Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Ask HN: Am I the only one who does not want IoT in their solar panel setup?
218 points by harryvederci on Sept 30, 2022 | hide | past | favorite | 199 comments
All solar panel systems I can find seem to require an internet connection. The (seemingly) best option I have found so far is a system with an inverter that only requires an internet connection during the initial installation, which uses a wifi dongle so I can probably unplug that afterwards. Even then, the manufacturer states that not having it connected to the internet voids their warranty, and I'm afraid the software might even have some kind of countdown where the system will stop working when it hasn't been connected to the internet after a certain amount of time.

I am not interested at all in connecting something that should help me get off the grid to the internet. Combine that with security issues[0] and I'm starting to think we're collectively folding our arms behind our back and intentionally falling forwards.

Maybe I'm too paranoid. Thoughts?

[0]: For example: https://csirt.divd.nl/cases/DIVD-2022-00009/



I bought a house with solar panels 6 years ago.

There was an ethernet cable plugged into the inverter panel, which ran to a Comcast box.

We switched to FIOS after a couple of months, and never switched the connectivity of the solar panels. Everything worked fine, but ...

A few months later, I received a strongly worded email from an organization I had never heard of, saying that I needed to get the solar panels back online, because it was a requirement of the contract the previous owner had entered into to sell the SRECs generated by the panels. Failing to connect the panels to the Internet could lead them to pursue action related to a lien they had on our property.

Wait, what contract? And what lien? This was the first I'd heard of either, and the lien hadn't come up during the title-insurance process.

It turns out the previous owner of the house had entered into a contract with an organization in our state that pays you a lump sum, if you sign away the rights to the SRECs your system generates. The owner had used that lump sum to help pay for the panels.

But then he sold the house to my wife and me, without ever disclosing the existence of the contract. So basically he sold the SRECs to both of us. He sold them to the organization by virtue of their contract, and he sold them to us, by virtue of the ownership of the panels transferring to us when we bought the house.

And the lien never came up because it was not against the real estate itself; rather, it was against the SRECs -- the Solar Renewable Energy Credits that are generated by the panels. They're not physical assets, but they do have substantial value.

It took the retention of a lawyer and a whole bunch of back-and-forth between us, the seller, our buyer and seller's agents, the lawyer who handled our settlement, and the SREC-buying organization to get the whole mess sorted out.

In the end, we got the lien and contract dissolved ... and then we hooked the panels up to FIOS, and I get handy little charts of all the SRECs we're generating :)


So, the solar panels conveyed with the house (as physical property), but the contracts covering them were never (re-)agreed to by you? And the lien wasn't against the physical solar panels?

I'd be curious how the state organization thought they had cause to pursue you?

How is that not the seller not lying about property (they didn't own the panels free and clear) and thus liable for effectively selling stolen property? In which case, state organization would be free to sue the seller, and you have no part in the proceedings.

What laws did it end up turning on?


You're correct that the seller never informed us of the contract, even though he was required to. Our state has a standard form that buyers fill out, indicating whether there are any encumbrances, liens, etc. that the seller needs to be aware of.

What the seller SHOULD have done is informed us that the contract would have to be transferred to us as part of the home sale, and we would have factored that into our offer.

But because they failed to inform us, they ended up selling the same property to two different parties.

And you're also correct that we were never a party to the contract, and that's basically what our lawyer spelled out in our demand letter. He basically said that this was a breach of contract between the previous owner and the organization (the two parties to the contract), and they needed to handle it themselves and leave us out of it.

Eventually, after a lot of back-and-forth, the organization agreed to not pursue this any further with us and nullify the lien. They had the option to pursue the previous owner for breach of contract, which would likely have involved him returning the money, but I don't know if they actually did so or chalked it up as a loss.


My guess is title companies haven't caught up to this - they've never had unreal/intangible property associated with real estate before.


Title insurers have not had to do due diligence for decades because record keeping has gotten so reliable there is almost never a problem. The underwriters just eat whatever occasional costs arise, and title agents make bank - at least on the East Coast of the US, out West they are typically one entity but the process is the same. Solar adds an interesting new wrinkle!


You would think hopefully that having title insurance. That the title company would need to do everything for hopefully no out of pocket cost.


Mineral rights? But that's probably specialized to certain parts of the US (upper midwest, southwest, NE Appalachia).

And water rights, but mostly a west-of-the-Mississippi thing?


Yeah, those would be things they might know about, but mineral rights usually DON'T convey (in some places they do) and water rights is usually a farmland thing.

I daresay this title company will now be on the lookout for things like this!


Water rights are very common for residential properties here in Utah. I currently live in a townhome, but several of the homes I had looked at had X "shares" of water from the reservoir in the mountains a few miles away.


Title companies are like an old folks home for politically connected lawyers. They do very little, and make enough money to spackle cash over their mistakes.


Did it end up costing you money to get that resolved? It drives me crazy that with housing contracts, they protect essentially everyone except the buyer. It's never anyone else's problem because of the contract somehow covering them (lawyers, agents, sellers, etc.), but somehow, the language never quite covers buyers.

We purchased a house, and within just days, the basement flooded because cleaning wipes were flushed down the toilet by the prior owner, clogging the system. So the sellers knew about the flooding or they flushed the stuff while moving out. It wasn't covered under any clause by interpretation, although it should have been by literal reading of the contract. Seller's can basically do whatever they want besides burn the house down.


We have a solar loan that we procured for our system. We're entitled to 100% of the generation, it's just a simple loan for the $40k system cost itself. But when we went to refinance, it was a pain in the ass explaining to the lender (Chase) that the loan has a lien against the SOLAR SYSTEM itself, and NOT the property. It shows up as a "second mortgage" in many systems, but it is EXPRESSLY NOT secured by the underlying physical HOUSE, just the solar panels and inverters themselves. We finally were able to get the lender to waive their positioning on the lien to clear the way for the refi, which was nice, but it was a _process_.


How can one sell their SRECs though? News to me. Have panels but never got credits for the SRECs. I do have net metering enabled.


Several states have programs that buy them at a price high enough to defray installation costs, to encourage adoption of solar. Part of the installer’s job is to do that paperwork for you, so you’d know if you had that option. You can also sell them to other states if you set up with a marketplace, typically SRECTrade. However, the fees and low sale price are prohibitive if you’re not in a state that statutorily buys them for a higher amount


This depends per country. But be aware that if you sell them, the buyer buys the right to generate electricity using fossil fuel, and sell them as if they are green/renewable. If your intent is to accelerate the transition to renewables, you should not sell them.


Are you still able to sell your SRECs with the system disconnected, or are you foregoing that as a cost of being offline?

I'd probably lose out on $500 per year from my system if I didn't sell them. At this point, though, it might _actually_ be asking the question of how much I value my privacy.


You can’t, because your production is verified by the data you upload to the SREC marketplace. You might even lose escrow you paid at installation.


If you sell your SRECs you're basically undoing any potential climate good your solar panel might be doing.


Of course, the SRECs and other subsidies are the only reason that there is actually a solar industry anyway. The economics of solar simply don't work and would not attract any investment at all if it were not for huge crony-ready subsidies paid by governments.

(FWIW, I'm not anti-solar at all, but I am a realist, and spent six years in the utility/commercial solar industry.


I'm all for subsidies, but I don't like SREC. The truth is our economies and populations are still growing, and so is our demand for energy. If we let solar supplement out energy instead of replace it, it's close to useless. Solar needs to be a pressure to reduce our carbon footprint, only then would it be tax dollars well spent.


That's not true because the purchase still raises the cost of buying credits.


Interesting - can you elaborate on this?


The idea is that you're selling your SRECs to someone (say a coal-powered bitcoin miner) so they can offset the pollution they're doing and not have to build their own controls. The argument is that if you did NOT sell, they would have to solve the problem some other way.

Just like if Tesla didn't sell electric vehicle credits to other manufacturers, those companies would have to find other solutions to comply.


While you're definitely not wrong, there are probably better vehicles for home energy users to meaningfully affect emissions. I make ~7 SRECs (7MWh) per year, and assuming I'm average, it'd take close to 100 homes not selling their SRECs for one year to offset an average coal plant for just one hour (http://www.energyjustice.net/coal/igcc/factsheet-long.pdf says the average coal plant puts out 667MW).

My state doesn't have time-of-use based pricing, but I do see this as one particularly effective way of shifting load (not installed capacity!) towards more emissions-friendly sources. The sun shines on home solar panels while we're not home; then absent any other incentives, we come home in the dark and do all sorts of energy-intense activities, like heating rooms and cooking dinner.


SREC mandates are used to achieve a percentage of renewable generation, not an emissions quantity.


It works out the same, someone doesn't have to build solar that would otherwise have to if the homeowners give them their credits.


