►
From YouTube: Industry 4.NOW - Intelligent Products
Description
In this session, we will look at how you can develop, maintain and launch connected, intelligent products.
You will also learn how SAP Enterprise Product Development suite of solutions can help you embed intelligence within your products and therefore create new sources of revenue for your business.
WE will also see some interesting case studies about how some of your peers have leveraged Intelligent, Connected Products to drive significant value to their business.
A
Okay,
so
I
think
we
should
not
let
people
who
join
on
time
wait
for
people
who
have
not
joined
on
time.
So,
let's
just
get
started
and
and
let's
see
who
joins
people
who
join,
let
them
join
in
as
as
we
go
along.
B
A
For
joining
today's
webcast,
we've
been
running
this
series,
a
five
part
series
on
industry,
four
dot
now
and
how
you
can
actually
drive
industry
4.0
conversation
within
your
organization.
A
A
C
A
Session
around
intelligent
asset
management,
we
did
that
with
maleki
actually
took
us
through
what
what
can
be
done
from
an
intelligent
asset
management
perspective.
Today
we
move
to
the
next
phase
of
the
conversation,
which
is
around
intelligent
products,
intelligent
and
connected
products.
Why
this
is
important
and
how
organizations
can
actually
drive
significant
amount
of
value
by
developing
and
creating
new
connected
and
intelligent
products,
which
you
can
then
deploy
as
new
business
models
as
well.
A
So
we
have
with
us
the
product
team,
ankit,
agarwal,
ayan,
pandey
and
saurabh
joining
us
on
the
call
they'll
take
us
through
about
45-50
minutes
into
the
session
around.
You
know
what
take
us
through
some
examples
and
also
talk
to
us
about
how
sap
can
actually
enable
customers
to
build
intelligent
products.
A
C
C
We
follow
as
much
capability
that
you
know
industry,
four
dot
now
or
industry
4.0,
whatever
you
might
call
it
brings
to
the
table
and
reach
out
to
you,
know
sort
of
be
going
beyond
what
we
are
normally
used
to
be
doing,
and
these
are
technologies
available
being
used,
and
so
the
question
is:
why
should
not?
Why
should
we
do
it?
C
The
question
is
more:
when
do
we
do
it,
and
that
is
where
you
know
we
are
trying
to
drive
this
conversation
towards
over
to
the
team,
to
take
you
through
that
entire
process
or
entire
cable
set
of
capabilities
that
industry
4.0
brings
to
the
table.
D
So
hi
everyone
and
good
morning,
good
afternoon
and
good
evening,
if
you're
based
out
of
new
zealand
so
yeah,
so
my
name
is
sarup.
I'm
part
of
the
enterprise
product,
sap,
enterprise,
product
development,
team
and
today,
as
a
part
of
this
session,
we
will
talk
about
what
are
intelligent
products
and
how
sap
can
help
you
or
your
organization,
make
your
products
smarter
and
build
and
design
intelligent
products
ankit.
If
you
can
move
to
the
next
slide,
please.
D
D
We
will
have
some
information
on
the
pricing
and
the
last
session
is
on
the
or
the
last
part
is
about
the
q,
a
yeah.
The
next
slide,
please,
okay.
So
what
is
an
intelligent
product?
D
According
to
us?
Intelligent
products
basically
rely
on
four
key
pillars
or
the
four
key
aspects.
The
first
is
the
technology
which
we
say:
okay,
whether
it's
the
sensors,
it's
the
processes
or
the
software
that
actually
lets
you
even
get
that
information
right
and
then,
once
you
have
that
in
place,
it's
the
intelligence
that
is
the
access
to
that
data.
D
To
drive
meaningful
insights
based
on
the
way
the
product
is
either
used
or
product
functions
or
performs.
So
so
you
then
kind
of
experience
that
so
that
these
are
the
experiences
and
then,
once
you
have
that
information
right,
this
data
is
then
fed
back
either
to
the
r
d
for
the
products
improvements
or
from
the
operations
again
back
to
the
the
oems
right
to
ensure
the
product
reliability
and
ensure
that
the
you
decrease
the
downtime
so
more
on
the
operations
and
the
feedback
mechanism
next
slide.
Please,
if
I
want
to
come
to
that
yeah.
D
D
If
I
go
to
the
left
side
right,
you
have
right
when
you
conceptualize
that
product
right,
you
have
that
information
and
that
information
goes
flows
to
the
build
phase,
then
to
the
install
phase
and
to
the
operators,
and
it
not
only
stops
there,
but
that
information
is
again
fed
back,
and
this
is
what
we
want
to
convey
with
talking
about
the
digital
thread,
and
this
is
what
we
are
working
on
right
now.
As
a
part
of
our
product
portfolio
and
yeah,
so
this
will
really
help
you.
D
So
if
we
go
to
the
next
slide
and
get
please
to
say
it
in
a
simplified
terms
right,
what
we
are
talking
about
is
if
you
want
to
capture
the
value
at
stake,
and
you
want
to
design
an
end-to-end,
you
need
to
design
an
end-to-end
digital
thread
right,
so
the
product
teams
they
get
the
information
from
the
when
they
design
the
products
they
feed
the
digital
thread
to
a
digital
product
model,
which
we
call
as
the
the
digital
thread,
which
is
then
consumed
by
the
extended
enterprise
which
which
are
your
customers
or
their
consumers.
D
And
then
there
we
have
how
the
customer
actually
uses
that
product.
We
get
that
information
from
the
operations
and
this
information
is
then
fed
back.
So
there
is
a
feedback
mechanism
right
which
is
then
consumed
by
the
product
team
right
to
kind
of.
So
this
is
a
kind
of
a,
I
would
say,
a
loop
mechanism
that
goes
and
on,
and
this
is
how
you
build
and
design
those
digital
products
or
the
intelligent
products.
The
next
slide,
please.
D
In
order
to
design
the
the
the
intelligent
products
right,
what
are
the
top
most
priorities
for
the
the
organization?
We
talk
about
the
top
line,
the
bottom
line,
and
now
the
most
happening
word
is
the
green
line
right.
So
when
we
talk
about
the
top
line
right,
how
do
you
define
new
products
right?
How
do
you
create
new
business
models
right
when
we
look
at
the
bottom
line?
I'm
just
just
taking
some
of
those.
D
Some
of
these,
I
would
say
tiles
from
there,
but
there
are
a
lot
of
other
things
you
can
do,
whether
it's
lowering
the
product
cost
right,
lowering
the
lifecycle
cost
or
when
we
talk
about
decarbonizing
the
products
right.
How
do
you?
How
do
you
reduce
that?
How
do
I
increase
the
reliability
so
that
I
don't
over
over
service
or
under
service
my
assets
and
reduce
those
time
down
times
right?
So
this
is,
I
would
say,
it's
more
about
the
intelligent
products.
D
B
Thank
you,
sir.
Before
I
start
on
colleagues
few
words
about
myself:
my
name
is
ankit
agarwal
based
out
of
norway.
