| |
5.2.
The Sensor-Opportunities-Based Approach
67
2001
] have been explored in di
erent project. Coarse Location can be also gained
through passive-infra-red sensors, mechanical switches or IR-barriers.
For On Object and On
Human outdoors the primary technology choice is GPS1.
However, also humidity, inertial
[ Vildjiounaite et
al. 2002 ]
, and pressure sensors
can also deliver location information, but on a lower and more coarse level. The
variety of location systems based on Mutual
Collaboration is huge: di erential GPS,
ultra-sound, radio etc. Sensing Bio
Signs/Emotions with In Environment sensor-settings is di
cult: [ Do-
nato et
al. 1999 ]
and [
Fernandez and Picard 2003
] report vision and audio for
reasoning on user’s Bio
Signs/Emotions. Augmented objects measuring force and
touch [
Ark et
al. 1999 ]
can give some hints about Bio
Signs/Emotions. However, most promising are On
Human measurements such as
[ Healey and Picard 1998
] [
Michahelles et
al. 2003c ]
. Activity can be well sensed with special purpose system, such as commercially avail-
able smart white boards. Load-cells, passive infra-red, pressure and capacity sensors
can be used for low-level detection only. On
Human sensing has been well explored
for motion activity [
Farringdon et
al. 1999 ]
. Location system can give hints reasoned
from semantical location descriptions.
Interaction among humans has not been explored well yet. In
Environment sensing systems based on vision, load-cells and audio could help to perceive characteristics
of interaction, such as collocation, gestures or speech. The On
Object field is blank as objects are not involved here. For On
Human the same sensors can be used as
for Activity if measurements are correlated among interactors. Location system do
not really help here, as collocation is not significant for interaction.
5.2.2 Discussion
As a result of the review presented in the previous section, this section discusses the
appropriateness of di erent sensor placement for the six sensing dimensions. Table
5.2 summarizes how well di
erent sensor placements work for di erent sensing tech-
nologies classified by the six dimensions of sensing (see Section 5.2.1). In particular
we di erentiate between not
applicable, if a combination does not make sense, pos-
sible for instances with very low quality of perception, and good and very
good for more promising solutions.
One finding is, that In
Environment placement is the primary choice for ID sensing.
As our focus is on human sensing it is not surprising that On
Object is well suited for 1Actually
GPS is a mutual technology requiring a receiver in collaboration with satellites
in space. However,
as satellites are so ubiquitous and invisible anyway we consider them as a ”natu-
ral”resource and view the receivers
only.
|  |
|
| |
|
|