Only if your base assumption is that everyone has to build solar, or would otherwise do it out of the goodness of their heart.

If your base assumption is that nobody is going to lift a finger for the common good, than the purpose of SRECs is so that a polluting Bitcoin miner pays you to put solar panels on your home, so that you get some good out of it and we don't all have an incentive to be polluting Bitcoin miners.

People who believe that everyone will naturally do the right thing out of the goodness of their hearts are often nicer to be around, but tend to get screwed by their inaccurate view of reality.


Yes, the end result is that energy generated in the state is sufficiently renewable per the government mandate, with homes contributing to it where it makes more financial sense than building greenfield solar or wind. In states where total demand is falling, conventional fuel plants are also likely able to be closed.


I think the OP's point is that it would be MORE renewable if the homeowners sat on their credits.


Yes, I can see that, but that would have to be balanced against the lower number of installations due to the higher costs to the homeowners. For many people, the payback period would be too long without the credits. As solar becomes less expensive, these programs will cease to exist, and already don't exist in many sunny states.


Something similar happened to me with solar city power purchase plan or something before they got bought out by Tesla


We just got a rooftop system from a very hands-off GC-type company and I'm so pissed off about it.

If I'd known in advance it'd have multiple Internet of Shit crapware devices attached that'll probably stop working in 5 years and require expensive replacement and renovation to keep working past that, and that it can't produce power at all unless those pieces of near-future landfill rubbish are happy, I'd 1000% never have gotten it.

Mea culpa—I guess I should have asked, but it just never occurred to me—but I'm fucking never getting another solar system unless I install it myself. There are other problems (that have solidified my usual preference not to let someone else GC a project for me, because they usually just skim money and do a terrible job—they fucked up our roof bad, for one thing, many thousands of dollars of damage so now we get to sort that out too, the hassle's not worth a few hundred in savings per year assuming all else goes well which at this point I'm sure it will not) but IoT shitware would have been a deal-breaker on its own if I'd known it required that to function, not just for monitoring or whatever.


It was paused but California was proposing Net Metering 3.0 rules in which you would have to pay $10/mo for each kw/hr of grid-tied solar panels. So for a small 5kw/hr panel setup (15 panels at 370w each), they proposed that new solar homeowners would pay $50/month to PG&E simply for having 15 panels. Most people I know have double or triple that.

Will Prowse on Youtube has great videos on off grid solar including equipment that is all UL certified. Getting over the hurdle of roof mounting panels or ground mounting up to code (seems to mainly be around using conduit for the wiring), I could buy a pallet of 10 solar panels, a converter/invetor all in one from EG4, and an EG4 5kw lithium iron phosphate battery for around $6000.

Battery: https://signaturesolar.com/eg4-lifepower4-lithium-battery-48... Offgrid Converter: https://signaturesolar.com/eg4-6-5k-off-grid-inverter-6500ex...


It's $8/month per-kilowatt and for PG&E customers the Market Transition Credit makes it $6.38/month per-kilowatt (at least initially). So you'd be paying ~$32 to PG&E for a 5kW system, not $50. It's really hard to know if the numbers CPUC settled on are really achieved the right balance, but they are trying to deal with a real problem here. There are a bunch of fixed costs associated with being attached to the grid that are currently baked into electric rates and since solar lets you avoid a bunch of electric charges (especially with the current NEM 2.0 rules) this represents a fairly large subsidy. As solar grows in share, the value of each additional solar installation falls unless its paired with storage to shift that generation to when there's the most net demand.

Now I don't actually know if CPUC actually stuck the landing here. The proposed rulemaking is a pretty dense document and it seems non-trivial to model costs from parsing that.

> I could buy a pallet of 10 solar panels, a converter/invetor all in one from EG4, and an EG4 5kw lithium iron phosphate battery for around $6000.

3.7 kW of panels (assuming these are 370 watt panels like you mentioned above) and 5 kWh of storage is not going to go very far in an off-grid system unless your electric use is pretty low. The thing about going completely off-grid is that you need a substantially larger system than in a grid-tied scenario both in terms of generation and storage.


I hope you're not too upset about that - there's a pretty good justification for it: It's both expensive and valuable for the utility to be there for your burst needs beyond whatever your panels generate at any particular time. But with net-metering by itself, there's no way to charge you for that expensive value if your net pull from the utility is zero, even though you might pull a lot at certain times.

You could go off-grid to avoid this, or install your own batteries, but both of those are obviously expensive and inconvenient in their own ways.

I'm not saying they got the price right ($10/kWh) or the model (more for larger arrays), but the overall principal is fine.


1)This sort of thing is covered by connection fees independent of usage fees

2)People providing green energy into the grid should not be charged more than others.


If they are providing "green energy" into the grid, are they doing it for free or is there some kind of financial consideration for them? If so.. are they actually being charged _more_ than others who simply use power from the grid, or are they receiving an entirely different class of service all together?


Then charge peak price for peak demand.


Everyone (including non-solar customers) is getting moved to time-of-use billing with some sort of phase in schedule and I believe it's already required for new solar hookups. Part of CPUC's proposed replacement for NEM 2.0 combined fairly agressive TOU billing with moving credits for grid exports to an "avoided cost" model so that you get very little for exporting in mid-day when there's currently an excess of solar generation and much more in the late afternoon when there's a big mismatch between demand and renewable generation.

I think you can definitely argue about the specific values they chose, but the broad strokes make sense.


After going with Tesla panels, this is my plan next time. I could have had 3x the system for 1/2 the price. At that cost/ scale I could truly be off the grid.


Funny thing I look at Tesla and my system and I’m seeing that my all in price for my panels + 3 powerwalls in 2020 is within 10% of what people in my town talk about paying for panels today. A lot of that cost gets eaten by installation and inefficient practices.

I’m kinda with you on building my own off grid system - but, I don’t think I would’ve had the know how of where to start without the knowledge I got from getting my Tesla system installed.


Yep. I'm also planning on bumping the size by a pretty large amount. I love my tesla system, but I should have gone about 50% larger in both battery and panels.


The electric utility still has to pay for grid upkeep. Most of their costs in fact are grid upkeep. Until you go off grid you need to contribute to that. The status quo in california is that solar owners get to use the grid as a free limitless lossless battery, and that was just never going to work. It means the rich are being subsidized by the poor.

And tbh if enough people go off grid I expect the state to enact a "grid tax" that every homeowner has to pay, regardless of if they're connected, because otherwise electricity costs for on grid users would skyrocket as fixed costs are spread between fewer users.


> And tbh if enough people go off grid I expect the state to enact a "grid tax" that every homeowner has to pay, regardless of if they're connected, because otherwise electricity costs for on grid users would skyrocket as fixed costs are spread between fewer users.

Electricity in CA is already 3x the rate of rural Wyoming or Arkansas or what have you. Why exactly are their fixed costs so high, and why is it the responsibility of individuals capable of generating power more efficiently to subsidize that ineptitude?


@Syonyk has some good stuff here: https://www.sevarg.net/tag/solar/ - I'm sure he'll comment on this eventually, but he has built his own solar in Idaho.


I skimmed most of his off grid article and he states this:

"If you want to get a rooftop solar system that powers your home with the grid down, you can do it! The hardware is out there. But such a system will be significantly more expensive than a normal grid tied system, and it will likely never “pay off” in terms of money saved."

That may be true if he was paying $0.10 kw/hr in Idaho, but that's definitely false for a few reasons.

-In California, it's $0.43 kw/hr so the ROI for solar+battery is getting much faster with the price of panels, inverter, battery coming down. Price of electricity will continue to go up in the short term.

-A lithium iron phosphate EG4 5kw/hr battery for $1600, or even an EcoFlow Delta Pro for $2500 is a good deal considering around 8000 discharge cycles

-Most critically, net metering (grid tie) means that all that excess solar power you are producing in the middle of the day is being bought back by the power company at CHEAPER rates than what you are paying, so that needs to be accounted for in grid-tie vs off-grid pricing. If you are off-grid producing x amount of your power needs and you are using 100% of that power via lithium battery (at night), you are getting full retail price for producing your own power at $0.43 kw/hr or $0.49 kw/hr peak rate from 5-9pm in California. Off-grid doesn't mean truly off grid if you already have electric hookup, it just means an off-grid panel for some or most of your house. Think of it like this... if you have 15 solar panels grid-tied, you are being "taxed" ~4 panels by your power company.


> In California, it's $0.43 kw/hr

PGE is ridiculous. But apparently whatever serves San Diego is even worse.

There's Silicon Valley Power, with rates that are a fraction of PG&E's, smack in the middle of the Bay Area. Can get as low as $.10 Kw/h during off peak hours.


