This Robot Chef Has Mastered Crab Bisque : The Salt The system, a pair of robotic arms, learned to cook by mimicking the motions of a top chef. Even though it can't smell or taste, its maker says the robot should be able to make 2,000 meals by 2017.
NPR logo

This Robot Chef Has Mastered Crab Bisque

  • Download
  • <iframe src="https://www.npr.org/player/embed/400367891/400750556" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript
This Robot Chef Has Mastered Crab Bisque

This Robot Chef Has Mastered Crab Bisque

  • Download
  • <iframe src="https://www.npr.org/player/embed/400367891/400750556" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript

RACHEL MARTIN, HOST:

When you think of home cooking, you probably think of your dad or your mom, maybe your grandma - probably not a robot. That could change though. This week at a technology trade fair in Germany, a company called Moley Robotics unveiled a chef that will never nick itself with a knife, sneeze in the soup, or forget how long the turkey has been in the oven. But first, this robot has to learn a few recipes. And since Le Cordon Bleu isn't currently accepting robots, it's studying under Tim Anderson. He was the champion in the 2011 BBC "MasterChef" competition. He joins us now from London.

Hi, Chef Tim. Thanks for being with us.

CHEF TIM ANDERSON: Thank you for having me.

MARTIN: OK, before we talk about the process of teaching a robot to cook, can you describe what this robot looks like?

ANDERSON: Well, the robot is fixed into the ceiling of a kitchen unit. It's a modular kitchen, so it's basically two robotic arms attached to these very sophisticated robotic hands, which have the same range of movement as human hands do. So if you look at it when the arms are retracted, it just looks like a normal kitchen. But then they come down and it's almost like having two hovering hands that sort of float above the work surface and do the cooking.

MARTIN: So how is it learning? Because when I first heard this story, I thought, why do they need to teach a robot to cook? Can't you just program a robot to cook?

ANDERSON: That was my assumption as well, and that's how, you know, previous kitchen automation has worked. But basically how it works is I strap on some motion-capture gloves and wristbands and cooked in a kitchen studio with 3-D cameras set up. And I ran through the recipe five times. And then the programmers and technicians took the motions that I made and chose the smoothest and cleanest ones and spliced them together, fine-tuned them. And so now you have a robot that will repeat my exact movements in a very controlled way every single time.

MARTIN: What was the recipe that day?

ANDERSON: It's a crab bisque.

MARTIN: Did you just pick that out of the ether or, I mean, was there any rationale to that?

ANDERSON: Yeah. Well, we brainstormed - I think we came up with about six dishes to begin with that we thought would be a mixture of feasible, impressive, delicious and applicable to further recipes. So we tried a stir fry, we tried some sushi, we tried a steak. We tried a bit of pasta. And what we ultimately settled on was the bisque, 'cause it sort of ticked all the boxes.

MARTIN: So, I mean, how was it? Was the robot good - I mean, at doing this?

ANDERSON: It made mistakes at first, but all chefs do, I suppose. But now it's more or less foolproof. It makes a great bisque every single time.

MARTIN: Are you worried at all that you might be training your replacement?

ANDERSON: That did occur to me. It could happen. I'm not going to say it can't happen - someday. But it's a long way off if it does happen. And I'm just not sure it'll ever get to the same cooking capabilities that a human has. The main reasons being it's very, very difficult for the robots to deal with inconsistencies in ingredients, and also the robot has no creative capabilities. The robot can't taste, it can't smell. It can't really understand flavors, which means it can never really a put a dish together on its own.

MARTIN: That's a problem, yeah.

ANDERSON: So I don't think I'll be out of a job. And also, it's already given me a job because, you know, I'm just one of probably many future chef robot mentors that we're going to need.

MARTIN: (Laughter). Chef Tim Anderson speaking to us from London. Hey, Tim, thanks so much.

ANDERSON: No problem. Thank you.

Copyright © 2015 NPR. All rights reserved. Visit our website terms of use and permissions pages at www.npr.org for further information.

NPR transcripts are created on a rush deadline by Verb8tm, Inc., an NPR contractor, and produced using a proprietary transcription process developed with NPR. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.