Sensors,
Journal Year:
2024,
Volume and Issue:
24(4), P. 1236 - 1236
Published: Feb. 15, 2024
In
the
era
of
Industry
4.0
and
5.0,
a
transformative
wave
softwarisation
has
surged.
This
shift
towards
software-centric
frameworks
been
cornerstone
highlighted
need
to
comprehend
software
applications.
research
introduces
novel
agent-based
architecture
designed
sense
predict
application
metrics
in
industrial
scenarios
using
AI
techniques.
It
comprises
interconnected
agents
that
aim
enhance
operational
insights
decision-making
processes.
The
forecaster
component
uses
random
forest
regressor
known
aggregated
metrics.
Further
analysis
demonstrates
overall
robust
predictive
capabilities.
Visual
representations
an
error
underscore
forecasting
accuracy
limitations.
work
establishes
foundational
understanding
for
behaviours,
charting
course
future
advancements
components
within
evolving
landscapes.
Sustainability,
Journal Year:
2024,
Volume and Issue:
16(19), P. 8336 - 8336
Published: Sept. 25, 2024
This
paper
explores
new
sensor
technologies
and
their
integration
within
Connected
Autonomous
Vehicles
(CAVs)
for
real-time
road
condition
monitoring.
Sensors
like
accelerometers,
gyroscopes,
LiDAR,
cameras,
radar
that
have
been
made
available
on
CAVs
are
able
to
detect
anomalies
roads,
including
potholes,
surface
cracks,
or
roughness.
also
describes
advanced
data
processing
techniques
of
detected
with
sensors,
machine
learning
algorithms,
fusion,
edge
computing,
which
enhance
accuracy
reliability
in
assessment.
Together,
these
support
instant
safety
long-term
maintenance
cost
reduction
proactive
strategies.
Finally,
this
article
provides
a
comprehensive
review
the
state-of-the-art
future
directions
monitoring
systems
traditional
smart
roads.
Scientific Reports,
Journal Year:
2025,
Volume and Issue:
15(1)
Published: April 21, 2025
Abstract
Multi-channel
sensor
data
often
suffer
from
missing
or
corrupted
values
due
to
failures,
communication
disruptions,
environmental
interference.
These
issues
severely
limit
the
accuracy
of
intelligent
systems
relying
on
integration.
Existing
restoration
techniques
fail
capture
complex
correlations
among
channels,
especially
when
losses
occur
randomly
and
continuously.
To
overcome
these
limitations,
we
propose
an
autoencoder-based
recovery
algorithm
that
recursively
feeds
reconstructed
outputs
back
into
model
progressively
refine
estimates.
A
dynamic
termination
criterion
monitors
reconstruction
improvements,
automatically
stopping
iterations
further
refinements
become
negligible.
This
recursive
input
strategy
significantly
enhances
computational
efficiency
compared
conventional
single-step
methods.
Experiments
multivariate
datasets
show
proposed
method
outperforms
one-time
autoencoder
maintains
robust
performance
across
diverse
scenarios.
approach
provides
a
scalable
adaptable
solution
ensure
integrity
in
networks,
enabling
improved
reliability
operational
industrial
technological
applications.
The International Journal of Advanced Manufacturing Technology,
Journal Year:
2024,
Volume and Issue:
unknown
Published: Aug. 29, 2024
Abstract
High-pressure
die
casting
(HPDC)
is
a
permanent
mold-based
production
technology
that
facilitates
the
of
near
net
shape
components
from
nonferrous
alloys.
The
pressure
and
temperature
conditions
within
cavity
impact
cast
product
quality
during
after
conclusion
filling
process.
Die
surface
sensors
can
deliver
information
describing
at
die-casting
interface.
They
are
associated
with
high
costs
limited
service
lifetimes
below
achievable
total
cycle
count
inserts
therefore
ill-suited
for
industrial
use
cases.
In
this
work,
suitability
long
short-term
memory
(LSTM)
recurrent
neural
networks
(RNN)
substituting
physical
virtually
ramp-up
or
end
sensor
life
investigated.
Training
LSTMs
data
233
cycles
different
process
parameters
provides
which
then
applied
to
99
further
cycles.
prediction
accuracy
investigated
time
interval
lengths
in
solidification
cooling
phase.
For
longer
intervals,
deteriorates,
potentially
due
highly
individual
hardly
ascertainable
buildup
distortion
internal
stresses.
Overall,
however,
developed
excellent
temperatures
good
pressures.
Journal of Applied Mathematics,
Journal Year:
2024,
Volume and Issue:
2024(1)
Published: Jan. 1, 2024
The
expected
healthcare
(HC)
inflation
rate
(IR)
(HCIR)
is
an
important
variable
for
all
economic
agents
within
HC
systems.
In
recent
years,
during
the
COVID‐19
pandemic,
Iran
has
experienced
a
high
HCIR
in
its
health
system.
this
context,
robust
approximation
of
will
be
helpful
tool
authorities
and
other
decision
makers.
Using
monthly
time
series
data
Iran,
we
developed
various
forecasting
techniques
based
on
classical
smoothing
methods,
decomposition
ETS
(error,
trend,
seasonality)
approaches,
autoregressive
(AR)
integrated
moving
average
(ARIMA),
seasonal
ARIMA
(SARIMA),
multilayer
nonlinear
AR
artificial
neural
network
(NARANN)
with
several
training
algorithms
including
Bayesian
regularization
(BR),
Levenberg–Marquardt
(LM),
scaled
conjugate
gradient
(SCG),
Broyden–Fletcher–Goldfarb–Shanno
(BFGS)
quasi‐Newton,
Powell–Beale
restarts
(CGB),
Fletcher–Reeves
updates
(CGF),
resilient
propagation
(RPROP)
algorithms.
Initially,
upon
criteria
possible
combinations,
selected
superior
model
each
method
separately.
After
that,
best
category
involved
6‐
12‐multi‐step‐ahead
prediction.
stage,
error
are
calculated.
According
to
our
findings,
six‐step
window,
Holt–Winters
multiplicative
pattern
SARIMA
showed
less
bias,
though
compared
alternatives
like
NARANN‐lm/br,
difference
was
relatively
small.
next
process,
by
doubling
it
observed
that
(ANN)
(i.e.,
NARANN)
strictly
outperformed
models.
As
result,
shorter
steps,
can
provide
better
prediction,
while
longer
windows,
NARANN
implemented
vigorously
Finally,
used
10
models
predict
future
trend
till
end
July
2024.