BTW, in California the acronym PG&E is for Pacific Gas and Electric

Confusingly when I moved to Oregon years ago, its PGE (without the ampersand) for Portland General Electric.


Often the gas part came before the electric and some of the companies still have names that imply that.


That would be ... SDGE!


Yeah, the off-grid article is from almost five years ago now, things have changed, and power companies are more cognizant of it, too.


I totally feel your pain and would feel exactly the same way in your shoes. This also ties into the broader problem of contractors treating their clients horribly because their services are so in-demand. Basically, contractors can show up, do a horrible job, and they will completely disappear on you as soon as they get their money.

There's so much demand in their industry that all the best people are working on high paying corporate job sites (like building hospitals and new FAANG campuses), that the average homeowner is left dealing with garbage companies who treat their clients like shit.

Hopefully rising mortgage rates and falling house prices can cool demand a little and we can get some decently competent contractors again.

WRT solar, I was super impressed with jerryrigeverything's solar installation that he did himself. Basically you can get a kit that you install yourself and it's much cheaper than having a garbage contractor rip you off and ruin your roof in the process.

His video: https://www.youtube.com/watch?v=jSa1tvrrFZg


If you can find a contractor you can trust, even if they're not versed in the installation of solar, you can work with them. I'd much rather have an electrician and roofer that I trust work on a kit like that than depend on some random contractor I don't know.


Right—contractors are fine, it's GCs that (mostly—I'm sure exceptions exist) blow.

I'd rejected GC bids for things before because they were crazy-expensive and they were proposing mediocre work despite their very high rate, then gone on to basically GC the project myself, engaging contractors for most of it and DIYing a few things here and there, and those have all been great. My mistake was making an exception this time, I guess, but OTOH the savings from solar also wouldn't have been worth it if I'd had to do that, even with a hypothetically-cheaper installation cost, so I'd not have done the project at all in that case.

As it is I don't see how the system won't end up losing us money by the time everything's accounted. Probably 5-10 years off our roof lifetime even if they fix all the acute damage but don't do a full replacement (I very much doubt they'll do that without us spending more money fighting for it than it's worth), plus it'll cost more to replace when it's time because of the damn panels.

But they didn't like our proposal that they take it all down (which they'll have to do to fix the roof anyway—it's bad), fix our roof, take all the equipment back and zero out our bill so we can all pretend this never happened. Sigh.

Been up for weeks and still hasn't generated a single goddamn watt of power so far, either. The IoT shitware refuses to do its job and of course they're not in any hurry to get out and fix it. Pretty sure it's gonna end up with attorneys involved, which is something I've managed to avoid my whole life so far and I'm dreading.

Blue Raven Solar, for the record. May as well name-n-shame, but my understanding is this kind of thing's extremely common in the solar industry so it's not just them. I suspect heavily-astroturfed reviews, in their case. Again, I really should have known better.

One point of advice for others: call out a roof person you trust before starting one of these projects. Ours, it turns out, would have told us our roof was too steep to safely install solar on without totally fucking up the shingles in the process, and that would have been the end of it.

[EDIT] Heh, incidentally I think the near-impossibility of putting the panels back up without wrecking the roof again is what might save us and get them to back out the whole project. IDK how they're gonna re-install the system after the roof's fixed.


The problem is that GCs have the same incentive as real estate agents. It's more profitable to do 2 quick jobs (even at 80% of top pay) than 1 quality job (at 100% of top pay). Quantity > quality.

Lesson learned from watching my parents build a home & personal projects: treat any payment of funds as a sign off on the work and expect you'll never hear from them again.

If a GC baulks at 1/2 withheld until completion/inspection, move on.

If you have any concerns about the quality of work, refuse to pay until it's completed.


Yeah, a good GC is either so astronomically expensive you can't even talk to their lawyer's cat, or simply not to be found (or you marry one or into the family).

I would get legal involved sooner rather than later; but a good lawyer can be as hard to find as a good GC.

Maybe you should just get the roof fixed and sell the house and try again.


> Maybe you should just get the roof fixed and sell the house and try again.

LOL, the "unplug it and plug it back in" of major house mistakes, I guess.


Real talk, if it isn't obvious foundation damage and the roof isn't currently on fire, it'll probably sell just fine.

You can hire a home inspector after it's "fixed" and see what he says. Just disclose that you have solar and that you had an issue and it was resolved, provide paperwork, and most people will yawn.


Side note: home inspectors are typically lazy GCs who realized it was easier money.

They're not actually liable for anything more detailed than what you could notice yourself, which... why are you paying them again?


When buying houses, I've really only hired them because I don't want to get on a roof myself if I don't have to :-) Or crawl around in an attic, for that matter.

Most folks are pretty bad at spotting problems with houses, because they simply don't know what to look for (and why would they? It's hard to know everything), so an inspector's probably a good idea.

Me, I spend the entire time at an AirBnB distracted by noticing every single place work wasn't done quite right, every cut corner or lazy shortcut, every interesting (or "interesting") material choice, "oh, looks like they replaced this shower insert but didn't want to fully redo the drywall so took the lazy route and added trim", "uh oh, this tile shower floor has serious drainage problems", "interesting, this doorway must have been added", et c., but most people aren't like that.


Yeah, you have to find the home inspectors that the realtors hate - those are the ones who may actually do something useful.

But if they don't see anything wrong, then any other ones won't see anything, either.


It would never have occurred to me, either! There was certainly none of that crap attached a decade ago when I had solar installed on my previous house. Glad to know this is something to screen for - but chagrined that it exists, naturally.


Whenever quoting an install of basically anything, ask for the list of part numbers, and then you can look them up yourself.

My furnace and AC have no internet connection; the thermostat I installed myself does.


Why did you choose an internet-connected thermostat?


I got a discount on it and a discount to have it (which doesn’t seem to actually change what it does) and I can change it from my phone.

None of which is exceptionally amazing and if it annoys me it gets replaced with a mercury switch.

Since I’m WFH and the house is always occupied and it has one zone none of the supposed energy saving stuff really applies.


> the manufacturer states that not having it connected to the internet voids their warranty

I was aware that my inverter could connect to the Internet, and probably would be connected to the Internet during setup, but my full intention was to disconnect it following the install. I was pretty happy that the one that the vendor gave me was poll-able from an internal Web server, so I could scrape the metrics myself.

It turns out, none of this was to be. I found out about the Internet connectivity requirement for the warranty _after_ it was on my wall and producing power - and when you've spent five figures on a system with a 25-year warranty, unfortunately I had to suck up my pride and let it phone home to the manufacturer's cloud. It now sits on a separate VLAN for "untrusted" devices.

What's more, I also found out that the manufacturer had intentionally removed the local web server feature, instead making it totally reliant on the cloud service for metrics. It has a laughable rate limit of 300 calls per day, which gives me roughly 5-minute granularity (I could get better granularity if I made my metrics-gathering scripts poll differently overnight, but still), and I really dislike having to reach out to a vendor for data that's produced in my own home.


I see you also have a Solar Edge inverter. I had no idea about the warranty, I didn't read the ToS because I wasn't asked to agree to one.


How does this even make sense, if the power goes out or the internet goes down your warranty can be voided? At least I know what question to ask when I go shopping for these…


This would obviously not fly in Europe.

I have setup my SolarEdge system 4 years ago, (HDWave with an LCD) and never did I sign anything that stated 'must be connected to internet, or else', again not even possible with European warranty laws anyway, as the vendor has to prove that the malfunction occurred due to not being connected to the internet.

When I was doing research on inverters and the control units, I looked into a few systems. Back then, you had either string based (Sunny boy, GoodWe was what was standard offered) or per panel inverters/optimizers.

Some vendors of those setups even wanted to place an (additional) router! Maybe to ensure the wifi signal from the inverter was close, though they wanted to place it in the meetering cabinet. No way of course. Are you crazy?

The Enphase control boxes where the worst, that still stick with me today. They actually setup a VPN to their mothership, and use that to be always connected, completely unacceptable of course. Just think about it, joe schmoe gets a solar setup, happily lets it connect to the internet, and then exposes his entire home network to Enphase.

SolarEdge was the least worry some, as (after research) showed they have a USB/RS-232 and RS-485 internal port that can be used (https://github.com/jbuehl/solaredge) and I've been using that to scrap the data for years.

I do know that with the newer HDWave (without LCD) things are a bit trickier, but should still have those RS-485 connectors, as those are needed for other purposes. Of course, I don't know what software updates broke over the years.

My system has _never_ connected to the internet, and is happily humming along :)

All in all, this is completely unacceptable behavior, and I'm surprised that they keep on getting away with this. Installers think its great, they show up to give you an offer, and show how great some of their customer setups are functioning, because yes, they also keep having 'full admin' access. Insane.


