Humanoid Robot KOBIAN Learning to Be a Comedian

This robot will make you laugh, at least if you speak Japanese and can understand its jokes

3 min read
Humanoid Robot KOBIAN Learning to Be a Comedian
'Funny, who me? Nooo...'
Photo: Takanishi Laboratory/Waseda University

We've seen lots and lots of robots making people laugh. Some robots are just doing funny things like dancing. Others are just doing silly things like falling. But as far as successful robot comedians, well… I could count the number of those on my 10 hands.

Just a little binary joke to get you warmed up there, folks. Thank you, I'll be here, uh, until I'm not.

Anyway, how should a robot be deliberately funny? Japanese researchers have read the books and done the experiments, and last week at the 2014 IEEE International Conference on Robotics and Automation (ICRA) they presented a paper with their findings.

So. How do you program a robot to make a human laugh? It's in the paper, of course:

C. Methods for making humans laugh

Part of the methods for making humans laugh are published by comedians or researchers. In order to achieve a robot's behavior for making humans laugh based on these methods, we studied 6 books[16-21] in which the methods for comedy and the ways for funny conversations are gathered comprehensively. These methods in the books are picked up from wide range of skits or ways of expressions, but most of the methods are commonly the same between other contents. We extracted the methods that are especially common in these books and show them below.

1) Funny behavior

  • Overblown: Exaggerate too much
  • Equivoque: Homonym or parody
  • Blue jokes: Make a dirty joke
  • Sympathetic story: Saying the empathic small things that almost everyone has experienced once
  • Induced laugh: Comedian laughs aimed to the audiences' infectious laugh

2) Funny context

  • Running gag: Use the same story again and again
  • Unexpected: Do completely unexpected behavior

3) Funny character

  • Self-flattery, Self-deprecating humor
  • Imitation

Seems clear enough, right? Based on these techniques, the Waseda University researchers, led by Professor Atsuo Takanishi, came up with a variety of sketches for their KOBIAN humanoid robot to perform.

Here's a video showing some of the jokes. Be warned: a lot of the humor here is translated (ish) from Japanese, and/or based on people or skits that are famous in Japan, making it simultaneously incomprehensible to most of us and way funnier than it might be otherwise:

The researchers recruited volunteers to determine if they found KOBIAN funny, wiring them up with EMG sensors and accelerometers and pointing video cameras to their faces to detect smiling and laugher.

According to the following graph, the volunteers found a few of the routines quite funny, although in many of them KOBIAN's attempt at humor fell flat. 

Still, the researchers say robot comedy can make people feel better, and they were able to quantify that with their experiment. Both before and after watching KOBIAN, the volunteers were asked to complete a questionnaire on their mood. The researchers used a psychological test called POMS, or profile of mood states, used to assess how people are feeling.

The results show that the mood of the volunteers seems to have improved after watching KOBIAN, as indicated by POMS scores associated with negative feelings like stress and anger, which the researchers say "significantly decreased." 

The researchers are now planning to create more complex skits, combining different comedy techniques, and conducting experiments with a large number of subjects. They also want to examine "the effect of embodiment of the robot to the reaction of the subjects, and to feedback the humans' laugh reaction to the robot to achieve the interaction between a robot and a human through laughing."

We don't quite understand what that means, but we're pretty sure the result will be hilarious.

"Bipedal Humanoid Robot That Makes Humans Laugh With Use of the Method of Comedy and Affects Their Psychological State Actively," by T. Kishi, N. Endo, T. Nozawa, T. Otani, S. Cosentino, M. Zecca, K. Hashimoto, and A. Takanishi from Waseda University in Japan was presented at ICRA 2014.

The Conversation (0)

How the U.S. Army Is Turning Robots Into Team Players

Engineers battle the limits of deep learning for battlefield bots

11 min read
Robot with threads near a fallen branch

RoMan, the Army Research Laboratory's robotic manipulator, considers the best way to grasp and move a tree branch at the Adelphi Laboratory Center, in Maryland.

Evan Ackerman
LightGreen

This article is part of our special report on AI, “The Great AI Reckoning.

"I should probably not be standing this close," I think to myself, as the robot slowly approaches a large tree branch on the floor in front of me. It's not the size of the branch that makes me nervous—it's that the robot is operating autonomously, and that while I know what it's supposed to do, I'm not entirely sure what it will do. If everything works the way the roboticists at the U.S. Army Research Laboratory (ARL) in Adelphi, Md., expect, the robot will identify the branch, grasp it, and drag it out of the way. These folks know what they're doing, but I've spent enough time around robots that I take a small step backwards anyway.

The robot, named RoMan, for Robotic Manipulator, is about the size of a large lawn mower, with a tracked base that helps it handle most kinds of terrain. At the front, it has a squat torso equipped with cameras and depth sensors, as well as a pair of arms that were harvested from a prototype disaster-response robot originally developed at NASA's Jet Propulsion Laboratory for a DARPA robotics competition. RoMan's job today is roadway clearing, a multistep task that ARL wants the robot to complete as autonomously as possible. Instead of instructing the robot to grasp specific objects in specific ways and move them to specific places, the operators tell RoMan to "go clear a path." It's then up to the robot to make all the decisions necessary to achieve that objective.

Keep Reading ↓ Show less