Page 311 - Handbook of Electronic Assistive Technology

300 HANDBOOK OF ELECTRONIC ASSISTIVE TECHNOLOGY

it offers a guaranteed fail-safe method of calling for attention via a pager or similar device
if connection to the tablet, or the tablet itself, fails.
Planning should ensure that possible failures in the system are considered. It should be
clear what functions will be available if the main device fails, or if any other components
of the system fail. A failure mode should be in place that allows the input device to control
some function of the system or some other form of failure compensation, (e.g., restarting
the system or sending a distress signal to a pager).

Mode Switching
By the very nature of integrating functions, the user will need to switch between modes
or outputs. This could be achieved in a number of ways and it should be clear how mode
switching is achieved. There should also be a clear path between all modes (i.e., no ‘dead
ends’ out of which a user cannot switch).
In some cases, the mode switch may require a timing element (e.g., a long press or a
timeout). In these situations, the implications of this should be considered if there are
other time-sensitive controls in the system.
For switch access to mode changes the concept of a ‘½ switch’ is useful to consider
(Clayton, 1999; Judge and Colven, 2006) – this is a switch placed in a location that is pos-
sible for the user to access, but not the optimal location which would be reserved for the
primary function switches, and also potentially a location where the individual is unlikely
to accidentally press it.

Wheelchair Systems
When considering the use of wheelchair controls as part of an integrated system, con-
sideration needs to be given to the specific type of integration mode that the control sys-
tem offers – for example, whether it provides proportional or switched/directional mouse
output.
When aiming to achieve integration using a wheelchair controller with single switch
access, particular consideration needs to be taken around the mode switching and failure
mode. This setup will require either a timeout, long press or similar for mode changing and
these may in themselves clash with the output desired (e.g., a timeout will need to allow
an entire scan pattern to be completed on the target device, or a long press will need to be
compatible with the likely driving style of the wheelchair).

Models of Integration

Integration can be conceptualised in a number of ways – traditionally, it would be seen as
‘sending’ a single input (e.g., switch) to a number of different devices. While this model
still has relevance, the development of accessibility frameworks and more pervasive com-
puting has revealed a number of other possible models. The following section attempts
to broadly describe models of the different schemas for achieving integration. This is not
designed to be an exhaustive, or entirely future-proof, description of the different modes,