I'm slowly working to start a solar-install company — your post helps confirm my belief that offering a private no-internet option... is a worthwhile differentiator. I have a lot to learn, but the Sol-Ark inverter or Outback (Radian) inverter might be the way to go. Otherwise, it might have to be some obscure inverter without reliable warranty.

Here's a related forum thread: https://diysolarforum.com/threads/sol-ark-needs-a-serious-ch...


I would bring up all the things in this post in your advertising. Installation suitability reports about roof types. Functions without external configuration, off site monitoring, and shortened lifespan due to outdated software. The purchaser owns the equipment so not a permanent lease against the property.

Just creating awareness should drive a lot of business.

I think some variety of IOT is valuable, but it should have a web server backup that provides both html interface and raw data which does not require dedicated apps.

There should definitely be a disconnect from the grid option for the extended wind related power outages of the Northern California.


I don't think thats enough of a differentiator for people to head to your product, no offense. Solar price, payback terms, incentives, green ambition and neighborhood also putting up solar are probably the biggest drivers. The amount of people who would want no internet connection is incredibly niche.


I'll be installing in East Bay, California. And I'm only planning to install one or two systems a month. ...I think I can find my little tribe of privacy people. But, yeah, I'll have some other differentiators.


That thread has more bullshit than a farm...internet connections back to China = EMPs? Those people are out of their minds.


Later in the thread, there was discussion around using inverters in offline mode. Sorry for throwing the whole thread at you.


I work in the industry on the commercial side. Access to that data and the ability to sell a homeowner's power generation and storage out from under them to the utility (double-dipping, because often the homeowner pays for the systems) is a big thing right now. If you buy solar or batteries and the company wants you to constantly be connected to their network, then they are somehow making money off of you, much like Sony will charge you for a TV then put ads on it.


> Access to that data and the ability to sell a homeowner's power generation and storage out from under them to the utility (double-dipping, because often the homeowner pays for the systems)

Can you elaborate on this? How does this work?


First, the data is helpful in aggregate to learn about homeowner power usage and sell and stuff. Second, these companies are often selling some sort of contract or carbon token, or some sort of virtual aggregation of power to a utility. For example, Tesla is famously paying huge amounts to sell their customer's power to SCE right now in a pilot program. I can tell you from experience, the prices they're paying their customers per kilowatt-hour are way higher than the market rate for that sort of thing, so it's not going to last. Eventually, these Tesla customers will be stuck in some contract, probably, selling Tesla their power for pennies on the dollar.

None of the VPP software running in the cloud cannot be run locally. What we should have is a small computer optimizing homeowner's devices (solar, batteries) on-site, but instead all that crap is in the cloud. It's "Generation as a Service" at this point, and it's bullshit.

What we need is an open-source building energy management supervision software with a diverse driver and plugin ecosystem. I want to write it, but I have a family to feed.


> What we need is an open-source building energy management supervision software with a diverse driver and plugin ecosystem. I want to write it, but I have a family to feed.

I work in the industry too, and would love to see something like this -- not just so homeowners can maintain first-party control over their assets, but also so there can be a "reference design", so to speak, with best practices and peer-reviewed code & data tables, etc. that any manufacturer can use (instead of crappily reinventing the wheel every time).

Have you ever spoken to your coworkers about this? I tried to bring up open-sourcing some of our stuff at work and got a lukewarm reception... people were cautiously interested but wary of navigating the bureaucracy (especially legal).

But still, it's SUCH an important field -- energy security is national security and climate policy rolled into one -- that I wish there was more of a coordinated effort to make these things better.

NREL used to lead efforts like OpenPV and that was a great start, but then the previous admin shut it down :( It should probably be a citizen-led effort and not subject to political whims...


Yea, none of the companies I have worked for would have gone for anything like that due to the tight coupling between their code and algorithms. The thing is, the algorithms to optimize these things don't have to be that complex, a simple solution with if/else statements would get you a TON of energy savings and optimization of time of use.

I do plan on writing this software, but I am not sure when. I've been considering taking a year or so off (if I can save enough, or perhaps if my company dissolves or I am fired) to write a framework that you could add drivers for hardware to and add your own algorithms to. Basically, it's just a simple loop of "read telemetry from devices -> save telemetry to local database -> data processing steps (ELT/ETL stuff) -> optimize the use of these devices for some metric (energy use, peak power use, comfort, etc.) -> store results -> send new commands/parameters to the devices -> sleep(interval)" and then adding hardware device interface drivers and algorithms in there.

Edit: If you need to aggregate devices it's a bit harder, but I think it could be doable with distributed algorithms or with open-source interfaces between the utilities and individual customers. I am not sure of how that looks right now, but I prefer a world where millions of individuals run their grids the way they want (within safety regulations) and then they can sell their excess power to the grid via a standard protocol. Right now we're doing a lot of vacuuming of data that isn't necessary, since the data can easily be stored locally for these applications, with a distributed protocol. I actually think this is an application that could use blockchain in a neat way, but I would prefer to avoid that, if possible, due to the implications.


> It's "Generation as a Service" at this point, and it's bullshit.

Wow that really is bullshit! Is this a US thing? My solar equiment is german... I think. Vendor is Fronius. There is a locally hosted monitoring web app, it can even create a wifi access point if necessary. It apparently works fine even without the network.

There is a cloud service too. It's meant for me as an end user but the solar company that I contracted took over the account I created in order to "provide remote support". I really have no clue what they're doing with it but I can't figure out if they're selling anything I'm generating.

Where I live (Brazil) the solar energy I generate gets converted by the power company itself into kWh "credits" which will offset energy consumed from other sources. These credits expire if unused so I have zero incentive to generate more than I consume. Not sure if this is the same as the carbon token you mentioned.


Yea, I am in the US, so I know mostly about stuff here, but I also think the US tends to use "* as a Service" more often than EU, right? Inherently, there isn't anything wrong with internet connective services, such as if your solar company is using your data to actually optimize the hardware on your home site. However, most of these today are middle-man cash grabs, IMO.


I read a story about a solar installer that blew my mind. The customer posted a negative Yelp review about the installer. In retaliation, the installer remotely deactivated the customers internet connected solar panels and held them ransom until the negative review was removed.

If the installer can do that, so can the analytics company and probably any competent hacker. I don’t think security is top of mind for this industry.

So, no internet connected solar panels for me.


I've skipped out on products _specifically_ because they have an app. I see apps as a liability, not a feature. Give me dumb products please. My air purifier doesn't need a god damn app, just an on/off switch with some LEDs, please!!!


When I got my security system in my home installed, I asked the sales representative about getting a key fob to arm/disarm the system. Her response was basically, "Oh, haha, you're too young to be asking for one of those. You can just use the app!"

Right, because I really want to have to take out my phone, unlock it, click an app icon, wait for it to load and authenticate my credentials (assuming I'm not forced to sign in -- again), find the arm/disarm button in some cluttered UI, and then click the button. As opposed to just pulling out the key fob and clicking the arm/disarm button? No apps, please.


Not to mention that the whole app thing will rely on a server and a functional internet connection, so you can forget about interacting with your alarm system if there's a cloud or network outage.

I feel like modern-day developers have effectively forgotten how to do local network communication. I know about NAT traversal and similar problems, but you can still architect your system in such a way that your cloud backend is a dumb NAT traversal helper/proxy, which then makes it trivial for your client to offer a cloud-less, local connection option for those users that know how to forward ports/have IPv6/etc (or when you go out of business and your cloud service goes offline).


Did you end up being able to get a fob?


Yes, for like $60 or something ridiculous like that.


That is total BS, my condolences to your wallet...


Agreed. Anything with an app is a liability I want to steer away from.

Not only due to the obvious security, privacy and availability problems, but because the lifetime of apps is very short in the grand scheme of things.

I'll bet there is zero chance a phone app for a solar install today is still going to be working in the 25 year lifetime of that solar panel system.


I have an 8-camera home surveillance system. It really wants to talk to the mothership in China, but it will work without it.

However, the only interface to its control box/recorder is through a plugin for Internet Explorer 7. I have to access it through a virtual machine running Windows XP.


SMA inverters (at least) do not need an internet connection to be configured. They of course support ethernet but it's not mandatory at all. Everything is self hosted and they can be run air-gapped


It's a "me too" situation. One company releases appliance X with IoT features and all the competitors panic because it's something they don't have that consumers will see as missing, so they rush to throw something together. This kind of herd mentality drives a hell of a lot of feature bloat.

Problem is that consumers really do see "internet/mobile connected" as a desirable feature even if they're just going to use it once for the novelty and then forget about it.


Some of these features probably lead actual usage. If you'd asked in 2009 how many people wanted to adjust their air conditioner temperature remotely, command their vacuum to clean, or see video of their front door when the delivery driver drops off a package, I'd bet those numbers would be low, but we today have IoT poster children like Nest, Roomba, and Ring making those desirable features.