B
I
am
part
of
the
same
team
at
saurabh,
so
basically
we
are
part
of
a
plm
cloud
solution
and
within
our
solution
we
have
capabilities
to
support
intelligent
products
from
both
from
the
oem
side.
So
even
if
your
equipment
manufacturer
or
if
your
operator
so
and
how
our
customer
have
done
that
in
the
past,
I
will
walk
you
through.
B
Some
of
the
customer
use
exam
customer
use
cases,
so
we
have
both
oems,
who
are
leveraging
the
digital,
twin
and
digital
products,
technology
to
enable
new
business
model,
and
we
have
also
a
lot
of
operators
who
are
monitoring
their
their
assets
based
on
virtual
sensors
and
digital
twin
with
our
technology
so
going
into
the
customer
use
cases.
So
I
will
start
a
customer
use
case
from
from
from
the
utility
industry.
So
it's
from,
I
cannot
name
the
customer
as
they
are
not
a
reference
customer.
B
Yet
so
it's
for
so
this
project
was
done
for
a
utility
company
based
out
of
out
of
france,
but
the
specific
project
was
done
for
for
a
plant
that
they
own
latin
america,
and
it
was
so.
The
target
of
the
company
is
to
basically
increase
their
share
in
the
in
the
renewable
energy
sector.
So
traditionally
they've
been.
B
B
So
typically,
what
happens
in
the
utility
industry
specifically
for
pv
plant
is
that
at
the
end
of
every
quarter,
you
need
to
send
your
colleagues
or
your
maintenance
engineers
to
to
inspect
that
if
anything
is
wrong.
That's
the
first
aspect
and
second
aspect-
is
to
clean
the
panels.
So,
as
efficiency
of
pv
panels
is
highly
dependent
on,
you
know
how
clean
they
are,
and
traditionally
the
most
of
the
maintenance
activities
are
based
on
the
time.
Time
schedules
inspections
right
so,
but
without
technology
they
wanted
to
see
if
they
can
switch
to
health.
B
Based
maintenance
and
religion
to
do
that
was
first
of
all.
Most
of
these
pv
plants
are,
they
are
located
in,
you
know,
quite
remote
location,
so
it's
challenging
to
send
inspection
engineers
and
imagine
if
so,
for
instance,
for
this
pea
plant,
it
was
not
a
huge
park
huge
plant.
It
was
relatively
quite
small
for
for
a
pv
plant,
it
was
two
megawatts,
even
though
it
was
two
megawatts
still.
B
There
were
around
9000
panels
in
the
whole
pack,
a
whole
whole
plant
and
out,
and
in
and
in
on
top
of
that
they
had
also
around
two
to
three
thousand
fuses.
They
which
were
installed
in
in
that
specific
pv
plant.
B
So
they
were
looking
into
our
technology
that
using
the
information
of
the
existing
sensors
can
we,
you
know,
give
them
outputs
or,
let's
say,
produce
virtual
sensor
data
which
can
give
them
refined
information
on
what
has
gone
wrong,
for
instance,
which
fuse
has
been
broken
at
which
specific
location
and
also
when
they,
when
do
they
need
to
do
any
maintenance.
So
basically
answer
all
four.
W
is
what
we
call
it.
B
So
what
occurred
why
it
occurred
when
it
occurred
and
where
it
occurred
so-
and
this
is
how
we
so
this
is
very
high
level
flowchart
how
it
looked
like.
So
they
are
our
sap
s,
4hana
plant
maintenance
user,
but
this
technology
also
work
with
you
know
if
you
are
also
on
older
erp
system,
for
instance,
if
you
are
on
ecc
also,
that
would
also
work.
So
that's
not
a
problem.
B
So
in
this
pv
plant
they
had
iot
sensor
data
installed,
for
instance,
for
temperature
for
irradiance
for
voltage,
and
they
had
also
current
and
power
values
which
they
were
actually
producing
only
on
a
few
locations.
So
in
total
on
this
overall
iot
sensor
data
values
which
they
had
was
about
eight.
B
So
then
what
we
did
we
used
that
input.
This
iot
sensor
data
available
as
of
today
on
the
plant
in
what
we
call
a
simulation
based,
digital
twin,
and
you
could
imagine
a
digital
twin,
basically,
a
replica
of
your
asset,
so
this
could
be
a
very
refined.
You
know
fem
based
or
cat
models,
etc,
or
it
could
also
be
as
simple
as
a
python
script
really
depending
on
the
use
case.
B
So
for
this
specific
use
case,
we
we
focused
on
a
1d
model,
so
you
know
no
cad
models,
no,
no
refined
finite
element
models.
A
very
simple
component
based
model,
so
where
you
know
you
can
just
use
the
sketch
of
the
of
the
plant,
how
different
components
are
clubbed
together
and
then
you
produce
a
digital
pin
of
that.
B
So
that
was
used
as
the
input.
Then
out,
as
as
output,
we
produce
a
multiple
sensor
data
at
multiple
locations
that
also
included
current
current
and
power
at
different
locations.
So
now
we
were
able
to
produce
for,
in
total,
around
40
to
50
additional
virtual
sensor
data.
So
now
customer
heads
in
total
50
to
60
time
series
data
that
they
can
further
use
to
do
anomaly.
Detection,
because,
if
they're
directly
going
through
this
loop,
then
first
of
all,
you
will
never
know
what
should
be
your
actual
performance
and
what
you're
actually
producing.
B
So
with
this
technology.
You
can
also
determine
what
you
should
be
actually
producing.
What
should
be
your
best
performance
and
what
your
actual
performance
is
right
and
that's.
One
aspect,
and
second
aspect
is:
if
you
are
not,
if
because
machine,
if
you're
doing
directly
and
normally
detection
based
on
machine
learning,
then
you
need
lot
of
data.
And
typically,
when
you
talk
about
these
kind
of
assets,
you
you
always
struggle
with.
B
You
know
getting
a
good
debt
good,
getting
good
data,
because
simply
either
you
need
to
install
a
lot
of
sensors
or
or
simply
storing,
data
is
not
available
or
asset
is
too
complex
that
you
cannot
obtain
obtain
the
sensor
data.
So
on
those
cases
you
can
always
get
it
via
virtual
sensor
where
digital
twins.
So
then,
only
by
installing
eight
to
ten
physical
sensors.
B
And
then,
whenever
any
of
these
anomaly
is
detected,
then
you
would
get
an
alert
in
your
backend
plant
maintenance
system
with
very
refined
information.
For
instance,
if
a
fuse
has
been
broken
that
this
specific
fuse
at
this
specific
location
has
been
broken,
and
this
is
how
it's
impacting
the
performance
of
your
plant,
and
so
this
is
why
you
should
take
some
action.
So
then,
as
a
plant
operator,
you
can
define
your
limits
that
you
know.
If
your
performance
is
below
than
this
value,
then
you
will
send
your
maintenance
engineers.
B
So
then
you
actually
know
when
to
do
maintenance
and
where
to
do
maintenance.
Instead
of
doing
a
typical
time
based
maintenance,
you
can
switch
to
what
we
call
intelligent
maintenance
or
digital
twin
based
maintenance.
