What to power my fembots with?

[D-wolfmaster_2022]

I am planning something special for my gallery, but I have some problems that my imagination alone just won't help with. The big issue is what to power my robots with. I had the idea of solar batteries, but I want to explore the issue some more before finally settling. Do you guys have any suggestions?

[BebesIsPerfect]

Solar batteries is a good idea, I think. It could possibly even be used as a secondary power system if the primary power source gets compromised. Now regarding the Primary source, I'm not sure how technical you want me to be or how feasible you want it to be. Anything unfeasible that I can think of just feels like a creative cop-out.

Naturally, there's a wide range of different style of batteries (Lithium Ion, Alkaline, Zinc, etc.). Do you want her to have to recharge every now and then? Lithium Ion would probably be the best one as its so commonly used. If not, it could be interesting to have her consume a power source and need a semi-constant supply of this kind of battery (kind of like how the Plutonium was injected into the time machine at the beginning of Back to The Future, only something more common than 1985 Plutonium). And it could be kind of funny to have a fembot powered by a comically large amounts of AAA batteries, or even function on just one solitary Lithium coin battery. I'm not sure if I was much help, but...

Solar power... I guess Wind power or Hydroelectricity aren't too feasible on a robot, eh?

[D.Olivaw]

What you power them with is going to depend on how much power they consume, and on how long you'd like them to be able to go between recharging.

Human bodies burn about 100 watts at rest and up to 500 W during vigorous activity. A desktop computer is in (very) roughly the same range. A fembot probably needs to be doing a lot more computing than a desktop, and her other systems will also draw power, especially whatever she's using to move around.

Boston Dynamics' Atlas humanoid robot consumed roughly 3,700 W during activity and 1/4 that (~1,000 W) when idle. Maybe the fembot is lighter than Atlas (5 ft, 190 lbs), but you probably want her to be doing much more complicated stuff than Atlas, especially things that are very computationally intensive like understanding her environment, engaging in nontrivial social interactions with humans, and making and executing plans in the face of changing circumstances.

You can conjecture whatever efficiency you like for the hypothetical robotics technology behind the fembot. If that technology resembles a mere refinement and continued development of what we have today, then the power demands will be correspondingly higher than for more speculative or advanced options.

Note that all of this power consumption ends up as waste heat that needs to be gotten rid of somehow.

So that's demand. Now we have to consider supply. The limitation is the volume of the fembot's body, into which we have to cram her power source unless we plug her into a wall outlet.

You mentioned solar. The theoretical maximum is around 1,000 W per square meter, since that's just the power of the sunlight that reaches Earth's surface. In practice solar panels aren't perfectly efficient. Commercial-grade solar cells are around 15-25% efficient, whereas the best, cutting edge ones approach 50% efficiency. That means you're producing 150-500 W per square meter of solar panel. You may compare those numbers with your consumption numbers to decide on viability.

For the storage of our internal power cell, lets think in terms of watt-hours (Wh). If our fembot ran at an average power of 1,000 W, then a 1,000 Wh battery would power her for one hour. For eight hours of operation we'd need an 8,000 Wh battery. Li-ion batteries can achieve power densities as high as 700 Wh/L, while there are as-yet unperfected or theoretical technologies that could deliver densities up to maybe 1,500 Wh/L. For comparison, gasoline's energy density is 5,000 Wh/L, of which your average car engine can extract perhaps 1,500.

If you wanted the 1,000 W fembot to run for 8 hours on one battery charge, you'd need a Li-ion battery with 8,000 Wh / 700 Wh/L = 11.4 L. A human body has a total volume somewhere around 75 L, so our fembot would be around 15% battery by volume. If you wanted the batteries to last 16 hours, 30% of her insides would have to be taken up by batteries.

From here, it's up to you to play around with values until you come up with a result you like!

[D-wolfmaster_2022]

Solar batteries is a good idea, I think. It could possibly even be used as a secondary power system if the primary power source gets compromised. Now regarding the Primary source, I'm not sure how technical you want me to be or how feasible you want it to be. Anything unfeasible that I can think of just feels like a creative cop-out.

Naturally, there's a wide range of different style of batteries (Lithium Ion, Alkaline, Zinc, etc.). Do you want her to have to recharge every now and then? Lithium Ion would probably be the best one as its so commonly used. If not, it could be interesting to have her consume a power source and need a semi-constant supply of this kind of battery (kind of like how the Plutonium was injected into the time machine at the beginning of Back to The Future, only something more common than 1985 Plutonium). And it could be kind of funny to have a fembot powered by a comically large amounts of AAA batteries, or even function on just one solitary Lithium coin battery. I'm not sure if I was much help, but...

Solar power... I guess Wind power or Hydroelectricity aren't too feasible on a robot, eh?

Yes, I am looking for something feasible. Thank you, Bebe. Your suggestions are a great help in my search. I am not too familiar with how to power machinery. After reading what you've listed, it is evident that I must do some more research.

