Biomedical Engineering Reference
In-Depth Information
workers against poisonous gases, e.g., carbon monoxide. The heroic achievement
is nowadays substituted by an electronic device that may provide warning in a
more effective manner. The electronic device may give a warning signal before the
poisonous gases reach an unhealthy level, that is at levels where the canary starts
to feel uncomfortable and dizzy.
The process of developing new and uncontroversial methodologies for ex-
ploring competitive artificial sensor systems, that are able to measure and iden-
tify complex qualitative parameters, with the aim to provide an additive consen-
sus with the corresponding human senses, may in no means be ambitious but is
still a challenge. Research has, for the last two decades, been exploring an inten-
sive perceptual-related development phase, presenting a technology “evolution”
in time, outgoing from a single gas (olfaction) sensor, e.g., commercially available
S n O 2 sensors provided by the Figaro Engineering Company as seen in Fig. 6.22.
This sensor and other types of sensors have made it possible to design smell sen-
sor arrays with a selectivity range that has made it possible to detect more com-
plex air bounded compounds. The development was perceptually followed by
a tongue (gustatory) sensor system, Winquist (1997), and further expanded with
an additional tactile sensor system by chewing detection, Winquist (2000). The
integration benefits experience an artificial sensing ability, that have been shown
in Sundic (2000) and Wide (1998). These papers investigate the fusion methods in
the integration process between an artificial taste and smell system.
The concept was further pushed to make an integration of the five sensors:
olfaction, gustatory, vision, auditory and kinetic parameters. In Wide (1999), an
electronic head comprising the five sensing abilities was presented as shown in
Fig. 6.13. The device is described as a virtual instrumentation, Wide (2001), in
providing qualitative estimation and decision-making of a dynamically changing
environment by combining data from different artificial sensor systems into a sin-
gle set of meaningful features. The single sensor information provided is of less
benefit than the aggregate of its contributing sensors.
Furthermore, the virtual feature estimation has been communicated to the
operator via a computer based face — awatar, able to express the overall impres-
sion of the tested object, Loutfi (2003), as can be seen in Fig. 6.14.
The artificial head concept has been used as a demonstrating platform by
using the various and optional number of perceptual sensor systems in specific
experiments. Tests have been performed in a number of preferably food-related
application with satisfactory results. However, after the methodology has been
demonstrated in different applications, there seems to be a huge and specific task
to verify the system performance and a general reliability to apply the technology
in industrial applications is needed. After all, an operator seems to be more trustful
in making complex and perceptual measurement tasks, even if the human aspect
of verification in advanced and non-linear application seldom is elucidated.
An interesting concept is considered in the development of tactile sensing
in an artificial hand, performing different gripping patterns. The industrial user
and research often state that the best robot hand for picking, holding and placing
Search WWH ::




Custom Search