So
this
was
the
first
first
use
case,
so
technology
used
a
high
level,
if
I
say
it
were
used
even
over
iot
machine
learning
and
and
digital
trends,
then
again
a
second
use
case
from
the
from
the
renewable
industry
as
well.
So
here
it's
for
a
wind
turbine
monitoring.
B
It
was
done
for
for
a
utility
company
based
out
of
norway,
so
here
again
a
similar
use
case.
They
wanted
to
again
switch
or
into
look
into
if
they
can
do
in
sort
of
smart
and
intelligent
maintenance
over
traditional
time
based
maintenance.
B
So
first
aspect
was
looking
at
the
performance
of
wind
turbine
and
second
aspect
was
maximizing
the
asset
life
and
minimizing
the
operational
costs,
and
typically,
what
happens
is
that
whenever
you
know
they
have
these
maintenance
cycles
in
order
to
do
maintenance
in
their
wind
tower?
So
so,
then
they
will
shut
down
their
operations
and
they
will
manually
climb
down
this
winter
wind
tower
and
check
all
the
locations
in
the
tower
where
there
are
issues
with
concern,
for
instance
with
high
structural
fatigue
with
high
vibration,
etc.
B
And
this
is
of
course
a
very
time
consuming
process
is
inves.
It
also
takes
a
lot
of
money
because
you
know,
because
you
need
very
refined,
equipments,
etc
and
also
you
need
to
they
also
logistic
costs,
but
so
and
then
they
use
a
lot
of
time
as
well,
because
they
don't
know
where
to
actually
inspect
so
they
inspect.
Everything
so
use
case
was
that
they
wanted
to
see
if
they
can
see
when
to
do
maintenance
and
where
to
do
maintenance
right.
So
if
I
go
to
next
slide
so
now
what
so?
B
There
are
different
use
cases.
For
instance,
as
I
mentioned
for
the
previous
pv
plant
case,
we
focused
on
the
component,
based
where
you
don't
need
any
any
3d,
3d
models
or
cad
models,
but
in
this
use
case
customer
was
interested
in
specific
sections
or
you
know
very
defined
information
of
the
wind
turbine.
So
we
we
used
an
fvm
based
approach,
finite
element
based
approach
via
cad
model.
So
then,
once
we're
in
simulations,
we
were
able
to
pinpoint,
for
instance,
on
which
specific
section
you
need
to
do:
maintenance,
for
instance.
B
Now
we
know
from
simulation.
That
is
this.
This
specific
sections
marked
in
the
red
we
have
the
relatively
higher
value
of
damage
or
what
we
call
in
structural
engineering
terms
structural
fatigue.
So
this
is
where
you
need
to
do
some
some
maintenance,
but
but,
for
instance,
the
parts
or
sections
lower
than
this.
The
value
of
the
damage
is
relatively
low.
So
then
they
no,
they
don't
need
to
spend
time
inspecting
here.
B
If
anyone
is
from
utility
industry
background,
you
might
know
that
in
wind
energy
sector
there
is
a
problem
with
the
your
misalignment.
So
this
is
basically
when
your
wind
turbines
are
not
aligned
with
the
wind
direction,
so
bigger
the
difference
between
the
alignment
of
the
wind
wind
turbine
or
what
we
call
rotor
with
wind
direction,
the
the
lower
the
performance
so
and
in
order
to
avoid
that,
typically,
there
are
control
systems
on
place
in
wind
turbine
which
can
follow
on
the
wind
direction
as
much
as
possible.
B
B
Why
our
system,
or
less
rather
system,
gave
them
an
automatic
suggestions
that
these
are
the
improvements
you
need
in
the
system
and
after
implementing
those
improvements,
they
were
able
to
able
to
increase
their
performance
by
15
to
20
percent
and
that's
very,
very
big
value
for
a
winter
wine
operator
for
one
wind
turbine,
okay,
the
next
case
so
now.
First,
I
started
with
utility
use
cases.
So
now
I'm
talking
about
a
public
infrastructure
use
cases.
So
here
again
it's
a
bridge
that
we
are
monitoring
based
out
of
norway.
B
It's
a
really
old
bridge
and
it's
as
you
might
know,
norway.
The
terrain
of
norway
is
quite
mountainous
right,
so
the
structures
there
are
usually
usually
a
bit
complex
and
this
bridge
is
80
years
old.
So
when
so
now
the
traffic
load
in
this
on
this
bridge
is
more
than
twice
of
you
know
what
it
was
designed
for.
So
that's
first
aspect
and
second
aspect
is
that
it
is
really
old
and
and
the
location
where
it's
located.
B
This
is
the
only
way
to
connect
to
to
two
parts
of
norway,
northern
on
norway
and
southern
norway.
So,
for
instance,
if
something
goes
wrong
here,
then
there's
a
big
issue
of
logistics.
So
then
yeah
you
will
then
basically
the
whole
transportation
between
northern
norway
and
southern
norway
will
crash
so,
and
you
cannot
overnight,
replace
this
bridge.
B
So
then
they
wanted
to
see
if
you
know
if
they
can
monitor
the
performance
or
the
integrity
of
this
bridge
in
the
real
time-
and
this
was
and
then
we
helped
them
do
it,
and
this
was
done
again
via
digital
twin.
B
So
basically
we
so
we
we
suggested
customers
to
install
some
iot
sensors,
so
in
total
they
stored
five
iot
sensors
in
the
span
of
the
bridge,
and
then
we
use
that
iot
sensor
in
our
system
to
produce
virtual
sensor
information
on,
for
instance,
structural
fatigue
or
damage
similar
to
wind
turbine
case.
So
on
a
few
model,
for
instance,
where
exactly
vibration
is
happening.
What
is
the
structural
fatigue
in
the
real
time
when
so
many
trucks
are
passing,
etc?
B
Parameters
like
that
and
actually
and
very
interesting
case
happened
a
couple
of
months
ago,
that
one
part
in
one
of
the
peers
broke
and
so
actually
that
specific
piece
started
vibrating
and
as
soon
as
the
vibration
happened
system
immediately
detected
it
and
gave
notification
to
the
to
the
maintenance
and
operation
engineer
that
this
specific
peer
has
much
higher
value
damage,
and
then
they
went
to
site
right
away
within
with
the
next
next
hour
and
then
they
saw
it.
Actually
it
was
that
actually
something
was
really
wrong.
B
B
So
this
was
a
really
good,
really
good
use
case
and
also,
if
you
can,
if
you
are
more
interested
in
what
this
specific
use
case,
we
also
have
a
youtube
video
available
where
customer
talks
about
how
exactly
they
benefited
from
the
solution.
So
when
we
share
the
debt,
we
can
also
share
that
youtube
video
link,
so
please
feel
free
to
watch
that
video
and
customer
customer
really
talked
about
in
in
detail
about
this
specific
use
case.
B
Okay
and
then
now
we
are
now
we
go
to
case
so,
as
you
can
see,
as
I
talk
about
different
use
cases
that
our
use
cases
are
not,
you
know
not
limited
to
any
industry,
so
we
have
almost
use
cases
from
each
and
every
industry.
