W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > January to March 1998

Re: Force Feedback Mouse for Blind Users (ref. Hardware for Bind Accountants)

From: Evan Wies <evan@immerse.com>
Date: Tue, 3 Mar 1998 16:15:21 -0800
Message-ID: <01BD46BF.9154E600.evan@immerse.com>
To: "'w3c-wai-ig@w3.org'" <w3c-wai-ig@w3.org>
My apologies for re-hashing a three-week old thread... I have been
tardy in responding to a couple comments on force-feedback mice.

Al Gilman wrote:
>First, I suspect that blind users will be better off with an
>absolute frame of references for the motion of the hand.  The
>conventional mouse is differentially coupled to the display
>cursor.  A tablet or touch screen is "DC coupled" to absolute
>display coordinates.  I suspect that blind users will be better
>off if the sandbox you prepare for their hand to play in is one
>which tracks position, not just accumulates changes in position.
>Second hypothesis: The tangible GUI approximates a touch-screen
>interface in resolution.  Don't try to put too fine a point on
>it.  This is the application class I think you are looking at as
>far as UIs for the blind and VI are concerned.

Evan Wies responds:
The FEELit mouse is actually an absolute device, so your wish is
granted.  We do tricks in software which dynamically changes the
scale of the workspace so that you get both fine-pitch control and
a large virtual workspace, despite a constrained physical workspace.
The balance between these two can be user-configured.

Al Gilman wrote:
>Third, I have a hunch that the real win from force feedback is in
>speeding up the process.

Evan Wies responds:
Here's some data to support your hunch.  We have conducted a formal
Fitts Task experiment with three conditions -- FEELit with forces
"snapping you to targets", FEELit without forces, and a conventional
mouse.  There was a 61% performance improvement going from FEELit without
forces to FEELit with forces and a 49% performance improvementgoing from
a conventional mouse to FEELit with forces.  With force-feedback, targeting
errors (where the user misses the target) were reduced 26% from a conventional
mouse and reduced 70% from a FEELit without forcces.  This is without any
long-term training on the FEELit (whereas users were well trained on the
conventional mouse).  

Charles McCathieNevile wrote:
>Coupling Al's first and third points (below), if the force feedback mouse 
>can give an indication of when the user is at the edge of the display, it 
>could be helpful. The feature that may be even handier is the ability to 
>expand the display to an arbitrary size. On my screen, I can open a 
>window that is bigger than the screen, and then have to, as it were, 
>position the screen over the section I want to see. A force-feedback 
>mouse could identify edges of a window bigger than the one I could see. 
>On the other hand, if it s possible to enlarge the window for ever, 
>things will get lost. This is the value of being able to get around the 
>display using a keyboard, which need not require a spatial model at all.

Evan Wies responds:
The FEELit mouse will be able to change its size dynamically to accomodate
different screen resolutions.  (We will also have to find a solution for
multiple monitors.)  This feature is not implemented now, but will be when
it hits the shelves.  I don't think we considered some of the finer points
of your comment, so thank you for your suggestion!


Evan Wies
Research Engineer
Immersion Corporation
Received on Tuesday, 3 March 1998 19:15:46 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 13 October 2015 16:21:00 UTC