James The Robot Bartender Knows When You Want A Drink

James The Robot Bartender  will chat, serve and recognize when you want drinks thanks to the partnership of a team of professors from Germany, Scotland and Greece.

The birth of robot bar tenders isn’t particularly newsworthy as there have been other attempts before James. However James, the Joint Action in Multimodal Embodied Systems robot is different to its predecessors due to his social skills.

People are Messy

While it’s been proven that robots are more than adept at mixing a whole host of drinks, the reality of them functioning as well as a human bar tender in a real life situation has always been unlikely. As a rule, robots like clean orderly situations. The chaotic atmosphere of the average pub serves to confuse robots. Robot bartenders in the past have only been able to mix drinks for patrons that have expressly ordered the drinks via a smart device. In reality though, patrons don’t like to use devices to order drinks, they prefer to use the traditional signals to indicate that they want a drink.

Determining an Order

During the project, the European team decided that in order to be able to teach James the social skills necessary for a bar tender, they would have to research people’s behaviors when ordering drinks in pubs.

Cameras were placed in a number of pubs throughout England and Germany. When the footage was watched, the team discovered that 90% of people who wanted a drink would simply move closer to the bar where they could stare directly at the bar tender who picked up on the signal and took the order.

For robots, the subtlety between somebody leaning against the bar and somebody staring over the bar at the bar tender is difficult for them to process. The team had to enable James the robot bar tender to process the information that people were presenting through body language and react in the appropriate way in a split second.

A new Bartender is Born

The results, published in Frontiers of Psychology show that progress is definitely being made. The project runs until January and the team hopes to bring robotics forward to the point that robot-human interaction is far more natural.