Day 1 - Haptics in Mobile Interaction - Paper

This was one of the more anticipated papers at CHI on day 1. The presenter started off with a summary of reasons why haptics is an important tool in mobile interaction, chief among his reasons being -
1. Visual and auditory senses are often impaired in a mobile scenario e.g. driving
2. On a small screen there is only so much information you can put.2
3. Some mobile devices like PDA's or cellphones stay on your body all the time.

The most powerful features of a haptic interface that they presented were
1. It could possibly use a unique tactile feedback for different menu items on a mobile device, so that users did not have to use their auditory/visual senses to interact. The potential applications of such a interaction are mind-boggling, imagine using this as a non-intrusive way of using your cell phone in meetings, receiving more specific alerts while you are talking on a call, and so on.
2. This device could be made location aware using GPS. Well nowadays everyone wants to jump on the gps bandwagon just because it is so easy to show an application, but in this case I felt a genuine use case, where tactile feedback could be used in conjuntion with gps input to direct the user towards a destination. Cool, very cool.

The presentation was good enough for me at this point already, but guess what the best was yet to come! To cement their work, they had done considerable research on different types of tactile feedback and what they meant to users.
They had used piezo-electric actuators which was basically an array of thin sheets all of which moved, as controlled by a wave. They created a stretch skin effect on the users finger which was then identified by the user based on factors like direction of wave, amplitude, speed etc. Broadly they made the following classification in the paper -
1. Primary distinguishing factors - Direction of wave and waveform being used.
2. Secondary distinguishing factors - Amplitude of wave and its direction.
Apparently the secondary factors were difficult to distinguish between especially as speeds of waveforms increased.

Based on a number of experiments it seemed promising that conclusions the paper had drawn were pretty good.

A good question was raised by an IBM researcher who had done previous work in this field. He pointed out that their experiments had concluded that user performance in distinguishing different haptic feedback deteriorates rapidly with increased stress level in the user, like while driving or crossing the street etc. As this prototype was still tethered to a PC, the tests were all done under optimal conditions. It will be interesting to see whether they face the same results as the IBM guys, and if so how do they alleviate them.

And yes, I forgot to mention this paper won the best of CHI award this year :).