Fault is a undesirable factor in any mechanical/pneumatic system. It affects the efficiency of system operation
and reduces economic benefit in industry. The early detection and diagnosis of faults in a mechanical
system becomes important for preventing failure of equipment and loss of productivity and profits. In this
paper, we present our ongoing research results on intelligent fault detections and diagnosis (FDD) on mechanical/
pneumatic systems. Using data from sensors and sensor network in an integrated industrial system, our
proposed FDD methodology provides the analysis of necessary sensory information (for example, flow rates
and pressure, as well as other digital sensor data) for the detection and diagnosis of system fault. In this experimental
study, the leakage of pneumatic cylinder was the "fault." It was shown that the FDD analysis was able to
make diagnosis of leakage both in location and size of the fault. In addition, the systematic fault and localized
faults can be detected separately. The proposed wavelet method gives rise to the fingerprint analysis to recognize
the patterns of the flow rate and pressure data - a very useful tool in intelligent fault detection and diagnosis.
Intelligent fault detection and diagnosis (FDD) depends on smart sensors which not only can render sensory
information but also can make easy the subsequent detection and diagnosis. At the heart of every intelligent
FDD system, there are sensors which work collaboratively with one another as well as the intelligent system.
Miniaturized sensors present unique advantages in facilitating the installation of sensory devices for the purpose
of diagnosis. In this paper, we present ongoing research on intelligent FDD and characterization using
miniaturized sensor. With miniaturization, sensors can be readily made and integrated into intelligent diagnosis.
Characterization and modeling of such innovative sensor designs are presented. Using new smart multi-function,
telemetric, and integrated sensors as "intelligent nodes" in systems will provide necessary sensory information
(for example, an integrated sensor which measures pressure, flow, and temperature--all in one module). The
characterization and studies of miniaturized sensor are presented, with practical relevance to applications. The
concept of intelligent nodes is further augmented with wireless sensors and system integration.