Sign In

Communications of the ACM

ACM TechNews

Crowdsourced Online Learning Gives Robots Human Skills

View as: Print Mobile App Share:
Service Robot

Photo courtesy of Javier Pierini/Getty

Roboticists are experimenting with using crowdsourcing to teach robots more general skills. By allowing users to pilot real or simulated robots over the Internet in trial experiments, the researchers hope to create machines that can simulate a human's flexibility and dexterity.

"Crowdsourcing is a really viable path toward getting robots to do things that are useful for people," says Brown University's Chad Jenkins.

Crowdsourcing also can be used to develop better human-robot interactions, says Worcester Polytechnic Institute's Sonia Chernova. She has led a team that developed Mars Escape, an online game in which two users each control an avatar, one human and one robot, to collect information on teamwork, social interaction, and communication.

After more than 550 game sessions, the researchers looked for patterns in the data, such as methods that players frequently used to retrieve objects, and phrases they exchanged when doing so. The researchers then set up a mock real-life version of the game in which visitors were paired with a robot powered by software based on the Mars Escape data. During testing, most of the visitors said the robot behaved rationally and contributed to the team's success.

From New Scientist
View Full Article

Abstracts Copyright © 2011 Information Inc. External Link, Bethesda, Maryland, USA 


No entries found

Sign In for Full Access
» Forgot Password? » Create an ACM Web Account