> Some of these features probably lead actual usage

This is an interesting comment and from my experience I'm sure it's often true.

Around 2008 or thereabouts I was the lead engineer on my company's first foray into what is now called IIoT: Industrial Internet of Things. At the time we just called it Remote Monitoring. There was (is?) even an magazine by that name.

We built large, complex medical instruments that required periodic servicing both by the customer and by our Field Engineers. Engineering's value proposition for the project to the Technical Support/Field Service group was that we could reduce the frequency of expensive and unscheduled field service callouts and provide considerably more debug information from an instrument when a customer called in with a problem. We should be able to reduce diagnosing an instrument failure from hours to minutes. And in the future, we'd probably be able to tell when a consumable part was due for replacement and schedule it ahead of time.

It was a freaking battle uphill both ways in the snow to get them to accept that. All they wanted was a slightly faster version of the tool they already had which was already inadequate for the task. They couldn't actually stop us from implementing this, but there a a strong reluctance to the point of "well it sounds nice but I don't see how it would be helpful to us."

Postmortem:

Fast forward two years. The system is in the field. Tech support keeps telling us how they can't imagine that they were able to get along without it and keep asking for new features!


Perhaps if you asked people randomly. If you'd asked the people who were already setting up home security cameras if they'd like phone alerts, I'm sure they'd have said yes.


I have a Solis inverter [0], while it has WiFi built in, and we have the data logging dongle, it doesn't need to be connected to the internet to function, and you can configure the dongle to talk to MQTT locally (but I haven't done this).

It doesn't need to be connected to the internet during configuration either, we set it up and got it going before connecting the data logger to it.

That said, I do use their portal to see my generation and consumption. It's pretty handy. But I also do scrape their data and store it locally as well.

[0] https://www.ginlong.com/rhi_3pinverter1/29118.html

* I'm not affiliated with them in any way, it's just the one we have


I read the "warranty" part, thought "no way", but sure enough Enphase says in section 5 of their warranty: https://enphase.com/download/2022-04-30-enphaseenergy-microi...

>The Covered Products should be continuously connected to the internet during the warranty period, except where interrupted by causes outside of the Covered Owner’s reasonable control. This will help ensure that potential defects in the Product can be diagnosed remotely and that the Product can receive over-the-air firmware updates.

Interesting use of "should".


I have a Fronius inverter which has both WiFi and Ethernet - but works perfectly well without it.

My installer said that they need access to the inverter during set-up in order to configure it properly, but after that it could be disconnected.

Personally, I use it to get firmware updates and to use the API in order to monitor it.

Although, that said, I'm not sure what sort of security issues you're imagining. There's no privacy issue - the inverter only has data on what you've generated.


One obvious one is if the panels can see the load to the house, they can determine occupancy.

There are devices that can monitor your breaker box and tell you when appliances have turned on.

You could perhaps prevent this by having it dump everything between the meter and the house so it can't tell if it's going to the grid or not. I don't know enough about it to say if that's reasonable.


As far as security - Fronius inverters (and other manufacturers) have a built-in API for remote shutdown / remotely disconnecting the solar power at your house. Some countries have so much household solar power that the solar feed-in can overload the power grid. Governments have required the ability to remotely deactivate a user's solar system to prevent grid overload & maintain grid viability.

Here's a Fronius document on the remote disconnect API setup:

[PDF] https://www.fronius.com/~/downloads/Solar%20Energy/Quick%20G...

"This Application Guide describes the procedure to setup a Fronius inverter system to comply with the SA Smarter Homes regulation of remote disconnect/reconnect when selecting the “SA Power Networks / Fronius –Fronius – API control of internet connected Fronius” Relevant Agent option."

Here's a Fronius webinar (audio with slides, hour long) discussing the remote disconnect functions, jump to 2:42 if the link doesn't automatically take you there. The first couple of minutes should give the overview:

https://youtu.be/KR8jxANwii4?t=162


I guess they're more worried about what degree of control their cloud has over the device, and since it upgrades trough the Internet, I guess they have potentially absolute control. The idea is minimizing the risk a device like this would become if compromised.


My parents have an internet controlled heat pump. It works fine without, but you basically can't change the settings locally. A couple of years ago they bricked it, charged us 400€ for 'repairs' which did jack squat, then it magically started working a few weeks later.

Don't buy internet connected appliances.


Three years ago I bought a small window air conditioner. It came with the usual front panel controls and a remote. The remote turns it on and off and lets you set temperature and fan speed, which is all I expect of a remote. However, it requires an Apple or Android smartphone, their phone app, and internet access to access "extra functions", none of which were described in useful detail in the sparse documentation that came with it.

No, I'll just use the remote.

An acquaintance was telling me about the new RV she and her husband bought. It has a brake controller for trailer brakes. Not unusual, I have them on two of my vehicles; there's a big lever to manually apply the brakes and a "gain" control to adjust how rapidly they apply. But her RV's brake controller is only controllable through a smartphone, apparently.

A few years ago I went to the store to buy a new alarm clock. All the clocks they had for sale were just Bluetooth accessories for a smartphone. Not particularly useful when there's no smartphone to link to it, and ignoring the question of why you wouldn't just use the phone's alarm instead.


I just keep away of anything requiring an app or internet connection. I just can't wait until this IoT hype comes to an end...


Sure. But that also has to be weighed up against the risk that the device has a firmware bug which will render it inoperable.

Or that a potential fault won't be identified and the owner won't be alerted.

There are risks on both sides. But, in the UK at least, there are several inverters which don't require an IP address.


This is why I don’t want 5G. It is not going to make my cellular service better or cheaper, but it is going to enable every internet-of-shit device manufacturer to buy bandwidth from telcos and a new generation of “smart” devices will be able to call home without asking you for wifi password.


I think low speed protocols like LoRA or relatives of it are more likely to be used for this. They're very low power, low overhead, and hard to block.

That's exactly the privacy endgame though: stuff a chip with cellular capability into everything you can and stream data to data brokers for double-dip profit whether the customer likes it (or even knows it) or not. The market for data about people seems to be almost infinite and bottomless.

This can only be fixed with legislation. I predict it will become a major battle in the future.


If I understand correctly, LoRA is more suited when you want to have a fluid decentralized networks, when nodes come in and out all the time and you aren’t concerned by latency.

This is the opposite if what telcos want. They want to have a low latency connection to every device in the house. Imagine your toaster and your tv stream in real time to an ad server what your are doing. You put a piece of bread in the toaster, while you wait you walk to your TV, start using the remote and you have a peanut butter commercial on Hulu. And this link inside the consumers homes is handled by 3 US cellphone carriers, not much competition. Or maybe it is just me being paranoid.


What does 5G have to do with that? There have been devices using 3G for years. It is very cheap and the infrastructure is pretty much everywhere.


A fully deployed 5G network will be able to support a much higher device density. Up to a million devices per sq km. Basically an antenna on every block and they will be able to connect every IoS device and scatter around a bunch of sensor, and still have some bandwidth left.


> There have been devices using 3G for years. It is very cheap and the infrastructure is pretty much everywhere. --- My Resmed CPAP machine came with a cellular tattletale. It was a separate module with a cable to the mainboard. I unplugged it. The DME provider got all pissy about it, but I just ignored them.

I'm guessing the main reason it was a separate circuit board was so they could install a compatible cellular module depending on what country it was to be shipped to.


5g deployment is going to be cheap enough for manufacturers to buy bandwidth and cut the consumer out of the decision entirely. They’ll get the user data for the cost of buying the bandwidth. And it’ll become a de-facto standard you won’t be able to buy a device without it.


In the "5G" bundle of technologies, one of the things was specific features for connecting larger numbers of low power devices very cheaply. Think almost every >$20 gadget in your home having its own GSM modem.


Depends on where you are, but 3G's being deprecated.


When I set up my Fronius inverters, I did connect them to Ethernet, but disabled any cloud features and set up a simple script that fetches the stats I'm interested in and stores them to a local influx database.

I didn't go as far as isolating them from internet/rest of my network- I am mainly concerned about them losing my data/going bankrupt/trying to charge for "cloud" features.


Some places have a legal requirement to do this; I think in order to be able to protect the grid (not quite sure why the smart meter can't handle this though; maybe to do with frequency matching or diagnostics)

But the main reason they do this is pretty mundane - LCD panels often fail after a few years in the kinds of environments inverters are often installed in (hot, sunny). So they move the display to be an app instead, and you need an internet connection.


I'm not 100% sure, but I think i read it was mandatory/required by law to connect your inverter to the internet in The Netherlands, because they need to be able to "turn off" the inverter in case of emergencies when the power grid becomes overloaded.

