Reviwer 1
Well written paper. Units (A V s N ...) should not be
written
in italics.
Could you provide values for a b c d in Eq. (7)?
----------------------
Reviewer 2
The paper presents an interesting study on the hybrid
prognostics approach for MEMS. The approach uses a
combination of data-driven failure model with a particle
filter to learn and estimate the system's state. The paper
has a logical structure and is easy to follow. Overall,
it will be a valuable contribution to the conference.
There are, however, some important limitations that need to
be addressed by the authors. The difference in degradation
data between the studied specimen is not addressed by the
authors, esp. MEMS3. In this regard it would be interesting
to see mean health indicator (HI) graph with 95% CI
following the lifetime tests. Likewise, RUL estimation
graph, which is shown in Fig. 9 (b) needs to be presented
for all specimen to better estimate the precision of the
model.
Below are additional minor comments to the authors:
Provide a reference to the definition of reliability
presented in section 2
Provide a reference to the specific MEMS valve model, if it
is commercially available
Did the authors ensure temperature stability in the lab for
the accelerated degradation test?
Section 5.3 states that measurements were collected every
day. Yet fig. 7 shows features of lower frequency, e.g.,
see graph for MEMS3
Eq. 7 - dots instead of multiplication?
Please state what values of a, b, c and d parameters were
used to obtain results in section 5.4
The following sentence may contain an error: Furthermore,
the real RUL is still in the 95% prediction interval at
different time steps.
Ensure that all subfigures of fig. 7 have the same
horizontal and vertical scale
Same for fig.8 and fig. 9
------------------
Reviewer 3
The paper topic is really interesting and it fits well to
the workshop topics. I would like to ask 2 minor reviews.
Just some minor amendments are required. The first two
figures are not clearly readable, with too small fonts. The
references are not in the required format [ref#] and should
be corrected.