Want to learn how to tie a tie? Dance? Take photos? Cook? Well all those things can be learnt just by heading on over to YouTube and watching some tutorial videos, and interestingly enough that’s one of the ways researchers at the University of Maryland Institute for Advanced Computer Studies are using to help teach robots how to do things.
Now before you think that this would mean that one day robots could replace chefs at a restaurant, it’s actually part of a bigger picture. The idea here is to show the robot the different processes involved in cooking, such as moving, picking up objects, observation, and etc., all of which is expected to eventually teach robots to better interact with people.
According to UMD professor of computer science and director of the UMIACS Computer Vision Lab Yiannis Aloimonos, “We chose cooking videos because everyone has done it. But cooking is complex in terms of manipulation, the steps involved and the tools you use. If you want to cut a cucumber, for example, you need to grab the knife, move it into place, make the cut and observe the results to make sure you did them properly.”
The goal here is also to allow the robot to be able to draw from its databank of steps in order to complete a goal, as opposed to having the robot follow a series of pre-determined steps verbatim. The end result is pretty encouraging and according to Aloimonos, “By having flexible robots, we’re contributing to the next phase of automation. This will be the next industrial revolution.”
Filed in ..