internet connectivity in every appliance we put out and we need it by yesterday"?

It's worse than a "solution in search of a problem." It's a problem in and of itself. Networked appliances almost always need some remote back-end to talk to. But companies don't have much incentive to maintain those back-ends for the 10-year+ life of a washing machine, coffee maker, refrigerator, etc.

So you end up in at least one of two situations: Either you lose functionality over time as companies deprecate those back-ends even though the hardware is still good (i.e., forced obsolescence leading to e-waste), and/or your outdated device ends up being a target for botnets to propagate themselves.

IoT is such a damn scam in the long-term.

I am a practical person and I'm not above trading convenience for security, but I have quite a few "smart" devices at home that endlessly blink waiting for me to connect them to a network as I utterly see no point in that.

If the execution wasn't shit, I would love it. It's not important, but I would find it fun if I could control/get data from these devices over network.

The execution, however, is almost always shit.

Meanwhile, I'm here with:

* Smart Air Filter

* Smart Air Fryer

* Smart Meat Thermometer

* Smart Sous Vide

* Smart Coffee Maker

* Smart Blinds

"Smart Blinds" are much more useful than your other examples IMHO.

Is there any FLOSS washing machine software?

Reply on news.ycombinator.com