Biomedical Engineering Reference
In-Depth Information
casualties in chemical production. Workers seem to remain on site until
the explosion occurs. This is why the engineers have designed a safety
system. This system is an alarm whose goal is to predict a mean remain-
ing time (MRT) before the explosion occurs. First, a detection algorithm
was developed and tested in an experimental situation. It was then sug-
gested that a prototype of the artifact be introduced in a tryout situation.
This prototype displayed (1) the mean remaining time before an explosion
and (2) very precise information regarding the temperature of the prod-
uct, to within a hundredth of a degree Celsius.
The results of the introduction of the prototype on the experimen-
tal site speak for themselves. They show that the workers gradually con-
sulted the alarm more and more. Thus, in the first trials, they consulted
the interface for only 1.7 per cent of the time devoted to the supervision
of the production process. However, 3 months later, they consulted it for
31.5 per cent of the time (see Table 10.1). Therefore, the workers took own-
ership of the system, which is a priori
a good thing.
However, a more detailed analysis of information collection (based on
gaze direction) showed that the workers use the alarm instead of the ther-
mometers that were previously available (see Table 10.1). Thus, the workers
collect mainly the information that the device gives about temperature,
and not the time remaining before the explosion (MRT), although this is
the whole point of the system. In short, in the activity of the workers, the
alarm takes on the status of a thermometer.
However, this status is unacceptable from the point of view of the
designers. First, European regulations (NE 31) require that 'safety instru-
mented systems' should be separate from the other available tools.
Whereas the alarm was initially designed as a safety instrumented sys-
tem, it takes on the status, in the activity of the workers, of an 'aid to
Table 10.1
Comparison between the Evolution of the Time Devoted
to Information Collection Using the Prototype and Using
the Thermometers That Were Previously Available on Site
a
between the First, Second and Third Trial Sessions for the Prototype
First
trial session
Second
trial session
Third
trial session
Duration of the collection of
information on the prototype
1.7%
8.1%
31.5%
Duration of the collection of
information on the temperature
indicators that were previously
available
27.3%
24.5%
4.2%
a
In per cent of the total duration of work.
