I'm creating humanoid robots that sweat to keep their components cool. It would seem wise to allow them to refill their water supplies by allowing them to drink. But is a mouth and throat leading to a tank really the most efficient way?
What is the best method for a robot to intake water independently?
If it is just for the intake of water, you could in principle use a tube connected to a pump for sucking in water.
Since you are bound to a humanoid configuration, you have various possibilities on where to place this tube.
However, considering that you also want to make the refilling as quick as possible, you have to rule out location where the diameter of the tube (and therefore the intake volumetric rate) would be too small: this rules out fingers and leaves you with pretty much two choices: upper or lower torso.
In the upper torso the robot would do something similar to drinking: put its mouth close to a water body and suck it in.
In the lower torso the robot should somehow squat on the water body and, again, suck it in.
The best of the two strongly depends on the surrounding environment and other boundary conditions: for animals in the wild drinking is one the most dangerous situation, as the head has to be lowered and cannot be used to explore the surrounding for menaces.
There is no best way for a robot to intake water. It depends on if you want your robot to feel and act like a human, or if you are just using the shape. The position of your cooling system and how much liquid it will require.
If you want it to be and feel human, then you would use the mouth and throat system. Putting water anywhere else would give away the human feel.
If you want it to be efficient, then you would simply have an intake area they would use to pump it into themselves. Most likely in the chest or shoulder areas as the water will flow down into whatever container is holding it. It could also store it in its head, but you will probably want real sensors that look like eyes there, so it would be a bad idea. Too low (aka pelvis or legs) and you have to pump the water up to critical components, but you already need to do that to reach the head. You would also need to consider how much that part can deform or moves (stomach area might be bad).
@Cognisant mentioned water sloshing around, and if you are evaporating water, this will always be a problem as you can't constantly fill up your water tank everywhere you go.
I don't think your traditional sweating idea would be a good way to cool any robot. Sweating works when the water evaporates and takes away energy hence cooling us. However you will require some airflow, something to transfer the heat to the "skin" and a mechanism to ensure water overflow and affect your electronics.
I suggest, rather than sweating, you use special heat vents to expel the heat. Water is used to transfer heat energy from electrical components to a set of heat sinks and fans which transfer the heat to the air. The water stays inside a set of pipes and pumps and you really don't need to refill it. If the heat sink is cool, nothing happens (in a sweating situation, the water will just stay on the skin), if its hot, the energy is transferred quickly from the components to the heat vent and cooled using fans, not how breezy it is outside. In this case you can also use oils or other special liquids to replace water.
The final way, would be to have all your electronics submerged in a special liquid. I'm not sure what its called, but there are videos online of computers which use this. Its expensive as hell and would allow you to submerge all your electronics and cool yourself using the skin to transfer the heat to the air around you.
If there is already a system which funnels water contained inside the robot to its skin, then it might be the easiest to just reverse the process and suck in water through the pores and collect it in the same tank. Granted, the quality of water taken in this way is not always assured, so there might be a need for purification process somewhere along the way. And it probably wouldn't be good if it were an automatic system, so the robot should be able to decide whether it wants to "drink" or not when (partially) immersed in water.
Apart from that, if your robot needs to blend into a human population, you'd probably need to design it to be able to drink the human way because then humans would probably accept it better.
The most efficient way would be to have the water tank as high as possible on the robot so when pumping it to the components you're not fighting gravity, and to keep the access point to the tank as small as possible so it's not taking up internal space unnecessarily like a pipe through the body would. Something like a hatch on the forehead that opens to reveal an intake port, or just a hole with a screw in plug will do, assuming there's ample space in the robot's head for the amount of water it would use over its period of use.
Also it's worth considering that a water tank could be any size and shape so it could go anywhere in the robot, meaning it would probably be better to design the tank around the robot rather than the robot around the tank. Another thing to consider is weight (water is heavy) and how that weight will move in the tank, having water sloshing around inside a bipedal robot will likely cause problems, especially if there's enough weight to knock the robot off balance.
To develop humanoid robots that can drink water, taking a biologically-inspired approach is the best way.
Your robot should be able to pick up a glass of distilled water, then place it to its mouth and tilt its head back. This would allow the flow of water from the glass, through a tube into the redox flow battery.
I don't understand what you mean by "independently".
Humanoid robots which are mingling with humans will be able to recharge onboard water supplies at lavatory facilities which are usually fitted with running water.
Drinking from a sink in a public area requires some contortions or the drinker must bring a cup. But there are basins of water in these facilities. Usually this water is in place to receive waste (and reduce smell / facilitate disposal). The robots would easily be able to use these basins as reservoirs to recharge themselves. Germ and sanitation hazards will not be an issue for a robot.
One would need an onboard hose to pull water up from these reservoirs to the robot's own tank. This hose could be extended from a hidden receptacle. A hose which was not hidden within the robot body would dangle on the outside and might spoil the robot's humanoid appearance unless it were concealed under garments. I think this last is the best approach: an extensible long hose for water uptake, concealed under garments. In the lavatory, the robot can withdraw the hose and extend it down into the water reservoir. The hose will be wiped clean and tucked away after use.
Properly configured, the presence of this hose under the robot's clothing should not attract too much notice or comment.