The part i'm not sure about, is whether this system depends on the internet connection, or whether there's some other wireless system build into these things.


Definitely not paranoid. This is the time to get ahead of this.

For distributed energy scenarios in which various sites operated by different administrators (my_neighborhood_solar, big utility, your_personal_site) may exchange energy, then how do we securely design the communication between the smart meters?

For example, perhaps dedicated SIMs would get the traffic off our home networks, but the resultant network would need security.


No, you're a totally sane person. People often mistake "smart" with "automated". I'm all for automaiton, but not for devices that think, decide and do things without me. And also the main question is "who's in control". So, you're not alone.


Thanks for posting this, because it would not even have been on my list of questions to ask. I am considering a solar conversion, but an IoT connection is a complete deal-breaker for me. Now I know that, not only do I have to ask about it, I have to ask a lot of follow-up questions.


IoT does not make sense for residential applications, except as a kind of gimmick.

As for other domains like, industrial, manufacturing, agriculture... that ship has already sailed. HN would do well to remember there's more to IoT than WiFi lightbulbs.


It absolutely makes sense.

Knowing how much capacity you have, monitoring intake, output, charge rate, battery temp, etc. are very important things to know and keep track.


an LCD display panel isn't enough for that?

I'm on mains electricity, but off grid for gas and water.

If I want to check my water levels I.. tap the water tank. or if I am feeling adventurous I get the ladder out and unscrew the top.

If I want to check gas levels, I see if the little mechanical indicator has rotated to the red plastic section yet.

I'd hate for any of them to be reliant on some hobby/residential grade IoT set up since they're so simple to check the old fashioned way.


Geeks who pay attention to that sort of thing are also the sort of people who could do it locally...


Networked devices make a lot of sense even in the home - my Home Assistant is pretty damn useful. The problem is the 'internet' part. I'm moving everything to the local network, but it's getting increasingly more difficult. To the point that I'm avoiding wifi devices (in favor of zigbee and possibly bluetooth).


An old-timer I used to work with once told me, "A warranty is only as good as the company selling it." There is more truth to the saying then I want to believe. Many of you have proved this by your comments.

In many states the "contract" is invalid if it wasn't presented as a requirement in the sale or when it was agreed upon. It doesn't matter if it is in the TOS if the TOS isn't signed and dated by the person paying for the Internet access. Legally you have a leg to stand on. You would need to get a lawyer and possibly go to court. Then trust them to honor the warranty later. That sounds like fun doesn't it.

Good luck.


These solar companies are gonna pop up, do a bunch of installs, and then disappear or go bankrupt.

And the manufacturers will say you have to deal with the installer, and the installer will be gone.


Completely agree.

That said we shouldn’t reserve the “IoT” moniker for crappy cloud based things requiring constant internet connections to work. That’s the internet of shit (as popularized by https://mobile.twitter.com/internetofshit)

I definitely want some connectivity in a solar panel setup. But in the case of a solar panel array there is no point for that connectivity to use be more than local, or offer more than simple monitoring. A local polling api that offers a simple json blob with stats? Does that qualify as “IoT”?


Internet of Turds seems to involve some worthless "cloud service" that they try to charge you for.

It is sometimes accompanied by a local API that works without internet (Apple HomeKit stuff seems to somewhat work sometimes in an internet outage, and the Hue buttons keep working).

It can be quite hard to distinguish the two.


The HomeAssistant project has pretty good docs on which devices use what class of communication e.g “local polling” should be the minimum for any device I get (regardless of whether I want to use it with HomeAssistant)


I'm considering a GoodWe inverter supported by HomeAssistant, it would be good to know if those can be functional without internet connectivity. Anyone knows the answer?


no.


No as in… you think “IoT” means gadgets with cloud connections?


I have a Huawei SUN2000 inverter and the network dongle is an optional extra, which is not required once the initial setup is completed. During setup you need it to provide a WiFi connection to the mobile app, but the installer could use their own dongle.

When I had my solar installed I had no internet connection at the property, so the installer gave me instructions of how to pair the inverter when I did. I had a smart meter which gave hourly data, so there wasn't really much need in connecting the inverter up. It has indicator lights to tell you when something goes wrong.


Mobile apps only work for a few years. Apple requires that apps are updated at least every 3 years.

i.e. after a few years, you will not be able to re-setup your inverter again.


Being delisted is not the same as unable to install. You can still download previous purchases, even when the app was delisted.


That a good distinction. Thanks.

New home owners would not have the app previously installed, so it would still suck for them.


He can run an android emulator on pc/mac.


Most systems that are connected to the grid require it by local code from what I've read. When the city did the inspection on my system after the contractor finished up installation, one of the things they checked is that it was on an approved list of inverters. It's not clear to me that the approved list required that those inverters be internet-connected, but I wouldn't be surprised if there was a requirement about "off site upgradeable/diagnosable" and so on.

If you're off grid, you can easily get away with not having that sort of thing.


Will Prowse has DETAILED answers and breakdowns for you: https://www.youtube.com/c/WillProwse


I don’t even know where to begin.


My solar panels report the amount of electricity generated so I get my SREC credits. These are direct payments for generating clean energy.

What frustrated me was that the company used a meter that reported over 3G, and then wanted me to pay $500 to upgrade my meter even though it was under warranty. They eventually found an ethernet port in my current meter, but they still charged me to run the cable. Fortunately, they got the cost low enough that it wasn't worth arguing.


Nope, you're not. I solved the problem of nosy IoT-things by creating a VLAN specifically for these things where they get to spy on each other without being able to phone home. The inverter connects to the VLAN through an untagged port on an access point using an Ethernet cable (no WiFi for this type of application, please) which makes it possible to access the web features for those who have access to this VLAN. I proxy the management interface to the outside world through a reverse proxy, this way I can access all features on the device without it being able to phone home or auto-update or do whatever else it could do if it had internet access. I'm using a Fronius inverter which is an Austrian brand, I specifically did not want to have a Chinese device for this application. Updates can be manually uploaded when required, configuration can be done through the wired Ethernet connection, it is not necessary to use any specific "app" (which, of course, is available and pushed but it is not necessary).

I installed everything myself - built the barn on which the panels are mounted as well - so I know what is where. The only net-connected part of the solar installation is the inverter, once I hook up battery that'll probably have a network connection as well so it can join the fray on the IoT-VLAN.

This solution with a separate network - VLAN or physically separated is up to you - which is reverse proxied onto the wide world gives the best of both worlds in that you remain in control while you still can use the networked functions of whatever it is you're dealing with. You won't be surprised by updates, the equipment does not get to phone home, nobody gets to data mine you.


No, you're not the only one.

It's ironic to see on this same HN front page a story about police & face recognition. We should oppose all forms of social credit / new world order / cashless / corporate fascism.

It shouldn't just be "the police are bad, so let's deny them this tool," it's "the police will just enlist private companies to do their dirty work for them, so let's deny everyone this tool."


You can deny the police the tool even though third parties.

Example: https://www.nytimes.com/2019/07/09/us/politics/trump-twitter...

Twitter is a third party but a court ruled that Trump blocking critics violated their first amendment rights.

It is simple enough to ban the government from using AI in the process of law enforcement and that ban would extend to whomever they enlisted to help them.

If you worded the statute correctly.


You're right about facial recognition, but I was extending that same concept to other domains. For example, the police can go to the solar company and get them to make your panels stop working. Or something.


I once worked with developing PV-inverter software. The web "features" always made be uncomfortable given the projected lifespan of >10 years.


> I'm afraid the software might even have some kind of countdown where the system will stop working when it hasn't been connected to the internet after a certain amount of time.

Or, just an automatic firmware "update." Most are made in a country that we are not currently on very good terms with. No, I don't trust any that require internet access, or an "account" of some sort to activate it.


> Maybe I'm too paranoid. Thoughts?

Yes. They aren't using these features to serve you ads. They aren't using them to lock you in to their platform. They added these features because the vast majority of users are not particularly technical and they want some dashboard they can login to to see some cute metrics about their system. They added these features because downtime alerts are genuinely useful but it is difficult to implement without some sort of remote connectivity. And if your local utility needs the data for some reason, or needs the ability to coordinate power delivery to the grid, either now or in the future, there is a mechanism in which they can do so.

The number of users who consider this an anti-feature is such a small minority of users that it isn't really worth their time to accommodate. IOT certainly has its issues. TVs absolutely don't need it. For other appliances it can be a toss up. But there are also genuine use cases for it.


> They aren't using these features to serve you ads.

Yet.

Experience tells me that everything will eventually turn into an ad or data collection platform.