I
will
show
some
of
them
today,
but
I
can
say
that
we
have
used
cases
first
from
utility
from
public
sectors,
less
transportation
than
we
are
use
cases
from
marine
industries.
For
instance,
this
is
floating
fish
farm
and
then
we
have
cases
from
oil
and
gas.
B
We
have
used
cases
from
from
life
sciences,
really
yeah
each
and
almost
each
and
every
aspect
of
of
the
industry.
So
coming
back
to
this
specific
use
case,
so
this
is
from
a
fish
farming
use
case.
So,
basically-
and
this
is
also
based
out
of
norway-
we
here
we
are
monitoring
us
a
floating
fish
farm,
and
this
is
the
second
biggest
floating
fish
farm
of
the
world.
B
And
and
since
whenever
we
talk
about
floating
structures,
it's
always
much
more
complicated
to
do
maintenance
and
to
inspect
or
to
do
and
to
do
real-time
monitoring
of
those
of
those
assets
and
and
in
norway
they
have
a
really
high
high
regulatory
demands
on
the
fish
feed.
So
basically
the
fish,
the
waste
which
fish
produce
that.
How
do
you
manage
that?
It
should
not
be
released
into
the
water
right
away
because
then
it
it
spoils
the
quality
of
the
water.
So
that's
a
requirement
by
the
authorities,
so
they
wanted
to.
B
First
of
all,
of
course,
look
at
the
structural
aspects,
for
instance
the
integrity
of
the
asset
and
and
enduring
high
c
conditions,
for
instance,
how
it's
impacting
impacting
impacting
the
vessel
shape
shape.
What
are
the
failure
modes,
it
might
produce,
etc-
and
second
aspect
was,
of
course,
that
how
do
you?
B
How
do
you
manage
that
fish
feed
what
we
call
it
so
the
fish
waste?
So
how
is
how
much
fish
they
produce
and
how
it's
recycled,
etcetera
and
then
third
aspect
that
when,
if
during
you
know
critical
conditions,
if
there
is
some
issue,
then
they
cannot
again
install
sensors
everywhere.
First
of
all,
it's
floating,
so
it's
installing
sensor
doing
anything.
Basically,
it
leads
to
really
big
costs,
because
you
need
to
drive
all
the
way
into
the
ocean
and
then
doing
is
doing
any
sort
of
downtime
or
any
sort
of
you
know.
B
Maintenance
activity
is
highly
costly.
So
so
then,
if
you
can
do
health
based
maintenance,
then
you
only
do
maintenance
when
it's
required,
instead
of
just
following
your
maintenance
cycle
and
also
when
you
do
maintenance,
you
can
actually
pinpoint
that
which
specific
joint
in
this
structure
or
which
specific
location
in
this
structure
requires
maintenance.
B
So
then
you
can
do
very
what
we
call
key,
or
very
sport,
focused
maintenance.
B
So
that
was
the
another
aspect
of
this
use
case
and
and
if
and
after
our
meeting,
if
you,
if
any
of
you
is
interested
in
you
know
any
specific
use
case
and
wish
to
have
a
deep
dive
session,
I
will
be
happy
to
happy
to
help
you
there.
So
then
we
can
have
one
to
one
later
on
so
today
I
will
not
talk
about
in
in
very
deep
dive.
I,
it
is
more
like
to
show
you
what
kind
of
use
cases
we
have
from
the
industry
and
how
it
can
be
used.
B
So
now
this
is
the
use
case
from
steel
industry.
So
in
the
steel
industry
they
typically
use
a
lot
of
cranes
of
different,
different
criticality
of
different
size.
B
So
and
typically
imagine,
if
you
are
a
train
operator
or
sorry
still
manufacturing
a
company,
then
you
are
using
trains
of
from
different
oems
and
typically,
as
you
would
see,
that
different
oems
also
come
with
their
own
sort
of
as
a
monitoring
solution.
But
if
your
operator,
you
have
you're
using
different
oems
for
different
cranes
and
then
then
you
might
have
100
different
systems
to
do
monitoring.
So
then,
as
as
operator,
it
could
be
really
challenging
to
keep
track
of.
You
know
all
different
assets.
B
So
then
they,
this
specific
customer,
was
looking
at
one
platform,
basically
where
they
could
monitor
all
of
their
assets,
irrespective
of
whichever
oems
they're
coming
from,
where
they
can
do
all
the
sensor.
Data
collection
or
the
digital,
twin,
related
data
processing
or
the
notification
handling,
and
then-
and
that
was
one
aspect
and
second
aspect-
was
that
in
the
cranes
main
part
of
maintenance
is
the
lubrication
of
the
joints
and
then
again
similar
to
other
industries.
That
was
done.
B
That
was
done
for
over
time,
so,
for
instance,
once
in
every
few
months,
but
sometimes
they
realized.
Actually
they
did
not
need
to
do
lubrication.
So
in
a
way
they
were
over
maintaining
their
assets.
So
then
they
wanted
to.
You
know
optimize
that
so
then
we
were.
If
I
go
in
with
more
detail
here
and
so
and
then
we
were
able
to
give
them
very
refined
information
on
when
to
do
lubrication
and
when
to
and
where
to
do
lubrication.
B
So
then
again
they
can
switch
to
health
waste
maintenance
over
time
based
maintenance.
So
again,
a
similar
flow
to,
for
instance,
the
pv
case.
So
we
have
a
crane
here
and
on
that
crane
you
have
a
few
iot
sensor
streams
so
into
like
if
they
had
six,
if
I
remember
correctly
so
current
position,
totally
position
hosting
position,
weight,
cable,
weight,
etc.
B
So
then
these
were
used
as
input
to
our
simulation
based
digital
twin
and
then
out
of
that
digital
twin.
We
produced
information
via
virtual
sensor,
data
on
hosting
parameters,
lubrication
time
forces
at
different
locations
of
or,
let's
say,
different
joints
of
of
of
that
of
that
specific
crane,
and
then
that
was
fed
as
input
to
train
machine
learning,
algorithms
and
then,
whenever,
whenever
they
required
maintenance,
they
would
get
an
alert
into
their
backend
system.
That
yeah
now
is
the
time
that
you
need
to
do.
B
Lubrication
on
on
your
point
on
your
on
your
on
your
crane
at
this
specific
point
and
then
all
of
this
project.
B
So
this
I'm
showing
this
flowchart
for
one
one
crane
and
then
they
can
use
the
same
technology
or
same
platform
to
extend
it
to
their
other
assets.
So
then
they
have.
They
get
a
one
scalable
platform
where
they
are
doing
the
operation
or,
let's
say,
monitoring
for
all
their
assets.
Instead
of
going
into
100
different
solutions
from
under
different
oems.
B
Then
this
is
example
of
so.
This
is
a
ongoing
pilot
project
and
this
is
for
for
autonomous
tractor,
so
they
have
installed
three
three:
three
sensors,
three
iot
sensors
and
based
on
that
they
are
doing
it's
quite
similar
to
the
use
case,
the
from
the
previous
imnc
industry.
So
the
previous
crane
use
case.
