Millions of smart meters will brick it when 2G and 3G turns off

The big benefit is taxation: it allows for net-metering with the surplus sold at a low rate to be bought back at a much higher one. Obviously this was the plan all along...

Interesting how "new and special" that seems in the UK with only daily prices? There are many providers offering hourly pricing in the Netherlands if you want to. I've seen some projects by enthusiasts to link those APIs to smart home systems and essentially use very cheap energy by timing their larger loads to the low-price moments. Especially on days with a lot of wind and solar energy.

You can also cut people off, or shunt them onto a pay-as-you-go tariff without installing another meter

It's also another point of failure.

I have a SMETS2 meter, and appreciate the time of day billing and better feed-in metering for my PV, but I'll not be at all surprised if this is hacked one day to cut millions of people off.

There needs to be a guide to physically removing telematics from vehicles.

This happened with my wife's Fusion Energi.

It used AT&T 3G to communicate with Ford's MyFordMobile servers. Ford sunset MyFordMobile shortly after AT&T sunset 3G nationwide (US) two years ago. They had a service program which replaced the head units with more modern SYNC 4 ones that have 4G radios in them and work with Ford's FordPass service.

However, the dealer nearby us didn't know about the service and, because they didn't have the parts on hand, would have needed to hold that car for two weeks before doing any work.

So, just like that, we lost remote start (via app), location tracking and a bunch of actually useful features.

That's good to know. I have been getting along fine by driving older cars, but obviously that won't last forever. I have always assumed that whenever the day comes that anything which still runs was built to phone home, I'd have to learn how to take the dash apart and remove the SIM or the antenna; but perhaps the endless grind of technological obsolescence will do the job for me. Wouldn't that be ironically nice?

Who pays to replace them?

One could mandate a modular design so that the communication pcb could be replaced, i.e. less e-waste. I wonder if smart meter manufacturers stay in business long enough for it to matter.

I'm not sure how modular they are, if they're just replacing a transmitter or something, I've seen some places transmitting only locally (zigbee or other) then a module on the street syncing upstream

But let's say even $500 (it's probably cheaper) for something that lasts 8 years is still much cheaper than sending people to get the readings

Nobody cares about e-waste. If they did companies would stop producing new products at breakneck speed and consumers would stop buying them.

Instead they care about the optics, so you get the ocassional Apple video about how they replaced foo material with bar, while creating not servicable BS - oh, and things like paper straws.

Recycling is also mostly BS (things in the "recycle" bins just end in landfills abroad).

So this is just another drop in the huge ever increasing pile.

> and yet another few million pieces of e-waste are generated

That's money generated while the e-waste is not factored in. I know I'm stating the obvious. But it's a problem with being realistic, as you start seeing idiosyncrasies everywhere: I run my household on 4g as I can reach realistic speeds of 100Mb/s. Theoretically that goes up to 150Mb/s. A 4k stream (compressed without visual loss) requires 15Mb/s. So 10 people in a household could simultaneously watch 4k content.

Did I fail at math? Why does anyone need 5g? What for?

I'm dealing with this in a slightly different context right now - installing cellular remote access/VPN appliances in industrial manufacturing equipment. The equipment is required to have a lifespan of at least 10 years. Our adjacent equipment regularly lasts double that. Some of the equipment is built around castings and transformers that are 4-10x that age.

But they also want me to be able to dial into the equipment later, and won't give me WAN access. But no cellular vendor in their right mind will certify that the adjoining towers will have any particular frequency available for more than ~5 years, typically limiting it to a warranty of 3 years. This isn't the only pain point - what about the licensing server that the proprietary IDE for the PLC calls out to? The operating system on which said IDE is installed? Security protocols on the SCADA uplink? Replacement parts and critical spares? Good luck!

Unfortunately, a huge amount of industry is effectively subsidized by breakneck investment in consumer smartphones, automotive, and PC equipment. People don't care or don't understand that their choices on the showroom floor for the glossier, flashier, 'smarter' gear result in painful, expensive obsolescence just a couple years in the future. Manufacturers are only incentivized to sell new products and ignore old products, we need a massive shift in consumer preference if we want to avoid this.

I guess that at some point they will make a deal with Amazon to use their Sidewalk system to provide connectivity to meters.

I think in reality it is a more efficient use of bandwidth using GSM. The alternative means your utility would have to maintain infrastructure and use up alternative bandwidth. In reality I think there should be regulation to keep some really basic form operating like 2G with GPRS or something on all networks.

My municipal water meter transmits readings in the clear on the 915 MHz band. There's a neat program to catch the readings: https://github.com/bemasher/rtlamr

It's just a tradeoff. There are advantages to consolidating on the same infrastructure for many different use cases. There is a much larger critical mass of people working on cellular technologies than on the technologies necessary to communicate via power lines or custom radios.

But of course there are also disadvantages, and this is one of them.

There is always a balance between concentrating effort behind a single solution, and having that solution be a single point of failure.

Depending on where you are they’ll likely just come and replace it or they’ll start charging you for having to send someone out to read your meter.

It’s nice of them to ask. Typically it’s theirs for running their side of the contract with you, and oftentimes they’ve got a right to come service it.

Why would you refuse if you can get detailed information out of a smart meter about your usage? You can save a lot of power by knowing what is going on and if you have solar you definitely want a smart meter. You also don't need people coming by every so often to check how much electricity, water and gas you have used.

