Biomedical Engineering Reference
In-Depth Information
to resolving the alarms ('resolve alarms once all the required actions have
been performed') and (2) the ease and speed of detection of incoming
alarms, because of the contrast between the old alarms (which have been
'beiged out') and new alarms (which are coloured red or magenta).
The interviews carried out with the five supervisors in group 2 pro-
vide some additional, most interesting information about the process
of instrumentation. One of the supervisors claimed he 'preferred not to
resolve any alarms to keep them around … to always have them in sight'.
Before we go on, however, what are the consequences of failing to
resolve some of the alarms on the 'alarm manager' window? Two points
should be made here:
1. Unresolved alarms continue to be displayed at the bottom of the screen.
Because the alarms are displayed in chronological order, the more
recent alarms are displayed in the bottom-most part of the screen.
2. When the alarms are resolved, they are displayed above the unre-
solved alarms. Upon arrival, new alarms are therefore displayed on
the screen just above the 'conserved' alarms. There are typically few
alarms of this kind (between three and eight alarms for the entire
duration of the workday, for each supervisor in group 2). The activity
of the supervisors does not seem to be hindered by this. There is no
confusion between the unresolved alarms, whether they have been
processed or not. The supervisors act only on new alarms. The verbal
utterances of the workers in group 2 suggest that what is important to
them is to be able to tell alarms apart. Some of the alarms are resolved,
and those that are unresolved remain at the bottom of the screen.
Therefore, the C type schema leads to keeping some alarms grouped
in a specific location. The alarms that are viewed as particularly important
(or critical) are therefore highlighted, which helps to supervise and follow
them. Some other elements seem to confirm and extend this analysis.
First, in-depth interviews and the analysis of work diaries - the docu-
ments where significant events having occurred in the course of work are
recorded - show that there is a correlation between the alarms that are
'set aside' and those that are recorded in these diaries. The latter alarms,
generally, are alarms that have proven to be impossible to resolve via
remote actions, and in some cases led to setting off a maintenance opera-
tion or specific processing. Approximately 9 out of 10 'conserved' alarms
were also recorded in the diaries. Furthermore, because the supervisors
worked continuous shifts, there were handovers and, consequently, some
exchanges of instructions. Many instructions were exchanged between
the two workers, who relayed to one another. Thus, as shown in Table 11.2:
(1) when the worker finishing his shift belonged to group 1, he relied
