Humans are asked help robots recognize the multitude of objects found in the average home.
Swedish researchers are asking people to use your Xbox Kinect sensor as a scanner to make detailed pictures of things in 3D at home.
The Kinect @ home project requires mass participation accumulate many examples of common household objects.
The examinations will be based on an object library robots can consult as they navigate around the houses.
Coordinator Alper Aydemir said: "The factory floors can be tailored and tools used robots can know precisely in detail This is not the case of the spaces of everyday life and objects.".
While humans have no trouble recognizing objects such as a cup of tea, even if it is a different color shape and size to those we have seen before, fighting robots to complete a mundane task.
"One of the best ways for robots to perform these tasks is to learn to recognize a sofa, chair or refrigerator feeding them a lot of data," said Mr. Aydemir the BBC.
Instead of building the database objects themselves, the Center team Autonomous Systems at the Royal Institute of Technology Sweden have returned to help the many people who bought a Kinect sensor for Xbox games.
The Kinect sensor uses a combination of an infrared sensor, camera and personal computer chip to detect and interpret the movements of the players, allowing them to play without the traditional controller manual.
The robots are coming out of the factories and homes
Mr Aydemir and colleagues Rasmus Goransson and Professor Patric Jensfelt have created software that uses the information captured by the infrared camera, and collected through the website Kinect @ home, to create the analytical method.
"This way, people get easy access to increasingly better the state of the art 3D modeling techniques for free and viewpoints regarding researchers learn what works and what does not by the use of the data, "he said.
The project has been running a few days and already 141 items, including shoes, shirts and guitars have been scanned.
With the Kinect as a 3D scanner was as easy as making a video for YouTube, Mr. Aydemir said.
Persons wishing to participate must install an add-on for the web browser, he said, and take a little care when scanning objects.
"When data capture, must move very slowly and avoid big gaps," he said. "Other than that, it's as easy as clicking a button on the registration page."
The trio are planning to take their data widely available to researchers robot and computer vision.