Authors
Jonghwa Kim, Stephan Mastnik, Elisabeth André
Augsburg, Germany
Summary:
This article outlined a EMG-based hand gesture control for an RC car. They used three nodes placed on the users arm to recieve EMG input from the user based on the users gesture movement.
The user would use hand gestures to tell the car to go fowards(gesture 1), right(gesture 2), left(gesture 3), or to rest(gesture 4) .
The user would recieve training on how to use the device before they were given the chance to actually use it. The motions would be recieved and then processed by a middle man before it was sent to the RC car.
The researchers, in a user study with 30 subjects, found that the users were able to quickly become very good at controling the car using this type of control system. It is important to note that the users did recieve training on how to use the gestures and on proper movements to achieve better results. It is explained later in the paper that while this training is not ideal, a good system using training is a step in the direction of not needing any training.
Discussion:
On a level of one to cool this thing is awesome. Just the theory of being able to control something with electrical pulses from your brain is very intriguing. I would be worried about depolying a system like this in the real world for fear of electrical noise but it seems like an interesting idea, assuming they can limit the effects of the electrical noise.
I was really surprised that the users were able to catch on as quickly as they were. When looking at a system like this that is seemingly so new we wouldn't expect such a quick catch on. On the other side however, the gestures are much more natural than a joystick so that might have helped.
I am looking forward to controlling my personal assistant robot with my mind so I would highly suggest that the researchers continue with their work in this field.
No comments:
Post a Comment