B
Basically
so
here
they
are
also
looking
into
the
maintenance
aspects
from
the
from
the
structural
perspective
so
which
are
the
locations
where
you
have
high
value
of
structural
fatigue
when
you
need
to
do
maintenance,
etc,
and-
and
so
this
is
quite
a
small
company
which
we
are,
which
we
are
taking
a
pilot
project
for,
but
it's
very
interesting
use
case,
since
it's
totally
autonomous
so
and
and
if
everything
is
successful,
then
the
customer
plans
to
deploy
it
to
their
all
of
to
all
of
the
tractor
tractor
asset.
B
B
So
this
was
really
good.
Example
of
you
know
how
intelligent
product
can
really
support
totally
automated
process.
Now,
there's!
No!
You
know
if
you're
talking
about
autonomous
vehicle,
we
are
talking
about
very
like
intelligent
maintenance
strategies
and
so
yeah.
This
was
very
interesting
use
case.
B
And
so
now,
most
of
the
use
cases
so
far
I
have
talked
about
you
know
from
a
very
operator
perspective
that
if
your
operator,
how
can
you
use
the
intelligent
products
technology
or
in
order
to
do
your
intelligent
maintenance?
So
now
there
is
a
initially,
as
I
mentioned
before,
I
started
that
we
also
have
oem,
who
are
even
leveraging
new
business
model
within
this
technology,
and
this
is
a
very
good
example
of
that.
So
now
it's
an
oem,
which
is
a
typically
so
their
traditional
business
is
manufacturing
of
progressive
cavity
pumps.
B
To
those
who
you
who
don't
know
so
cavity
problems
are
typically
used
in
oil
and
gas
industry
and
they
operate
at.
You
know
really
high
high
depths,
and
then
there
is
no
way
that
you
can
monitor
their
the
health
of
their
of
their
pump
at
these
larger
at
these
larger
locations.
B
So
then,
as
an
operator
who
or
as
customer
of
our
our
you
know
as
an
oem,
if
you
have
water
pump
from
them
and
now,
if
you
start
operating
for
your
for
your
operations
as
required
and
then
if
you
want
to
do
operation
or
if
you
are
sorry,
if
you
want
to
do
any
maintenance,
you
need
to
stop
your
operations
and
you
need
to
pull
the
pump
out
and
then
you
need
to
do
some
sort
of
maintenance
and
again,
typically
or
or
most
of
the
times.
B
You
will
realize
that
you
don't
need
to
do
maintenance,
because
you
are
simply
over
maintaining
maintaining
it
and
and
if
you
could
simply
say
that
you
could
install
iot
sensors
here,
but
that
is
simply
not
possible
because
now
we
are
talking
about.
You
know
depths
really
high
high
depth,
where,
where
physically,
it's
not
possible
to
install
iot
sensors
as
of
today
right.
B
So
the
only
way
to
do
the
maintenance,
traditional
maintenance
is,
you
know,
stopping
your
operation,
take
pulling
the
pump
out
and
then
inspecting,
if
requires
maintenance
and
then
doing
the
maintenance,
if
required,
but
now,
with
this
new
business
model.
Now
the
oem
is
enabling
customers
their
customers
to
actually
monitor
the
health
of
of
of
their
pump
via
virtual
sensors.
So
now
they
don't
need
to.
B
They
don't
need
to
stop
their
operation
if
maintenance
is
not
required,
so
they
can
install
virtual
sensors
via
digital
twin
and
then
so
so
now
as
an
oem,
you
have
two
business
models,
so
you
are
selling
a
physical
compressor
or
physical
pump
similar
to
before,
but
on
top
of
that
you're
also
selling,
let's
call
it
a
digital
product
or
intelligent
product
or
digital
twin.
B
So
so
we
so
basically
in
this
in
this
project,
what
oem
did
so?
This
was
also
again
a
pilot
project,
and
now
we
are
in
discussions
discussion
with
oem,
where
oem
has
some
customers
interested
who
are
able
to
who
are
interested
in
this
virtual
sensor
based
technology.
B
So
for
this
specific
pilot
project-
and
these
were
the
you
know-
sensor
data
available-
and
these
were
current
voltage,
drill,
head
pressure,
etc,
and
then
we
used
this
information
in
a
digital
twin,
and
this
was
a
3d
advanced
digital
twin,
something
similar
to
this
and
then,
as
an
operator
they're
able
to
produce
these
sort
of
parameters
at
all
at
those
locations
really
deep
locations
that
we're
talking
about
so,
for
instance,
fluid
level
pressure,
torque
temperature.
B
So
so
now,
as
an
operator,
you
don't
need
to
first
of
all
do
any
downtime
just
to
make
sure
your
asset
is
working
working
well
or
your
pump
is
rather
working.
Well,
you
will
only
do
or
you
will
only
stop
the
operations
when
you
will
get
a
notifications
from
the
system
that
this
is
what
you
need
to
do,
and
this
is
what
something
going
wrong
right
with
without
any,
without
any
physical
senses
at
those
locations.
So
it's
sort
of
a
win-win
for
both
oem
and
operator.
B
So
for
oem
you
are
enabling
new
business
model
by
also
selling
your
digital,
twin
and
as
operator,
you
are
using
the
digital
twin
in
order
to
do
condition
based
monitoring
over
time-based
monitoring,
minimizing
your
maintenance
cost
and
minimizing
your
downtime.
B
Okay
and
then
this
will
be
one
last
use
case
that
I
will
quickly
talk
about
is
in
the
in
the
ev
space,
so
is,
for
it
was
done
for
for
a
vehicle
manufacturer.
B
This
is
basically
battery
health
monitoring,
and
so
this
is
example
of
a
battery
from
from
a
vehicle,
and
then
we
have
we
have
some
physical
sensors,
which
were
which
were
passed
down
to
input
as
a
digital
twin,
something
something
like
this
in
apd.
This
is
specific
product
that
I'll
take
it
in
the
next
next
step.
Let's
say
for
now,
as
in
sap
solutions,
intelligent
solutions,
and
then
we
were
able
to
produce
virtual
sensor
data
on
the
remaining
lifetime
of
the
battery.
B
What
is
the
state
of
health
and
temperatures
at
different
locations
in
the
battery
health
wear
rate,
as
so
in
the
battery?
As
you
know,
it
is
impossible
to
to
to
calculate
the
aging
of
the
battery
via
physical
sensor
is
just
simply
not
possible
right
and
also
from
from
the
other
use
cases
as
well.
There
are
some
parameters,
even
though,
if
you
install
iot
sensor,
you
cannot
measure
them.
B
For
instance,
structural
fatigue,
as
I
showed
you
from
some
of
the
used
cases
right,
that
you
cannot
never
manage
and
never
measured
by
physical
sensors
or,
for
instance,
this
aging
of
the
of
your
equipments
that
also
you
can
never
measure
so
yeah.
So
on
this
kind
of
use
cases
where
these
kind
of
parameters
are
important,
for
instance,
aging
of
your
asset,
and
now,
when
I
talk
about
asset,
you
could
think
of
any.
You
know
complex
asset.
