Today, several advanced transducer device and emerging signal processi
ng technologies are combined to establish the important area of so cal
led 'intelligent sensors'. They offer new or improved operational and
processing capabilities. However, as modern technical systems get more
and more complex with increasing requirements on system reliability a
nd availability, appropriate systemarchitectures have to be designed,
which will have strong impacts on kind and quality of signal processin
g tasks to be performed by intelligent sensors. Regarding this back-gr
ound motivation, this paper will provide to practitions some hints and
explanations for processing uncertain information by utilizing basic
fuzzy set theory. The proposed applications include uncertainty propag
ation, self-calibration, human-machine-interfaces, and semantic classi
fication in fuzzy combinatorical networks. The presentation emphasis o
n the relation between practical issues and appropriate modeling and p
rocessing of information with fuzzy techniques. The industrial applica
tion to an intelligent sensor for detection of oil pollution in water
is described.