Here in Switzerland 80% of homes are supposed to have smart meters by 2027 but it looks like it will take longer, we are at 20%[1]. Once you have one the energy company has to give you access to (by law[2]) it (rs232, Ethernet what every the device offers) so you can read the data from it.

[1] https://www.srf.ch/sendungen/kassensturz-espresso/kassenstur...

[2] https://www.fedlex.admin.ch/eli/cc/2008/226/de#art_8_a

The newer generation smart meters operate on their own network, so that its critical and separate infrastructure so this won't happen again in the future. Also you have to be pretty set on not getting a smart meter as it removes you from a lot of competitive energy deals, especially ones with a cheaper night rate.

Why? What does this matter to you? It's not like these people are going to lose their electrical service due to this.

It's a problem for the utilities, and a cost that will be borne by all their rate payers, but I don't see any individualized cost or benefit here that is dependent on having or not having one of these meters.

I don't understand how their problem affects you? Does installing a newer radio module require turning off your supply?

I worked on this, like 15 years ago, but ultimately after you go out of lab, there is so much garbage in the network that a lot of time this is not feasible.

That definitely exists, but AFAIK it requires special equipment to get the signal across transformers. Not much of a problem in urban environments, but e.g. in low-density areas in the US, where every house has its own transformer, it might be a problem.

A custom solution would only be worth it if companies looked that far ahead instead of maximizing short-term profits.

Homes tend to have internet connections!

Is there an existing mass-produced solution that can do it? The normal power line ethernet devices can handle ~300m and degrade quickly with noise. You'd need something much more specialised to go low speed high distance and able to handle lots of unsynchronized devices on the same line.

More common solutions installed in other places are dedicated mesh networks like Gridstream or AcuMesh. I'm really not sure why would the energy companies go for GSM instead. This is not just for the cities - Gridstream can reach a few km away.

On the other hand here's a buzzword-heavy ad from Thales that pushes for GSM https://www.thalesgroup.com/en/markets/digital-identity-and-...

Edit: apparently the powerline based data exists in the wild. (Keyword: PLC, but it's only one-way) Here's an interesting report from an Australian provider https://www.aemc.gov.au/sites/default/files/content/1494afbe... 10 years ago they were 46% rf mesh, 39% powerline data, 0.1% GSM. It's hard to find something more recent, but local mobile telcos had a lot of announcements / ads in later years, so the trend may have moved as well...

A decent global market overview from 2019 https://www.smart-energy.com/magazine-article/ami-penetratio...

signalling/communicating through the zero cross of the AC is trivial as long as there is a galvanic connection (wire). Once you get to transformers nothing much happens unless their is a module on the secondary to intercept it and then use other channels to ferry the data. In other words each transformer would need to have a communication device/router - I suppose that's way more expensive than ad-hoc modules per meter.

You've gotten some answers below, but one answer is: There are several common ways of handling this, from powerline communication to cellular to LPWAN approaches, and all of them are in use, with different trade-offs. There's no one standard for smart metering infrastructure. Older AMR approaches often would just broadcast a signal that you could pick up with a vehicle driving around the neighborhood, but they're one way only. AMI uses cell, PLC, LPWAN, and sometimes zigbee (to a neighborhood gateway). What your meter uses depends on your power company and when it was installed.

I consult in the industry, and certainly don't know everything, but everything that I've seen says that no, they don't. There are companies out there like Bidgely who claim they can do that to some extent, but they likely can only do that for really big loads unless your meters only powers a very few things.

You want that if done with respect to privacy. Your electric car should not charge when rates are high. Your water heater should only turn on when the wind is blowing and there is a surplus of cheap power. (This requires extra hot water storage to make it through a day)

No, the smart meters generally don't have this functionality. Other appliances can in theory hook into the market information from the supplier, but this isn't very common, and is generally under the user's control. The most common thing where users cede some control is home battery systems, where some providers give a discount or outright pay the users for control over using the batteries as grid storage ("virtual power plant"). But this is independent of the smart meter, a smart meter is mostly useful as someone with a home battery or EV because you can get acess to minute-by-minute energy pricing, whuch will generally save you money if you can shift your draw from the grid to when it's cheap.

(I will say for the average consumer at the moment a smart meter only gives you a mild convenience increase, and is likely not worth the trouble. I haven't got one installed myself, despite and now in part because my energy provider has been hounding me to install one, due to this push from the government. There's no reason to be an early adopter for this if the benefits aren't good, given how badly scuffed the first iteration was)

Octopus Energy were talking about being able to do this using machine learning given only real-time access only to the total power consumption.

https://youtu.be/g-RQ0dOPTtc?t=474&si=r0zhDUr3OaMieHOz

They offer things like discounted rates for charging electric vehicles. And are talking about being able to tell people when they're using more power than usual that they forgot to turn the oven off or they're using their dishwasher during peak or whatever.

Edit: he mentioned Voltaware which seems to be this: https://voltaware.com/blog/2021-07-01/voltaware-octopus-insi...

It is a shame there isn’t a nationwide grid of wires it could use to send those few KB of data through.

The comms module is normally separate from , but attached to, the electricity meter. Not user replaceable though. Slightly surprised (act, not really) that this is happening. Seems to have been dodged in Scotland & the North because cellular coverage there was poor enough they went with a 400mhz network instead.

I think it's the register making a hash at a pun: https://en.wiktionary.org/wiki/brick_it

The pun doesn't really work, but when has that ever stopped headline writers?

OpenBTS is a thing

So replace a $2 gsm module with much more expensive femtocell?