Furthermore, even if there is indeed no nefarious intent from the third-party collecting & storing your system's telemetry, if you don't get any benefit from it, why give it to them? It's still a risk they (or a some malicious attacker, or the government) use this data in the future in a way you didn't expect.


This. I'd not expect to become advertisement fodder for a product that I pay regularly for. The inverter, on the other hand, is something I paid for once, so for years two through 25 (the expected lifetime), I'm essentially a cost to them as they operate whatever service that it talks back to. There's an incentive there to try and recover that cost, and they can't do it by increasing my monthly bill.


> I'd not expect to become advertisement fodder for a product that I pay regularly for

It's not always advertisements. Sometimes, the company will hire some oxygen waster that wants to boost "engagement" in order to justify their job and salary, so now your system is using your telemetry data to find ways to waste your time (as well as track whether they've indeed managed to waste it) even though advertisements themselves aren't (yet) involved.

Not to mention, advertisements don't have to be served over the same device that collects the data. The inverter company can very well collect the data, sell it to a data broker where other ad companies will use this data.


Please tell me how the output of a solar system will ever be useful to an ad company. The fact that you own a solar system is easily obtainable in a number of different ways (permit data, google earth, etc...). The fact that you own a solar system might provide some minuscule value to an ad company but the production data of that system is literally useless for ad targeting. A solar system is passive generation asset. It literally sits there and soaks up irradiance from the sun. The variation of that output is completely independent of the people that live in the house.


The telemetry from a solar system includes the power being consumed in real-time, which can be used to infer occupation of the premises.

Data brokers accumulate various bits of data that are unreliable and useless on their own but in aggregate they can be used to "connect the dots". If an IP address & user agent pair or tracking cookie only lights up when the solar system detects significant power draw suggesting someone is at home, you can tie a pseudonymous IP or tracking cookie to the solar system data which presumably would include name or some other bit of identifying info.

Now repeat that with all the other information you've got and you're able to paint a very detailed profile off someone.


Everyone in this sub-thread is shifting the goal posts. There is no requirement that solar inverters need to be aware of household power consumption. You can certainly set that up if you want, but an installation in its most basic form would not include that feature. You simply connect the inverter to the house panel and what you don't use flows back through your utility meter and onto the grid.

And again, the solar data by itself is going to be nearly useless as it is entirely a function of the irradiance on the panels, and does nothing to reveal customer behavior.


A little fuzzy AI and it will know your routine to the minute. It can guess what room you are in and what activity based on power usage alone. It is not only monitoring the sun, but how much power is left it can feed back into the grid.


I once worked at a greentech firm. Someone's hackathon project one year used energy usage data to predict customers' stock prices.

Will it work on the micro level of an individual home? Maybe, maybe not, but the potential for that data to be used to model behavior is definitely there.


Ads are 100 percent irrelevant.

But the lock-in is there regardless of whether they intended it or not. And, by they way, you are insane if you actually believe they do not intend a lock-in, given that they're making direct (and probably illegal) threats to "void the warranty" if you don't stay connected.

If they go out of business in 5 or 10 years, and your system stops working, you are screwed. The same applies if they just decide it's no longer profitable to support old products. Even if they are still in business, they are not going to do anything whatsoever to help you keep running without them.

The fact that most users are stupid enough to sign up under those circumstances doesn't change the fact that it's a stupid thing to do.


> But the lock-in is there regardless of whether they intended it or not.

Are you familiar with these systems at all? You can't lock someone into an inverter brand. If you don't like your current inverter, you just buy a new one and plug it into your system.

> If they go out of business in 5 or 10 years, and your system stops working, you are screwed.

I work in the industry. I can assure you that won't happen. These are not fly by night VC funded IOT companies selling $100 trinkets. The industry went through a rough patch ~5 years ago, a fair number of restructuring. Literally zero incidents where products were EOL'ed due to remote servers turning off. There are a number of reasons for this, I can go into more detail if you really want me to. But the gist of it is that these are revenue generating assets many of which have large financial institutions holding the bag. They have the leverage to make sure that these types of systems never turn off.


> Are you familiar with these systems at all? You can't lock someone into an inverter brand.

Unless everything speaks standard protocols and can easily be swapped to a different control and monitoring system (including one the owner runs locally if they want to do that), you can lock somebody into a subscription, on pain of having to replace an expensive durable asset that should otherwise last a LONG time and requires expensive skilled labor to swap out.

And, without having read the ToS, I have a suspicion that you can also Alter the Deal under which that service is provided.

> The industry went through a rough patch ~5 years ago, a fair number of restructuring. Literally zero incidents where products were EOL'ed due to remote servers turning off.

5 years ago, how many installed, Internet-connected systems were already 5 or 10 years old? How many different backends were supporting them?

> But the gist of it is that these are revenue generating assets many of which have large financial institutions holding the bag.

I accept that that reduces the chance of them being turned off. I would in fact be interested in hearing more about this revenue model.

Of course, you now have me worried about the risks of them being remotely meddled with in various ways to suit the interests of whoever is providing this revenue. It's usually a bad sign if you buy something, but somebody else thinks it's still their "revenue generating asset".


> Unless everything speaks standard protocols and can easily be swapped to a different control and monitoring system

We are taking about AC and DC current. That is about as "standard" as it gets. Your panels produce DC current. Your inverter takes that current and converts it to AC. The AC current is then sent to your main panel. The inverter is a fraction of the cost of the entire system, and most inverters only last 8-10 years, so they are expected to be replaced a few times through out the lifetime of the system. It would be physically impossible to implement some sort of inverter level DRM that could then bind up your entire system. You could brick the inverter itself, but for reasons explained below that would be very unlikely.

> I accept that that reduces the chance of them being turned off. I would in fact be interested in hearing more about this revenue model.

I don't know the exact number, but the overwhelming majority of residential solar systems are financed, either through a lease or a PPA. The banks that offer this financing maintain lists of "bankable" equipment and in order to be eligible for financing the equipment used in a project needs to be on one of these lists. And to get on these lists manufacturers need to meet a number of requirements, one of which is that there are guarantees that in the event of a bankruptcy the equipment will continue to operate and that warranties will still be honored.

> Of course, you now have me worried about the risks of them being remotely meddled with in various ways to suit the interests of whoever is providing this revenue

That is a fair concern. But if you enter into a financial contract with an entity that has the ability and standing to shut off your solar system for whatever reason, then that is your own problem, and somewhat tangential from the topic in this thread. And FWIW, in practice it very rarely makes sense to turn these things off. It costs too much to repossess them, and the homeowner will just end up putting more money towards the utility bill to keep the lights on. Instead they just keep the electrons flowing and hope their collections department can sort things out.


> You can't lock someone into an inverter brand

Challenge accepted. Just sell a fully integrated system that refuses to do anything if the incorrect inverter is connected.


You can't DRM a solar panel. But if you want to devote your life to trying, I am not going to stop you.

EDIT: And the more I think about, it is very unlikely that you could do the same to an inverter. You could certainly brick the control electronics, but the power electronics, the stuff that matters, it would be difficult. So snipping a few wires and swapping out some boards is probably sufficient.

How about this, if anyone on this thread ever has their inverter remotely shut off because a company goes out of business, I will personally volunteer to develop a work around.


Note how the conversation is actually about removed features, and inability to opt-in to unwanted features. This isn't a vehicle where you must choose between 2 or 4 wheels, but is a service where your purchase is sabotaged if you opt to prevent it from being sabotagable via an unnecesaary network connection. To use identical rhetoric, an air-gapped network is useful, and consumers need the ability to directly monitor and maintain devices in their custody, so there should be a mechanism for them to do so.


That's great that they provide those features - none of that is a reason to require an internet connection to use the solar panel at all.


Are we talking about powering a whole 4 bedroom home? Or just a few panels, a small charge controller, small battery, and like 2000w inverter? Renogy (etc.) sells all of the latter on Amazon that AFAIK don't require any connectivity. But you can't put it in circuit with your house's existing wiring and still be up to code.


Depends on where you are. My local Code Gestapo is only concerned about AC systems connected to the grid. DC and PV wiring aren't part of their remit, and they're not in the least interested in being tasked to cover those, as it would mean more work for them.


Completely agree — thank you for the heads-up! Right up front, I wouldn't have expected this to be another sh*tware connectivity thing unless I was doing net-metering (it kinda does have to talk to the elec utility for that).

My general expectation is to do a small circuit for things that don't need to be connected at first (like electricity for the workshed), then get fully off-grid if I do a full system. No way do I want to be off-grid and relying on some of the components calling their mothership.

I'm also deeply suspicious of all of these schemes where solar is "free", whether they are fronting the cost to reap the solar credits for themselves or installing to charge you a fixed-rate electrical bill. Far too many opportunities for things to go sideways.


Where do you live, and what products are you looking at?

