Friday, February 25, 2011

Paper Reading #13: SemFeel: a user interface with semantic tactile feedback for mobile touch-screen devices

Comments
 Joshua Penick - http://jip-tamuchi-spring2011.blogspot.com/2011/03/paper-reading-13-mouse-20.html
Shena Hoffmann - http://csce436-hoffmann.blogspot.com/2011/03/paper-reading-13-twinspace.html

Reference Information
Title: SemFeel: a user interface with semantic tactile feedback for mobile touch-screen devices
Authors: Koji Yatani, Khai Nhut Truong    
Presentation Venue: UIST 2010: 22nd annual ACM symposium on User interface software and technology; October 4-7, 2009; Victoria, British Columbia, Canada

Summary
In this paper, the researchers describe a tactile feedback system for touch-screen mobile phones called SemFeel that uses five vibration motors to help the user interact with the mobile device without having to look at it. The vibration motors are placed so that they can produce both single-point vibrations and vibrations that “sweep” across the mobile phone.

Taken from paper: the different vibration patterns
 In their prototype, the vibration motors are located on the back of the mobile device with the five motors forming a kind of cross shape (one at the top, middle, bottom, right and left) that allows for eleven different vibration feedback patterns. Since people tend to hold the mobile device such that their palm does not touch it, the researchers created a sleeve to fill the gap to allow the users to better feel the center motor.

The researchers did two studies with their prototype. In the first experiment, the researchers asked the users to hold the mobile device in their non-dominant hand and to identify the vibration pattern after the system generated a pattern. If the user chose the wrong pattern, they were informed which was the correct pattern. They found that users had the most difficulty identifying the counter-clockwise pattern. Other than the counter-clockwise pattern, the users were accurate on average 89.6% of the time.

In the second experiment, the users were asked to hold the mobile device with their dominant hand and use their thumb to input four numbers on the touch screen without looking. They studied three different conditions: one where no tactile feedback was given, one where only the center vibration motor was used and one where multiple tactile feedback was used. When a user tapped a number, it would appear on a computer screen. If they failed to type a number, an ‘X’ was shown. The researchers found that the users input numbers more accurately and more quickly with multiple tactile feedback than with single or no feedback.

Note: Prior to the actual experiments, the users were allowed to practice with the system.

The researchers conclude by giving different examples of where their SemFeel system may be useful such as to provide navigational aid to the visually impaired, to enhance current interfaces such as a maze game and to help a user interact with a calendar or calculator without looking.

Discussion
It seems I’m reading a lot of papers about mobile phones lately. I think this is one of the more interesting ones. They presented the material in a manner easy to understand and provided plenty of examples. Though I’m not sure how well this system will work in future applications, I do support anything that can help people keep their eyes off their cell phones. I’d rather people not do it, but since they’re going to anyway, at least this way, they stand a better chance of not running into someone or something.

One area of future study that the researchers mention would be to see how accurately the users can tell the difference between weak and strong vibration patterns in their prototype. Since one of the main reasons for this system is to allow people to interact with their phones without looking, they also mention doing studies where the users walk while interacting with the mobile device.

3 comments:

  1. I also liked this paper and the prototype. We take the vibrator in cell phones as a given today so it was refreshing reading about people trying to find a way to enhance this type of feedback.

    ReplyDelete
  2. Probably needs a little refinement before going into wide scale use. 89.6% is a great batting average but not so hot for interpreting input.

    I also have doubts about a touch screen with vibrating feedback winning the market against the phones with those little keyboards that you can flip out, although I'm hardly an expert.

    ReplyDelete
  3. I can see this being largely unhelpful. Many people hold their phones differently, which would result in not feeling some of the feedback. In addition, some of the feedback responses may be confusing to learn. Although this does seem like it could be useful in situations where you can't (or shouldn't) look at your phone.

    ReplyDelete