@Thorne has covered most of your questions admirably. My only objection are on the subjects of sleep and death.
Simulating sleep while permanently mounted batteries get recharged is a tremendous waste of time and productive potential. Robots will have interchangeable batteries with secondary rechargeable power sources to keep their brains active during the battery transition process.
Also Robotic death would only occur by choice. Regular backups, snapshots of their living consciousness, would make accidental permanent death extremely rare. Similarly, aging and obsolescence will not be real for robotic life. All of their body parts, including their brains can be replaced with newer models as needed.
On the OP's invitation for unique ideas...
Whenever someone brings up the idea of artificial sentience, the most interesting aspect for me is that it is un-prove-able. As conscious, thinking beings, each of us can be pretty certain of our own sentience, but even among a 100% natural human populace, we can never be sure of the sentience of anyone else. Are the people around you right now independent sentient beings, or are they automatons, mindlessly playing their prescribed roles in a play for your singular observation?
This paranoia-seed will become even more insidious once we share the world with sentient-behaving devices of our own creation. At least in an all-human world, we can pacify our fears with the idea that all of those other humans were created by the same miraculous process that gave us our own life. In the AI future, into which we may soon arrive, such healing logic may loose much of its power. Machines of our own creation, will interact with us and our world will more skill and grace than homo-sapiens have ever expressed; but are those beautiful, brilliant machines alive?
...and once we question the "aliveness" of the silicon super-geniuses which have spared us from all unpleasantness in our post-scarcity world, how much easier is it to write off the "aliveness" and fundamental rights of the human people around us? The arrival of AI may very well herald a twilight for the human ideals of morality, equality and kindness.
Don't fear the monsters with the silicon brain whose morals have been hard-coded into them. Natural, un-programmable carbon brains are ultimately much more dangerous!