By the way, I love your new profile picture.

[D-wolfmaster_2022]

What you power them with is going to depend on how much power they consume, and on how long you'd like them to be able to go between recharging.

Human bodies burn about 100 watts at rest and up to 500 W during vigorous activity. A desktop computer is in (very) roughly the same range. A fembot probably needs to be doing a lot more computing than a desktop, and her other systems will also draw power, especially whatever she's using to move around.

Boston Dynamics' Atlas humanoid robot consumed roughly 3,700 W during activity and 1/4 that (~1,000 W) when idle. Maybe the fembot is lighter than Atlas (5 ft, 190 lbs), but you probably want her to be doing much more complicated stuff than Atlas, especially things that are very computationally intensive like understanding her environment, engaging in nontrivial social interactions with humans, and making and executing plans in the face of changing circumstances.

You can conjecture whatever efficiency you like for the hypothetical robotics technology behind the fembot. If that technology resembles a mere refinement and continued development of what we have today, then the power demands will be correspondingly higher than for more speculative or advanced options.

Note that all of this power consumption ends up as waste heat that needs to be gotten rid of somehow.

So that's demand. Now we have to consider supply. The limitation is the volume of the fembot's body, into which we have to cram her power source unless we plug her into a wall outlet.

You mentioned solar. The theoretical maximum is around 1,000 W per square meter, since that's just the power of the sunlight that reaches Earth's surface. In practice solar panels aren't perfectly efficient. Commercial-grade solar cells are around 15-25% efficient, whereas the best, cutting edge ones approach 50% efficiency. That means you're producing 150-500 W per square meter of solar panel. You may compare those numbers with your consumption numbers to decide on viability.

For the storage of our internal power cell, lets think in terms of watt-hours (Wh). If our fembot ran at an average power of 1,000 W, then a 1,000 Wh battery would power her for one hour. For eight hours of operation we'd need an 8,000 Wh battery. Li-ion batteries can achieve power densities as high as 700 Wh/L, while there are as-yet unperfected or theoretical technologies that could deliver densities up to maybe 1,500 Wh/L. For comparison, gasoline's energy density is 5,000 Wh/L, of which your average car engine can extract perhaps 1,500.

If you wanted the 1,000 W fembot to run for 8 hours on one battery charge, you'd need a Li-ion battery with 8,000 Wh / 700 Wh/L = 11.4 L. A human body has a total volume somewhere around 75 L, so our fembot would be around 15% battery by volume. If you wanted the batteries to last 16 hours, 30% of her insides would have to be taken up by batteries.

From here, it's up to you to play around with values until you come up with a result you like!

15-30% better, eh. Looks like there won't be much room for my rocket pack.

Anyways, these are some very helpful notes, Mr. Olivaw. I'll admit that I am not all that knowledgeable about electricity. These are all things that require further research on my part. For what I'm imagining a typical household humanoid robot typically runs for about 36 hours before needing a recharge. Does that sound realistic? The setting I'm crafting for my 'droids and 'noids to inhabit is not exactly hard sci-fi, but I want it to be as plausible as possible.

[D.Olivaw]

For what I'm imagining a typical household humanoid robot typically runs for about 36 hours before needing a recharge. Does that sound realistic? The setting I'm crafting for my 'droids and 'noids to inhabit is not exactly hard sci-fi, but I want it to be as plausible as possible.

36 hours without recharging implies that they are either extremely efficient or that their power source has incredibly high energy density, at least by our standards. Even if they run at only a few hundred watts on average (as, or more efficient than a human body), they'd need a battery about as big as the one in the example I worked through above. If they need to draw more power than that, the battery needs to be correspondingly more energy-dense to fit in a reasonable volume. You quickly reach explosive-like values.

[D-wolfmaster_2022]

For what I'm imagining a typical household humanoid robot typically runs for about 36 hours before needing a recharge. Does that sound realistic? The setting I'm crafting for my 'droids and 'noids to inhabit is not exactly hard sci-fi, but I want it to be as plausible as possible.

36 hours without recharging implies that they are either extremely efficient or that their power source has incredibly high energy density, at least by our standards. Even if they run at only a few hundred watts on average (as, or more efficient than a human body), they'd need a battery about as big as the one in the example I worked through above. If they need to draw more power than that, the battery needs to be correspondingly more energy-dense to fit in a reasonable volume. You quickly reach explosive-like values.

Explosive values, huh?
Alrighty then, so maybe 36 hours is a bit much. Like I said before, I don't understand batteries or power sources all that well.

[D.Olivaw]

Alrighty then, so maybe 36 hours is a bit much. Like I said before, I don't understand batteries or power sources all that well.

I don't really mean to say you can or cannot do certain things. Think of it like a framework for understanding the tradeoffs and constraints involved, and of course since it's fantasy you are free to ignore even those.

[D-wolfmaster_2022]

You have all been very informative. I give you my thanks. I shall ruminate on this new information with great care.