B
It
could
be
also
if,
for
instance,
compressors
it
could
be
transformers,
it
could
be
really
any
any
any
asset.
We
also
other
use
cases,
but
since
we
don't
have
enough
time
today,
so
we'll
focus
on
these
use
cases
and
then,
if
you're
interested
please
reach
out
to
me,
and
then
we
take
it
on
the
on
to.
B
We
can
show
them
to
you
later.
Okay,
coming
back
to
the
topic,
so
so
now
to
this
battery
use
case.
So
we
we,
for
instance,
have
this
physical
sensor,
so
that
is
measuring
state
of
charge.
What
is
charging
and
discharging
power,
and
what
is
the
temperature
at
one
specific
location
and
then
by
using
that
information,
we
are
able
to
get
the
virtual
sensor
information
for
different
the
other
parameters,
for
example
the
state
of
health,
health
wear
rate
and
what
is
the
energy
use,
etc.
B
So
if
we
take
a
full
cycle
so
then
there
is
a
sputter
where
you
know
battery
is
fully
charged
it's
and
then
now
battery
is
discharged
right
and
now
it
means
that
you
got
a
notification
that
you
need
to
charge
the
battery.
Then
you
start
charging
the
battery
and
then
battery
is
full
fully
charged
again
right.
So
this
is
one
full
cycle
of
the
battery,
and
now
the
charging
and
discharging
of
of
your
battery
is
measured
by
the
sensors
installed
in
the
battery.
That
is
nothing
new
right.
B
This
is
typically
we
have
in
every
battery,
and
so,
for
instance,
is
discharging.
This
is
charging
and
now
here
is
fully
charged.
Now
we
use
this
information
to
calculate
parameter
these
parameters
right,
so
how
your
state
of
health
of
battery
is
changing
with
every
cycle,
every
charging
cycle
of
your
battery,
so
basically
a
decrease
in
the
state
of
health
and
again
these
kind
of
parameters.
B
Then
this
is
the
decrease
in
the
state
of
health
when
fully
charged,
and
then
we
also
show
some
you
know
energy
used,
and
then
this
is
first
aspect,
of
course,
looking
at
the
aging
of
the
battery
another
aspect,
if
you
are
oem,
you
could
think
of
reducing
the
number
of
sense
physical
sensors
that
you
need
on
battery,
thereby
reducing
the
weight
as
an
oe
as
a
battery
manufacturer,
there
is
a
lot
of
focus
on
making
efficient
batteries
by
having
a
lightweight
battery
as
much
as
possible,
so
reducing
the
amount
of
physical
sensors
installed
on
that
battery
is
one
is
one
of
the
way
to
do
that,
so
you
can
just
install
minimum
number
of
physical
sensors
on
the
battery
and
then
use
that
information
to
calculate
other
parameters
via
virtual
sensor,
which
requires
no
hardware.
B
It's
totally
software
based
yeah
and
then,
as
as
said,
and
they
are
looking
then
looking
at
different
parameters,
for
instance
battery
aging
battery
temperature,
then
also
efficiency,
etc,
and
then
also
in
some
cases,
performance
as
well
for
some
some
of
the
batteries
yeah.
Okay.
So
then,
I
talked
about
all
the
use
cases
and
I
think
we
are
almost
by
the
time
so
I'll
quickly
walk
you
through
the
through
the
product.
B
So
I
talked
really
about
digital
twin
diet
and
what
exactly
is
a
digital
twin
in
in
our
sense,
so
you
could
think
this
of
a
physical
asset
and
on
your
screen
is
a
digital
replica
of
your
physical
assets
and
now,
in
this
asset
you
have
of
if
I
post
this
video
for
now,
we
have
this
one
iot
sensor
that
is
measuring
the
movement
in
this
specific
cantilever
beam,
and
this
measurements
is
given
as
a
feedback
feedback
to
this
specific
digital,
twin
or
digital
replica,
and
then
it
runs
multiphysics
simulations
in
the
real
time
and
gives
you
outputs,
for
instance,
for
parameters
like
structural
fatigue,
structural
stress
strain,
depending
on
the
use
case.
B
It
could
be
anything
and
now
you
can
generate
these
values
as
what
we
call
it
as
virtual
sensor
data.
So
to
summarize,
you
have
physical
asset,
you
have
a
digital,
twin
or
digital
replica.
Then
you
have
one
physical
sensor
and
then
digital
twin
uses
that
physical
sensor
to
calculate
virtual
sensor,
information
at
different
locations
for
different
parameters
of
interest.
B
And
then
this
capability
right,
which
we
talked
about,
is
intelligent
products
and
for
different
oem
models
for
different
operator
models
from
different
industries,
all
the
different
use
cases
they
have
been
covered
with
our
sap
product,
what
we
call
it
as
as
an
enterprise
product
development,
so
the
the
goal
of
this
product.
Our
this
specific
solution
is
to
cover
all
three
phases
of
your
product,
starting
from
define,
develop
and
deliver,
and
when
you
talk
about
define
phase
typically,
you
would
start
with
that.
B
So
now
for
now,
let's
think
I'm
talking
about
perspective
from
an
oem
for
now
and
then
I
will
switch
to
operator
perspective,
but
let's
for
now
focus
on
oem.
So
as
an
oem,
when
you
are
designing
designing
a
product,
you
start
with.
First
of
all,
what
are
the
requirements
which
are
coming
right?
So
requirements
could
come
from
your
market
insights
from
your
customers.
It
could
come
from
your
research
units,
so
there
could
be
different
requirements
right.
B
So
then
you
then
you
would
prioritize
the
needs
and
requirements
in
a
specific
what
we
call
it
as
a
requirements,
engineering,
because
someone
might
might
just
say
that.
Okay,
I
need
this
right,
but
how
do
you
achieve
this,
then
you
need
to
break
down
that
specific
high
level
requirement
into
smaller
requirements,
and
this
is
what
we
call
requirements
engineering.
So
that
is
what
we
call
it
as
a
defined
phase
and
then,
once
you
have
your
requirements
clearly
defined,
then
you
would
go
into
develop
phase
and
then
during
the
developer,
you
could
call
it
manufacture
phase.
B
You
have
to
collaborate
with
different
parties.
It
could
be
collaboration
with
your
suppliers,
it
could
be
collaboration
within
you
within
your
party
or
it
could
be
or
it
could
be
compliance
compliance
requirements.
For
instance,
if
you're
talking
about
process,
industry,
chemical
or
pharma
industry
there
there
are
there.
There
are
complaints,
require
requirements,
formulation
requirements
so
all
of
these
kind
of
requirements
or
which
you
need
to
take
into
account
during
the
develop
phase.
B
This
is
what
what
you
can
do
in
the
develop
phase
and
then,
finally,
once
you
have
developed
your
product
to
manufacture
your
product,
then
you
would
deliver
it
and
then
then,
as
said
as
oem,
you
can
also
generate
a
new
new
business
stream.
So,
instead
of
just
delivering
traditionally
your
your
product,
you
can
also
deliver,
let's
say
a
digital
twin
based
service
as
well.
So
now
you
can
also
sell
for
a
a
digital
product.
