Future Internet,
Journal Year:
2024,
Volume and Issue:
16(11), P. 402 - 402
Published: Oct. 31, 2024
Replacing
mechanical
utility
meters
with
digital
ones
is
crucial
due
to
the
numerous
benefits
they
offer,
including
increased
time
resolution
in
measuring
consumption,
remote
monitoring
capabilities
for
operational
efficiency,
real-time
data
informed
decision-making,
support
time-of-use
billing,
and
integration
smart
grids,
leading
enhanced
customer
service,
reduced
energy
waste,
progress
towards
environmental
sustainability
goals.
However,
cost
associated
replacing
their
counterparts
a
key
factor
contributing
relatively
slow
roll-out
of
such
devices.
In
this
paper,
we
present
low-cost
power-efficient
solution
retrofitting
existing
metering
infrastructure,
based
on
state-of-the-art
communication
artificial
intelligence
technologies.
The
edge
device
developed
contains
camera
capturing
images
dial
meter,
32-bit
microcontroller
capable
running
digit
recognition
algorithm,
an
NB-IoT
module
(E)GPRS
fallback,
which
enables
nearly
ubiquitous
connectivity
even
difficult
radio
conditions.
Our
methodology,
on-device
training
inference,
augmented
federated
learning,
achieves
high
level
accuracy
(97.01%)
while
minimizing
consumption
overhead
(87
μWh
per
day
average).
Sustainable Energy Grids and Networks,
Journal Year:
2024,
Volume and Issue:
39, P. 101452 - 101452
Published: June 18, 2024
Existing
energy
management
systems
are
becoming
increasingly
insecure
and
inefficient
due
to
the
rapid
adoption
of
smart
grid
technology.
Current
research
indicates
that
effectively
managing
dynamic
flows,
adjusting
changing
needs,
protecting
against
new
cyber
threats
remain
significant
challenges
for
system.
An
advanced
comprehensive
plan
grids
is
therefore
required,
capable
addressing
these
delicate
multifaceted
problems.
The
proposed
framework
addresses
through
unifying
several
key
aspects,
it
includes
an
data
acquisition
system
captures
real-time
from
various
sources,
enabling
monitoring
flow
analysis.
By
integrating
predictive
algorithms,
provides
precise
demand
forecasting,
which
essential
adaptive
management.
A
contribution
incorporation
AI-based
module
diagnostics
prognostics,
leverages
machine
learning
techniques
shift
reactive
proactive
maintenance
strategies.
optimal
power
(OPF)
optimization
represents
a
central
component
framework.
It
employs
computational
methods
ensure
efficient
cost-effective
distribution,
particularly
in
incorporating
renewable
sources.
Additionally,
architectural
strengthened
by
robust
cybersecurity
designed
safeguard
wide
range
threats,
maintaining
integrity
both
operational
consumer
data.
This
paper
also
practical
implementation
such
as
compatibility
with
existing
infrastructure,
investment
costs,
need
specialized
training.
solution
benchmark
operations,
ensuring
more
sustainable
systems.
International Journal of Production Research,
Journal Year:
2025,
Volume and Issue:
unknown, P. 1 - 17
Published: Jan. 9, 2025
Human
behaviour
recognition
is
one
of
the
most
fundamental
tasks
in
Industrial
Internet
Behaviour
(IIoB)
and
crucial
for
safe
reliable
IIoB.
Existing
methods
lacks
adaptability
transferability.
In
addition,
there
a
data
isolation
problem
among
different
users.
Therefore,
an
urgent
requirement
to
construct
secure
adaptive
human
model
IIoB
without
violating
privacy
Mamba,
structured
state
space
that
integrates
selection
mechanism
scan
module,
used
time
series
modelling
tasks.
To
tackle
aforementioned
problems,
Federated
Learning-based
lightweight
with
selective
models
proposed.
First,
we
design
integrating
Mamba
residual
structure
achieve
modelling.
considering
training
efficiency,
decentralised
dynamic
FL
framework
designed
collaborative
training,
including:
initial
source
users,
aggregation
strategy
based
on
weighting,
fine-tuning
module
small-sample
data,
improve
efficiency
accuracy
recognition.
Extensive
experiments
are
conducted
prove
superior
performance
proposed
method.
