This pancake flipping robot learns how to flip pancakes the old fashioned way, by having someone hold its hand - while it grasps the pan - and then perform the flipping motion for it, so it can get a feel for what it's like. (Note, however, that motion capture is also used - a technology available in relatively few kitchens.) Take a look at this video of a Barrett WAM 7 DOF (degrees of freedom) robot manipulator learning how to flip pancakes.
(Pancake flipping robot)
The skill is first demonstrated via kinesthetic teaching, and then refined by Policy learning by Weighting Exploration with the Returns (PoWER) algorithm. Compared to policy-gradient approaches, the reward is treated as a pseudo-probability, which allows Reinforcement Learning to use probabilistic estimation methods such as Expectation-Maximization (EM).
A similar kind of process is used by the usuform robot bartender in Anthony Boucher's 1943 short story Q.U.R.. Can a robot bartender make a perfect Three Planets drink?
Quniby said, "Three Planets," and he [the robot] went into action. He had tentacles, and the motions were exactly like Guzub's...
...I got one of those new electronic cameras - you know, one thousand exposures per second... So we took pictures of Guzub making a Three Planets, and I could construct this one to do it exactly right down to the thousandth of a second. The proper proportion of vuzd, in case you're interested, works out to three-point-six-five-four-seven eight-two-three drops. It's done with a flip of the third joint of the tentacle on the down beat.
(Read more about Boucher's robot bartender)