B
So
if
we
go
into
bit
more
detail
of
each
and
every
of
these
three
d's,
which
I
talked
about
so
first
of
all,
I
talked
about
these
requirements
part
so
that
requirements
could
come
from
voice
of
customer
again.
We
have
our
solution
with,
so
it
could
work
with.
If
customer
has,
you
know
their
own
solution
to
get
feedback,
so
we
can
help
with
integrations
or
if
they
don't
have
it.
B
We
also
have
our
own
solution
that
can
help
in
getting
the
voice
of
a
customer
and
then
once
you
have
that
all
the
requirements,
then
you
can
do
it
in
sort
of
break
to
break
them
down
in
what
we
call
it
requirement
engineering
and
then,
once
you
have
those
requirements,
then
you
would
go
into
the
develop
phase
and
now
in
the
develop
phase.
For
instance,
let's
take
example
of
a
crane,
so
in
a
crane
there
are
different
types
of
systems.
So
there
are
structural
system
that
you
need
to
design.
B
B
What
we
call
it
as
a
system,
engineering
and
and
then
during
the
development
phase,
you
need
to
do
collaboration
between
different
groups
in
your
units
or
you
need
to
do
supplier
collaboration
or
you
need
to
do
collaboration
on
the
specification
that
what
are
the
specifications
you
need
to
have
in
the
product
or
if
you
talk
about
talk
about
a
process
industry,
then
there's
also
product
validation.
That
is
quite
relevant
for
the
process
industry
and
then
you
you
had,
you
might
have
to
do
collaboration
on
different
type
of
documents
and
different
data.
B
So
all
of
this
can
be
handled
in
the
development
phase
and
then
another
aspect
is
digital
twin.
So
this
goes
into
direction
of
of
what
we
talked
about
in
the
new
business
streams
right.
So
so
anyway.
Typically
when
you're
designing
and
developing
your
any
physical
product
right,
you
always
have.
You
have
always
done
r
d
of
that
product
right
you.
So
these
kind
of
simulation
models
that
I
that
I
talked
about
earlier
right,
this
digital
twin,
digital
replica
of
your
asset
would
already
exist
in
yours
in
your
team.
B
So
now
it's
only
about
monetizing
on
it.
So
now
you
can
so
you
already
have
that
digital
twin,
but
then
you
can
sell
that
digital
twin
as
a
service
in
it
in
the
deliver
phase
that
I'll
talk
about
later
and
then
we
also
have
other
two
business
process
coming
coming
up.
So
now
this
intelligent
handover
is
already
has
come
up
in
q2
and
then
we
will
another
one
coming
out:
intelligent
routing,
so
basically
intelligent.
B
Handovers
deals
with
a
conversion
from
engineering
bomb
to
manufacturing
bomb
yeah,
and
then,
once
you
have
delivered
your
product
right,
you
can.
You
can
establish
collaboration
with
your
operators
with
your
with
your
customers
that
so
it
this
business
model
could
look
different
depending
on
you
know
what
operator
and
oems
are
comfortable
with
another
one
example,
as
an
oem
could
be.
You
take
responsibility
for
all
the
service,
so,
for
instance,
you
you
ask
your
operator
to
share
ot
data
back
with
you
and
then
based
on
the
ot
data
that
you
are
producing.
B
You
can
monitor
the
asset
based
on
the
digital
twin.
That's
one
business
model
and
second
business
model
is
you
are
selling
the
digital
twin,
for
instance,
for
that
specific
cavity
pump
use
case
that
I
talked
about
then
operator
themselves
take
care
of
the
maintenance
activity
based
on
the
digital
twin,
so
this
if
they
don't
want
to
share
the
ot
data
back
with
the
oem
so
there
so
as
an
oem,
you
have
two
two
different
choices.
Right
you
can.
B
This
is
how
you
can
think
of
your
of
your
digital
products,
or
how
do
you
want
to
collaborate
with
the
with
your
customers
on
on
digital
products?
So
you
are
basically
monetizing
on
the
on
the
r
d
models
that
you
owe
that
you
already
have
that
you
are
never
using
in
maintenance
and
operation
right
and
then
then
also
we
have
support
for
visual
services
and
support
and
and
and
space
visual
work,
instruction,
etc.
B
Okay,
so
now,
since
we
are
over
time,
so
I
will
skip
this
part-
maybe
maybe
yeah.
So
then
I
will
talk
about
these
parts
I'll
talk
about.
Maybe
this
specific
use
case.
So
this
is
one
of
the
business
process
that
I
had
explained
earlier
right.
So
now,
as
an
oem
right,
you
could
also
leverage
product
as
a
service
use
case,
for
example,
let's
say
instead
of
selling
any
asset
you
could
so
we
also
even
have
customers
who
are
doing
it.
B
So
we
have
a
customer
who
is
actually
selling
not
the
compressed
air
compressors,
but
compressed
airs
as
a
service.
That's
one,
so
you
are
basically
delivering
and
you
are
basically
delivering
product
as
a
service.
First
use
case-
and
second
use
case
is
in
is
quite
also
is
becoming
common
in
the
imc
industry
is
from
the
cranes,
so,
as
cranes
are
very
expensive
right
to
buy
for
operator.
B
B
How
exactly
an
operator
is
using
the
using
your
asset,
because,
let's
say
you
know,
if
you're
not
using
the
asset,
the
right
way,
then
of
course
the
breakdown
or
the
wear
and
tear
of
your
asset
will
be
much
faster
compared
to
you
know
how
it
should
be
used.
So
then,
as
an
oem
in
order
to
charge
your
operator
in
the
right
way,
you
need
to
do
the
real-time
monitoring
of
your
asset,
and
this
can
be
leveraged
by
the
digital
twin.
B
So,
as
said,
you
can
create
new
business
model
and
then
your
customers
could
pay
based
on
the
subscription
or
you
could
be
based
on
the
pay-per-use
models.
You
are
monetizing
on
the
engineering
models
that
already
exist
in
your
back
end,
and
then
we
have
integration
with
what
we
call
its
subscription
billing.
So
then,
with
the
subscription
billing,
you
can
automate
the
billing
process
in
s4hana
and
then
asset
user
can
provide
you
also
feedback,
and
we
we
enable
the
full
this
collaboration
between
oem
and
operator
via
what
we
call
it
as
asset
intelligence
network.
B
So
again
we
can
talk
about
that
if
you
are
interested
in
one
to
one
and
also
the
other
business
process
which
I
skipped
for
today
and
these
ones.
We
can
also
talk
about
that.
If
again,
if
something
is
interesting
for
you
and
later
on,
and
now
that
I
talked
from
oem
perspective
now
quickly
talking
about
from
from
an
operator
perspective,
so
I
talked
about
different
maintenance
strategies
right
what
operators
were
using
earlier
and
now
they
are
switching
to
this,
what
we
call
smart
maintenance
or
prescriptive
maintenance.
B
So
typically,
there
are
three
type
of
maintenance
strategies
that
and
maintenance
any
operator
would
define.