I have an 11.6Kw solar system with a SolarEdge inverter. It does not need an internet connection to function. I currently have it connected to use the app that reports performance data, but it's not a requirement for functionality.


<network> Of Things is great.

<Internet> of things isn't.


No. More specifically, I am not interested in IoT for many or even most of the things in my home.


My system is connected to the internet - but it only transmits small amounts of data out for reporting. It shouldn't be hard to harden their code to prevent hacks. If the local software is just making outbound connections and talking (no inbound connections or listening) it would be a pretty tricky entry point for a hacker.

You could also connect it to a guest wifi network.

The data it provides is rather useful as you can see the efficiency of each panel - useful to know when they need a cleaning :)

btw, my panel disconnected from wifi for a month and nobody said anything.


I don't blame you for being disappointed in the situation. I would be too.

I've replaced a bunch of various appliances and etc, furnace, hot water heater, washer and etc. All have various internet connectivity options but they're all OPTIONAL. They are there at least from the start as an option for the owner to participate in, in exchange for some level of convince. That seems like a premise I would also expect from a solar panel setup ...


No, you are not alone. In Cebu City, Philippines, if you buy, from a cheap vendor (https://web.facebook.com/solarcompanyEST2018/?_rdc=1&_rdr), a solar panel set, a controller, an inverter, and a battery, you only can get dumb offline equipment. IoT (connected) is not even an option.


Depends. If it’s SMA, the stuff I’ve seen only sends metrics out to a server so you can see the output in a mobile app, and has little to no local management capabilities: https://taoofmac.com/space/blog/2022/08/04/0800#solar


I had no idea. I'd be okay with that if the firmware was open source, but I can dream.

On the other hand, if something bad happens (virus, remote bricking, disabling some functionality that was in the contract in un update), you can probably sue them. Granted, if they face massive service disruption, you may not be able to gain much from it.



What features do you need? You must be looking at a different range of devices than what I use but I am entirely off grid and have no internet connected devices in my solar system. I have even avoided Bluetooth enabled things without much fuss.


> the manufacturer states that not having it connected to the internet voids their warranty

I dunno about much of the rest of the world, but in Australian Consumer Law such a declaration would certainly be unlawful.


They're very tricky about wording it, they'll say things (like the car companies do) such as "not being connected to the monitoring service MAY void your warranty" (if a fault happens and you don't detect or correct it - just like changing your own oil and replacing it with baby oil will void your warranty).

And you can be hella sure they'll default to any excuse not to honor their warranty, because they know that you'll probably not bother fighting them in small claims court.


In South Australia at least, it's a legal requirement that the inverter remain connected to the internet. South Australia Power Networks requires the internet connection so they can remotely shutdown your solar system. They are added as an additional remote user to the system during installation:

[PDF] https://www.fronius.com/~/downloads/Solar%20Energy/Quick%20G...


Huh, weird. I’d have expected any signalling like that to come through the power lines that the solar system is connecting to. Using an internet connection for it seems like adding unnecessary complexity, procedural fragility, and another point of failure—and it’s not like your internet connection is more reliable than your power supply, most have no effective battery backup and go down immediately if they lose power. And not everyone even has a persistent internet connection—living out in rural western Victoria, I use a cell phone for my internet supply (it’s the best non-business service I’ve experienced, actually, though NBN fibre should be better now: clear line of sight to an Optus tower 400m away that serves under a hundred people; the only alternative is satellite—under half the speed, over twice the price, and 600ms latency).

(Mind you, I would still expect this to be completely distinct from anything about voiding a manufacturer’s warranty even if it was a grid-connected system which is probably but not necessarily the case.)


From an academic research scientist that works on energy system analysis:

In addition to outputting telemetry data about a solar array's production - whether that is meant to be used for the homeowner to monitor their system's performance or a third party to use for SRECs verification - the IoT modules on modern PV inverters are increasingly being designed/included for the purpose of receiving external input signals. These would either come from the local utility, distribution system operator, or regional balancing authority - and would communicate the need to change the inverter module's mode of operation to ensure the stability of the distribution circuit that it was interconnected to.

Small scale utility customers are typically only billed for the "real power" that they consume (real power is metered in kilowatt hours, kWh). This billing can come in different incarnations - i.e. volumetric tariffs, time-of-use tariffs, net-metering tariffs (as with bi-directional solar metering). However, in order to operate a reliable AC power distribution system you need to also be concerned with the supply of "reactive power" (metered in kilovolts-amps reactive, kVA). Reactive power is consumed by inductive loads (like electric motors) in large volumes particularly when starting up. Moreover, reactive power cannot easily be transmitted over long distances in the same way that real power can. The consumption of reactive power can cause the voltage and current curves of an AC power flow to shift out of phase with respect to one another. This effectively reduces the apparent power delivered by the system. This can also cause other problems with maintaining the system's nominal frequency and voltage.

Large AC generator units, that are essentially comprised of a large spinning mass ( such as a natural gas turbine for example) naturally output a more even balance of real and reactive power components. However, when you try to replace these units with lots of small DC generators, like solar panels, whose AC power characteristics are determined by the design of the inverter modules, what you end up finding is that there is a deficit of reactive power supply. Inverter modules can technically be setup to produce reactive power - but they generally are not tuned to do so. This is because it detracts from the amount of real power that they produce. And remember, that real power is the only thing that you get paid for under a net-metering tariff. There are currently no markets which would compensate you as a homeowner for the production of reactive power.

So, with the knowledge that many of these inverter systems will have 10-15 year lifespans once installed - and that grid operators will likely need to call upon them in the future to dynamically control the supply of reactive power at certain locations (particularly as more renewables come online) regulators are increasingly requiring that all new inverters have networked communications capability so that they will be ready once such a command-and-control protocol is eventually developed and implemented.


Your comment seems highly relevant and sadly ignored. Some relevant links:

- [IEEE 1547-2018 - proprietary](https://standards.ieee.org/ieee/1547/5915/)

- [CA energy commission 2020](https://www.energy.ca.gov/sites/default/files/2021-05/CEC-50...)

- [IEEE power and energy commission report 2020](https://www.nrel.gov/grid/ieee-standard-1547/assets/pdfs/sma...)

AFAICT, the mandate to permit not only network connection but control is coming, but the security of this link was expressly deferred.


This is an interesting narrative of how this exact scenario is currently playing out in Hawaii - which, somewhat by necessity, has historically been among the first to blaze new territory when it comes to renewable energy policy implementation.

https://www.canarymedia.com/articles/batteries/hawaii-is-pay...


As someone who works in renewable energy and has a dose of skepticism on most IoT devices. I very much want to have a read on how much energy my solar assets are creating therefore I am very comfortable with an inverter shipping out production values. That way I can resolve them when it isn't producing due to soilage or any other issue.

I would chalk it up to being overly paranoid. However if you are in a foreign country and have a system that is connected via payments that is a different matter all together (i.e. that you don't own the asset and you are paying for the electricity produced).


I have several off-grid inverters, solar charge controllers and AIO systems, none of them have any IoT or any connectivity requirements.


You're correct, this seems like a terrible idea. Why not just make a cheap, repairable device? It would be better for everyone.


Wait, so is this because you want to feed back the generated power to the grid?

How are all those rural places too remote for the internet working?


Given this discussion (and as someone that lives in a very rural area), I am planning to go entirely off-grid with solar and batteries. I am no longer interested in a grid-tied system in any way, shape, or form.

Under no circumstances would I install a system that could be remotely powered down to save the grid, when the grid itself is the source of the problem.

It’s like the power companies want their cake and to eat it too.


Yeah. If I put in enough solar & batteries to fully serve this largish house, the local utility will still make me pay about half of my current electric bill. It'd make more sense to go off-grid and buy a generator for backup.


My electricity provider’s unreliability already forced me to buy a generator.

As far as I understand, that also brings up the true farce about grid-tied solar: you are still at the utility company’s mercy to keep the grid working, because the solar system only works when there is grid power. Insane.


I got solar at the start of 2020 with a Growatt inverter. Internet completely optional.


Absolutely not. I would not want any of that to be connected in my personal setup


not sure why it would need internet but I would DIY with an ESP8266 if it does (no cloud involved)


being able to DIY a high-wattage inverter would mean that this person already has everything they need to avoid asking this question in the first place,


'8266 is an MCU with built-in Wi-Fi. They're not talking about DIY'ing an entire inverter, just the web interface to it.


web interface to what? the inverter, right? and to monitor voltage and current of the input(s) and output (s) of an inverter you need to know how to connect up the sensors, keep them protected in whatever environment, and make sure they're reading correctly.

a person that knows how to "DIY" these things correctly already knows how to use a microcontroller to log the data and either present it or send it somewhere that can present it.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: