IEEE Technology and Society Magazine - September 2016 - 40

unaware or uninformed about the participatory critiques
of the positivist training we had absorbed as human computer interaction (HCI) researchers.
In 2004, we began a longer-term community intervention by establishing a computer lab in the Deaf community center. To support the facility, a senior
researcher and one or more postgraduate students visited the center every Wednesday to try out various technical research ideas and to solve technical problems in
the lab. The weekly visit pattern established in 2004
has been maintained to this day and has had a lasting
impact on building and maintaining trust between a
revolving cast of researchers (as students graduate and
new ones are recruited) and the Deaf community living
in proximity to that community center.
There have been several constants over the past
15 years. Firstly, the main players who run DCCT are
still in place. Secondly, the academic who initially introduced us is still working with DCCT, and with us, and
can act as an advisor, mediator, or "honest broker" if
relations become strained for any reason. Thirdly, two of
the academic supervisors continue to manage postgraduates working on the project at their respective local
universities. This long-term collaboration has resulted in
mutual trust and genuine friendships. Although weekly
visits could be seen as localized instances of bungee
research, the relationships underpin tolerance and willingness to find solutions associated with the inevitable
hiccups caused by bringing computer science students
with often limited social skills (and often different social
backgrounds) into a disadvantaged community.
As the relationships have developed, the nature and
focus of the weekly visit has also changed. After a couple of years, the Wednesday morning session became
recognized as the time that researchers visited, and
Deaf people started to avoid the lab at that time. Now,
instead of testing new technologies on Wednesdays, we
schedule other times for participant engagement and
research sessions. Instead, we attend weekly to support
accredited computer literacy training, and to assist the
Deaf technical support staff (see below).
We had thought that a monthly social gathering would
be an ideal opportunity to conduct surveys with larger
numbers of Deaf people, only to find that the Deaf people did not want to compromise the social nature of their
event. They preferred that we arrange other times that
did not overlap with those monthly sessions. We also
hired a small number of Deaf technical support staff to
keep the lab open for 5 or 6 days each week. As one of
the support staff began to assume more complex system administration duties, we deliberately avoided being
on site too often. If researchers were present, then we
would simply fix problems as he watched. Instead, we
provided email, SMS, and instant messaging support for

him during the week, and would only visit on Wednesdays or help solve catastrophic problems such as losing
network connectivity. This avoids undermining his local
status as system administrator and lab manager. We
also hired two Deaf assistants (only one of whom has
continued) who also collect statistics on how people are
using the computers, e.g., job hunting, watching videos,
email, etc. This supports both our research and the
DCCT management. The Deaf team developed its own
system for recording usage, including using a spreadsheet to relay the information to us. This regular reporting also provides regular communications operating in
parallel with bungee research visits.
Perhaps the biggest evolution over the years is that
the academic supervisor overseeing the whole project
(Tucker) has chosen to appear less often. In the early
days, he attended each week; now it is monthly or less.
The main reason for the supervisor to "back off" has
been to hand over more local responsibility and ownership to the Deaf community and more experienced students working on the project.
Currently the main assistive technology product that
has been developed with DCCT is SignSupport [14]. This
mobile phone app combines "packaged" videos of South
African Sign Language (SASL) with a simple UI navigation
and can be used to support structured conversational
scenarios. A demonstration system supports conversations between a pharmacist and a Deaf user when prescribing drugs (checking whether the user is pregnant,
taking other medications, explaining quantity, timing,
etc.). Sign Support has been successfully demonstrated
as a prototype, and a workbench has been developed to
specify and edit new types of structured conversation.
Next steps involve re-engineering the student prototypes
to bring the platform up to production standards, and
negotiating the (medical) regulatory and legal liability
issues around distributing such an application.

Rural Telcommunication Services in the
Eastern Cape
Whereas the Deaf community center is only a half hour
drive from our campus, the second project, called
Zenzeleni (see http://zenzeleni.net), is based in the
Mankosi community in the Eastern Cape, a 16-hour
drive, which takes even longer if the bus has to be used.
In this remote area, we are exploring affordable, and
solar powered, voice and Internet services in a rural
community. Over the years, we have significantly modified the way in which we visit the site, with a trend
towards fewer yet longer visits.
We began with very short visits following the operating procedure of the organization that originally
invited us to the site to explore applications in telehealth. Representatives of that organization would

IEEE Technology and Society Magazine

∕

september 2016

http://www.zenzeleni.net

Table of Contents for the Digital Edition of IEEE Technology and Society Magazine - September 2016