Paper Reading #1: Imaginary Interfaces: Spatial Interaction with Empty Hands and without Visual Feedback

Imaginary Interfaces: Spatial Interaction with Empty Hands and without Visual Feedback




By:
Sean Gustafson, Daniel Bierwirth and Patrick Baudisch.



Sean Gustafson is a PhD student under Patrick Baudisch working in HCI at Hasso Plattner Institute in Potsdam, Germany.
Daniel Bierwirth is a Co-founder of two companies that specialise Mobile applications. He holds an MSc in IT-Systems Engineering from Hasso Plattner Institute in Potsdam, Germany.
Patrick Baudisch is a Professor of Computer Science at Hasso Plattner Institute in Potsdam, Germany. He is also the Chair of their HCI lab. His research revolves around miniaturisation of mobile devices and their touch inputs.

Presented at UIST 2010.

Summary

In this paper the researchers are trying to achieve the ultimate level of miniaturisation possible by completely doing away with a screen. This system relies completely on the users short term memory and their ability to keep track of what they were "drawing" and be able to pick up where they left off with absolutely no feedback from anything but their mind. They used a few tests using random test subjects to prove that it was in fact possible for people to use such a super miniaturised device without any external feedback.


1. Using their non-dominant hand to create an L shape to be used as a point of reference, users were asked to try and draw 3 different styles of images from memory. They had to do Graffiti characters, repetitively draw a shape in the same spot and finally they had to draw a few diagrams that required more than one stroke. They hypothesised that users would have fewer errors with the graffiti and get more data on decreased accuracy when having to use multiple strokes for images. Over the period of this they were able to show that the users could recreate the objects that they had been asked to create, from their short term memory with a fair bit of accuracy. It was noted that most errors occurred during diagrams that required multiple strokes.

2. The 2nd set of experiment saw a new set of users. Again they used their non dominant had to create an L shape to use as a point of reference. Users were asked to draw an image, rotate 90 degrees and find a point on that image. Some were asked to not rotate and then asked to find a point on the image they'd drawn. As hypothesised, there were far fewer errors with users who did not have to rotate before selecting a point on their drawn image.

3. For third and final experiment the objective was to select points on an imaginary coordinate system using the non dominant hand as the axis. The hypothesis stated that there would be greater errors as the coordinate points got further away from the origin. Users were asked to make an L with the non dominant hand and then given coordinates to point to. As previously hypothesised the errors drastically increased as users got further away from the origin.

Considering the results of these tests, it was concluded that the system worked better when the user was stationary and worked best while things were still fresh in the users short term memory. As time increased, user memory faded and errors increased.

Discussion

While these tests may have proved the concept beyond doubt, its current implementation leaves much to be desired. A product like the one tested would probably be cheaper considering the lack of a screen, extremely portable, and the novelty of it would certainly make it very appealing. That being said, it relies far too heavily on a users ability to picture things and have a strong short term memory. In a consumer driven world where we're companies are working hard to make products simple and easy to use, a product that requires this much thinking and mental activity is immediately limiting itself to a niche market.

In addition to that, when considered from a reliability perspective this system just doesn't have the ability as currently implemented to compete. As technology improves, screens are liable to get cheaper and more energy efficient. While they may lack the novelty of a Star Trek like product, I feel the business community would greatly prefer a product that with guaranteed, tangible reliability as opposed to something that depends on how well I can remember images.