Computers,
Journal Year:
2025,
Volume and Issue:
14(4), P. 124 - 124
Published: March 27, 2025
Federated
Learning
(FL)
is
a
transformative
decentralized
approach
in
machine
learning
and
deep
learning,
offering
enhanced
privacy,
scalability,
data
security.
This
review
paper
explores
the
foundational
concepts,
architectural
variations
of
FL,
prominent
aggregation
algorithms
like
FedAvg,
FedProx,
FedMA,
diverse
innovative
applications
thermal
comfort
optimization,
energy
prediction,
healthcare,
anomaly
detection
within
smart
buildings.
By
enabling
collaborative
model
training
without
centralizing
sensitive
data,
FL
ensures
privacy
robust
performance
across
heterogeneous
environments.
We
further
discuss
integration
with
advanced
technologies,
including
digital
twins
5G/6G
networks,
demonstrate
its
potential
to
revolutionize
real-time
monitoring,
optimize
resources.
Despite
these
advances,
still
faces
challenges,
such
as
communication
overhead,
security
issues,
non-IID
handling.
Future
research
directions
highlight
development
adaptive
methods,
measures,
hybrid
architectures
fully
leverage
FL’s
driving
innovative,
secure,
efficient
intelligence
for
next
generation
Frontiers in Sustainable Cities,
Journal Year:
2024,
Volume and Issue:
6
Published: Dec. 19, 2024
The
purpose
of
this
study
is
to
assess
the
potential
machine
learning
in
advancing
Sustainable
Development
Goals,
particularly
Goal
11,
which
focuses
on
sustainable
urban
and
community
development.
To
reduce
impacts
increasing
urbanization
environment,
it
necessary
prioritize
development
smart
cities.
Smart
cities
use
information
communication
technology
techniques
enhance
sustainability
by
improving
resource
management
reducing
environmental
impact.
In
context,
artificial
intelligence
enhances
overall
quality
life,
a
critical
component
Machine
learning,
subset
intelligence,
crucial
promoting
This
application
cities,
ranging
from
energy
management,
transportation
efficiency,
waste
public
safety.
It
highlights
role
algorithms
improve
operational
minimize
expenses,
practical
ML
across
several
countries
demonstrates
its
ability
handle
challenges
increase
sustainability.
paper
discusses
variety
real-world
initiatives
that
have
successfully
employed
develop
as
well
in-depth
studies
used
obtained
results.
also
covers
implementing
into
city
projects,
such
data
quality,
model
interpretability,
scalability,
ethical
considerations.
emphasizes
importance
high-quality
data,
clear
models,
right
tools.
Energy Exploration & Exploitation,
Journal Year:
2024,
Volume and Issue:
42(6), P. 2241 - 2269
Published: Aug. 27, 2024
In
smart
cities,
sustainable
development
depends
on
energy
load
prediction
since
it
directs
utilities
in
effectively
planning,
distributing
and
generating
energy.
This
work
presents
a
novel
hybrid
deep
learning
model
including
components
of
the
Improved-convolutional
neural
network
(CNN),
bidirectional
long
short-term
memory
(Bi-LSTM),
Graph
(GNN),
Transformer
Fusion
Layer
architectures
for
precise
forecasting.
Better
feature
extraction
results
from
Improved-CNN's
dilated
convolution
residual
block
accommodation
wide
receptive
fields
reduced
vanishing
gradient
problem.
By
capturing
temporal
links
both
directions,
Bi-LSTM
networks
help
to
better
grasp
complicated
use
patterns.
improve
predictive
capacities
across
linked
systems
by
characterizing
spatial
relationships
between
energy-consuming
units
cities.
Emphasizing
critical
trends
guarantee
reliable
forecasts,
transformer
models
attention
methods
manage
long-term
dependencies
consumption
data.
Combining
CNN,
Bi-LSTM,
GNN
component
predictions
synthesizes
numerous
data
representations
increase
accuracy.
With
Root
Mean
Square
Error
5.7532
Wh,
Absolute
Percentage
3.5001%,
6.7532
Wh
R
2
0.9701,
fared
than
other
‘Electric
Power
Consumption’
Kaggle
dataset.
develops
realistic
that
helps
informed
decision-making
enhances
efficiency
techniques,
promoting
forecasting