Are there other functional smart furnace setups?

[afk2minute]

This discussion is very active, and we both are editting our posts simultaneosly, so it is quite hard to understand what we are talking about.
I will come back to this thread tomorrow with the fresh mind and new ideas (also it is a deep night here).

Well if you say that this solution can work here is a little improved version

setup1

With this layout you need only 2 separate solar panels per furnace, i dont remember ratio for accumulators but its quite near 1:1 so say 2 accumaltors (stack inserter is 132kW but there would be some idle time so 2 60kW panels should be enough, if the idle time is more than a half then only one solar panel can be used). Its wider, but 2 tiles is ok for a lower power consumption.
Non-beaconed setup has its advantage of no need of manually rewiring grids, they are always would be separate if blueprint-builded.

setup2

Sadly im not good with combinators to make a controller for these things.

[MeduSalem]

Belt Smart Furnace.JPG (195.89 KiB) Viewed 1794 times

Buffer Chests on in/output, each furnace/Beacon row can be turned off individually if necessary.

The combinators for controlling the stuff would have to be somewhere at the end of each row... since there's no space in between somewhere obviously. But basically XNight's design from over here can be used.

The nearside lane of the belt would deliver the items, the farside lane would be for the output products. I would even argue that regular Filter Inserters are enough or even better than Stack Inserters, because the Filter Inserter is a lot more responsive to sudden changes in the item flow on the belt. In my attempts to make a belt based Smart Furnace the Stack Inserters acted strangely, especially if items are distributed sparsely on the belt.

... now someone can knock himself out with the Combinator mess and distribution/sorting system for source items and endproducts on the belt.

[afk2minute]

So the speeded up with 3 SM3 beacons + 2 speed modules (im using speed modules for the "worst case scenario" instead of productivity) is +175% crafting speed bonus, that is 5.5. Recipe for iron plate is 3.5 seconds, so 0,6 seconds per plate (lets use 0,5 for some emergency capacity).
10 iron ore is 5 seconds time for this furnace. Grabbing time is less than a second that means more than 75% of idling time, correct me if i made a mistake somewhere...
If that is so then 1 solar panel per 2 furnaces can be used, which is great. 500 furnaces = 250 solar panels (i would prefer more accumulators than the usual amount, say 400 in this case).

[XKnight]

So the speeded up with 3 SM3 beacons + 2 speed modules (im using speed modules for the "worst case scenario" instead of productivity) is +175% crafting speed bonus, that is 5.5. Recipe for iron plate is 3.5 seconds, so 0,6 seconds per plate (lets use 0,5 for some emergency capacity).
10 iron ore is 5 seconds time for this furnace. Grabbing time is less than a second that means more than 75% of idling time, correct me if i made a mistake somewhere...
If that is so then 1 solar panel per 2 furnaces can be used, which is great. 500 furnaces = 250 solar panels (i would prefer more accumulators than the usual amount, say 400 in this case).

Actually you don't need so many solar panels, probably one accumulator for tens furnaces should be more than enough.
The purpose of these accumulators is to give inserters time to finish their rotation operation if we are out of energy. After that we can disable all inserters and wait until full charge again. So, it should be like this:
- 0 tick: Temp accumulator has 100% charge, and it is connected to the main energy network, because main network has extra energy.
- 10000 tick: Night. Temp accumulator has 100% charge because its network has lower priority than main network and inserters are using power from the main network. Also, if accumulators in main network are discharging (because it is night time) then we should disconnect temp accumulator from them (to keep it full charged).
- 20000 tick: Outpost is under attack and main network is out of energy, so inserters now are using energy from temp network. We detect this situation and send next order to all inserters: "new item type = NONE", as a result after ~60 ticks all inserters will finish their rotations, and we will be able to switch them off.
- 21000 tick: Energy level in main grid is restored, inserters no longer use temp accumulator, so we switch them on.
- 25000 tick: Day time: temp accumulator has 95% charge. Because accumulators in the main network are charging we connect temp accumulator to them.
- 25500 tick: Temp accumulator has 100% charge.

By the way, belt-based solution is already finished for several days using the same principle as in robot-based solution.
It requires 8 tiles per furnace row and its block size = 4, also it is protected from low energy due to the same reasons as in robot-based solution. But before posting it I really want to check approach with input belt super controlling, because it looks promising and because we can use the same belt for delivering raw materials and for delivering produced items. Also, my target is to get as close as possible to 6 tiles per furnace.

P.S. And again I've made the same mistake: I've counted beacons twice, current value is 8 tiles per furnace row.