I met a startup in Hong Kong, CafeX, who were working on a coffee-making robot, and my first reaction was that they were using cutting-edge technology to solve a trivial problem. I do like good coffee though, so I thought through the business case.
It seemed to fit into an awkward gap between "cheap coffee vending machine" and "cafe kiosk with trained barista". People grabbing a convenient coffee in places like gas stations and convenience stores, care about price over quality and will be fine with a cheap machine. People who like quality coffee will usually prefer a trained barista if one is available, unless they want to try the novelty of a robot.
The two niches I can see are:
1) Inside fancy office buildings, where the drones don't want to walk outside to get good coffee. (The robot is also good for impressing visitors).
2) Places without decent cafes around, i.e. shopping malls, theme parks, airports, where the robot might be as good or better than Starfucks burned-for-consistency beans.
That's not why Starbucks is so labor-intensive.
Starbucks developed a good automated coffee-making machine, but decided not to deploy it. Instead, they built a lower-profile manual machine which allows their employees to maintain eye contact with the customer. The whole point of Starbucks, the thing that justifies their high prices, is the ego boost the customer gets from making the barista perform. Starbucks is about sucking up. All their employees know this; it's in their training. Starbucks is a fast food operation with the sucking-up level of a sit-down restaurant.
Read what their CEO has to say on the subject.[1]
[1] http://www.kplu.org/post/robot-baristas-heres-starbucks-ceo-...
I seem to recall (and I can't remember where I saw it) that Heston Blumenthal uses a very expensive automated coffee machine to remove the variance inherent in human baristas.
http://www.theguardian.com/lifeandstyle/wordofmouth/2013/apr...
The tea-making robot seems to come with overoptimistic numbers.
The sandwich-making robot is a good idea, but it will need a complete redesign to make it cleanable. All stainless and Teflon, no food trap points, machine-washable, steam-cleanable. Whether it's cost effective is another issue. That idea has been tried many times. Here's AMFare, from American Machine and Foundry, in 1964.[1] That system worked quite well, but needed a constantly busy fast food outlet to justify the expense.
Then there's the version for the 1%[2].
[1] https://www.youtube.com/watch?v=FmXLqImT1wE [2] http://factor-tech.com/robotics/17437-robot-chef-that-can-co...