Historical data has value there though. Even if knowing that flight X is usually around $500 from anecdotal history is the foundation, there is opportunity. A third party has a great opportunity to provide even better information.

It might be bad for people that live in hub cities though.

That works until a third party is hired by all the airlines to give a pricing algorithm for a specific customer pool and routing so there is "your" price for flying, independent of which airline you try to use.

I think you are high on your frequency and low on your distance there. The frequency is inaudible, between 0.1 and 1.0 Hz depending on the diameter. The thump you feel in your body can go well over 1km.

You have about 100,000 nodes to work with, although there are subsets with no interdependencies. The opportunity for an automated system is to make decisions in a cascading failure situation fast enough to isolate problems, but this requires overriding some protective devices (relays) that would otherwise operate to protect things. It requires a holistic look for sure.

It is not that different from a master-system back 30 years ago, although managing 100-1,000x nodes of what was practical then. The real challenge is if it becomes easier or harder to game-- assuming it is using machine learning and not primarily hard-coded, at least there isn't an instruction manual in code to show how to overwhelm the system.

My concern is if it is designed to maximize ongoing reliability or to simply address first order impacts. The former is about 10x harder.

To me that is the big issue with trying to manage immigration; isolationism doesn't work, but now you are making it completely undesirable for people to immigrate given the potential treatment at the whims of the government.

I will be curious how long this can be sustained... by the administration, businesses, and communities alike. As I look back on my career it is interesting to imagine just how many great engineers I worked with would never bother with the US if they had to tolerate this type of treatment. In the "great" category it is possibly 30-50%, most of whom ended up spending most of their life here.

The benefit is more for distribution issues than diurnal loading-- it helps reduce loads on the11kV side.

In Sweden I am surprised there isn't more battery storage at a neighborhood substation level. The way they do their underground distribution (I think much of northern Europe is the same), they have small enclosures every 1km or so with switchgear and a 380V transformer; these types of locations are almost ideal for smaller battery systems, especially given how over-paneled most homes with solar are. I wish it was that easy for utilities to add distribution-level batteries in the US as it would really open up more distributed generation opportunities.

After about 3 years it started degrading from UV exposure. Home Assistant needed to use a different integration, and switching over was a real problem. It was fine when I first got it working and worked around a year and a half. I want a longer term solution though.

Oh, and my ecowitt wasn't really UV resistant. Maybe good for 3 years.

I paid $150 or so for the Ecowitt station, and $320 for Tempest (IIRC). When I first looked at it, I could not justify the Tempest... but it was penny-wise. The Tempest has been a joy compared to the persistent hassles with Ecowitt (and specifically Home Assistant compatibility).

Yes, and boy does it do a better job. Rain accuracy is a little more complicated with the pizeo sensor, but plenty accurate for my needs. (It isn't as good in very light and very heavy rain apparently, and liquid sunshine is the rain du jour here.)

I'll just add one more thought to this. For a residential or light industrial customer, putting rooftop PV plus a battery in should be a very attractive investment. This is one of those cases where the little guy has an easier solution to the problem than the big guy. The battery just needs to be big enough to cover loads for ~12 hours and you are good, bigger being better almost always though.

The problem for the grid operator is that getting GW and GWh via rooftop solar is a slow process.

I finally threw away my Ecowitt weather station and bought a Tempest. The ecowitt units were just too buggy; Tempest just works and can be fully local for integration into Home Assistant.

I just didn't really need any more crap. I looked, but nothing was really compelling enough for me to hit "buy."

A big part of that is squarely on Amazon's shoulders, becoming a broker for knock-off stuff of indeterminant quality. We bought a lot of stuff last year, some of which was useful and others that ended up being a bigger pain than they were worth. I was actually needing some "retail therapy" as well, so I am a pretty easy mark right now.