A robot developed by researchers at Cornell University could someday make your friendly neighborhood Starbucks employee redundant, as it has learned how to autonomously make an espresso using crowdsourced instructions provided by Internet users. We would never have the problem of misspelled names on cups again. (Seriously, how do you get Jannett from Shannon?! Come on guys!)
Officially known as PR2 but affectionately dubbed Robobarista by its creators, the machine has progressed from pouring hot coffee over ice cream to make an affogatto coffee to the somewhat more complex latte, which also requires the use of steamed milk, according to CNET.
Step 1: Get the robot acquainted with your kitchen
The Cornell University team behind Robobarista has designed it to follow verbal commands as well as collect information sent in by volunteers that tell it how to use objects that it had never previously encountered, such as an espresso machine. Once it had enough know-how to perform such tasks, it was able to do so on its own, extrapolating the actions that needed done.
“In order for robots to interact within household environments, robots should be able to manipulate a large variety of objects and appliances in human environments, such as stoves, coffee dispensers, juice extractors, and so on,” the researchers explained on the robot’s official website.
“Consider the espresso machine,” they added. “Even without having seen the machine before, a person can prepare a cup of latte by visually observing the machine and by reading the instruction manual. This is possible because humans have vast prior experience of manipulating differently-shaped objects. In this project, our goal is to enable robots to generalize to different objects and tasks.”
Makin’ the cup of joe
CNET uses the example of a toaster to explain the concept. While toasters aren’t all the same, people generally know what constitutes a toaster, and generally how it is used. Robots, however, function on pure logic and lack the intuition to deal with differences in appliance models. Thus, they need to have access to a through database to figure things out – hence the crowdsourcing.
Once PR2 had collected enough information about various objects, it was able to teach itself how to actually make coffee using an English-language instruction manual, CNET said. When it saw the instruction “push down on the handle to add hot water,” it knew that it had to locate a handle on the unfamiliar espresso machine using its knowledge of a similar-looking urinal handle.
Step by step it went through the instructions, referencing its database and comparing objects that were similar in order to produce a latte without any direct human intervention. The robot is still learning how to perform different tasks and add new drinks to its arsenal as well, so if you want to help out, you can visit the project’s home page and drop some knowledge on the Robobarista. It might be some time before PR2 can create a grande, iced, sugar-free, vanilla latte with soy milk.