So
one
is
very
traditional
that
is
reactive,
so
that
that
you
do
maintenance
when
something
has
broken
down,
but
that
of
course
comes
at
a
high
cost
right,
it
suddenly
is,
there
is
a
is
a
is
a
downtime
or,
and
also
sometimes,
if
you
don't
have
the
equipment
or
that
specific
part
in
your
inventory.
B
Then
suppliers
might
take
advantage
of
your
situation
right
because
then
they
can
charge
as
much
as
they
want,
and
in
order
to
do
that
either
you
would
have
a
very
large
inventory
and
then
you
would
have
large
cost
related
to
inventory
in
case
of
reactive
failures,
but
then
say,
and
then
second
aspect
is
proactive
or
time
based
maintenance
that
I
talked
about.
B
So
we
I
showed
some
use
cases
where
customer
was
using
only
simulations,
for
instance,
the
wind
turbine
case,
then
their
use
case
of
panel
and
the
crane
use
case
where
customer
was
using
simulations
and
predictive
technology.
When
I
say
predictive,
I
mean
machine
learning
technology,
so
really
so
you
can
define
your
asset
strategies
based
on,
of
course,
how
critical
your
asset
is
for
your
operation?
How
you
know?
How
will
it
how
down
how
some
down
time
will
impact
your
operation?
B
So
you
can,
of
course,
convert
all
this
value
in
value
directly
to
two
dollars
right
in
order
to
do
good
value
engineering
case
and
all
of
the
customers
which
I
mentioned
earlier,
have
you
know
saved
a
lot
of
money
on
on
their
maintenance
and
operation
activities,
without
smart
technology
or
with
intelligent
technology
and
doing
this
kind
of
you
know.
B
Smart
maintenance
also
allows
you
to
do
all
sorts
of
these
kind
of
analysis.
Right
root,
cause
analysis,
meaning
what
has
gone
wrong
risk
risk,
based
inspection,
failure,
mode
analysis,
reliability,
centered,
maintenance,
etc.
B
Yeah
so
now,
if
I
in
either
scenario
right,
how
let's
say
high
level
a
business
flow
would
look
like
so,
for
instance,
now
we
have
a
temperatures,
I'm
talking
about
a
very
high
level.
So
let's
say
we
have
a
transformer
right:
have
a
different
temperature
different
sensors
installed
on
it,
for
instance,
temperature
scatter
system,
yeah,
etc,
could
be
anything
so
now
we
would
read
the
data
in
in
our
epd
solution
and
then
we
would
deploy.
B
B
You
know
when
you
need
to
do
and
do
the
changing
of
oil
of
transformer,
for
example
right
or
it
could
be
aging
of
transformer,
or
it
could
be
any
engineering
kpi
basically,
depending
on
you
know
what
your
asset
is,
what
is
the
use
case
etc,
and
once
you
have
that
information
of
those
highly
refined
data
or
highly
refined
engineering
kpis,
then
you
could
use
that
information
in
machine
learning
based
solution.
B
Let's
say
oil
needs
to
be
changed,
I'm
just
saying,
for
example,
right
and
then
based
on
that
what
kind
of
maintenance
you
need
to
do
you
can
trigger
work
order
so
basically
telling
you
telling
your
maintenance
engineers
exactly
on
what
kind
of
equipments
they
need
to
bring
to
site,
because
now
you
know
what
has
gone
wrong
and
where
it
has
gone
wrong
right.
B
So
you
save
a
lot
of
time
because,
if
you
think
about
additional
perspective,
so
it
could
be
that
maintenance
engineers
went
to
the
went
to
the
site,
but
they
did
not
bring
the
right
equipment
because
they
did
not
know
what.
What
is
the
root
cause
of
the
issue
right
so
now
you
know
what
is
the
root
cause
of
the
issue,
so
your
maintenance
engineers
will
bring
the
right
equipment
the
first
time,
and
all
of
this
can
be
managed
by
what
we
call
is
sap
field
service
management.
So
it's
not.
B
One
solution
is
different
solution,
but
together
what
we
can
you
can
enable
you
know
the
full
end-to-end
intelligence
maintenance
process.
B
The
specific
this
digital
twin
capabilities
that
I
talked
about
is
defined,
deliver
phase
that
is
covered
within
epd
enterprise,
product
development,
okay
and
now
a
short
one
slide
on
simply
pricing
how
it
how
it
is
based.
So,
as
I,
as
I
mentioned,
our
pricing
is
epd
is
enterprise.
Product
development
is
not
one,
not
one
solution.
Rather
it's
a
it's
a
suite
of
different
solution.
B
So,
for
instance,
if
you
are
interested
in
the
defined
phase
of
the
of
the
solution
or
if
you
are
interested
in
the
collaboration
aspect
with
different
during
your
during
your
development
phase
or
you
are
interested
in
the
visualization
aspects
or
you
are
interested
in
the
digital
twin
aspect
that
I
should
talk
about
later,
that
is
connected
products.
B
So
all
of
these
other
aspects
are
based
on
how
many,
how
many
users
will
have
access
to
the
system
and
connected
products
is
based
on
what
we
call
device
or
you
could
simply
relate
it
to
the
asset.
So,
for
instance,
one
crane
would
be
one
asset,
so
how
many
assets
you
want
to
monitor
and
then,
of
course,
there
are
other
parameters,
for
instance.
How?
B
Because
the
cloud
products
is
based
on
the
amount
of
data
that
you
produce,
so
that
how
many
iot
sensors
are
installed?
How
many
yeah,
how
many
virtual
sensor
will
you
produce,
etc?
So
parameters
like
that?
But
if
you
want
to
try
out
different
capabilities
of
epd,
you
don't
need
to
buy
it
separately,
so
you
just
need
to
buy
on
a
single
material
code,
so
you
could
think
of
epd
as
a
as
tokens
right.
So
then
you
have
bought
a
specific
number
of
tokens
or,
and
then
tokens
are
basically
capacity
units.
B
B
Yes,
then,
yes,
then,
thank
you
for
having
me.
So
if
there
are
any
q
and
a
and
would
like
any
discussion
I'll
be
happy
to
do
that.
C
So
we
are
now
open
for
questions
and
answers.
What
I
would
request
is
if
there
are
any
questions,
please
put
those
in
into
your
chat
or
you
can
open
up
the
mic
and
start
speaking
about
it.
Ankit
I
have
received
where
you
know
I
have.
I'm
asked
whether
this
digital
twin.
A
C
B
Absolutely
so,
if
I
quickly
go
back
so
if
I
quickly
go
back
to
one
of
the
use
cases
right,
let
me
let
me
go
to
previewskit.
I
think
that's
much.
That's
easiest
to
explain.
B
Let's
take
this
one,
rather
that's
also
okay,
so
for
instance,
now
right
so
now
you
are,
you
are
based
on
the
digital
twin.
You
are
producing
all
these
refined
engineering
parameters
right.
So
now
you
have
simulation
based
digital
twin,
that
is
producing
the
parameters
that
are
going
as
input
to
your
machine
learning
engine
where
all
that
anomaly
detection
and
the
prediction
is
happening.
So
the
when
you
talk
about
predictive
maintenance
piece.
So
this
is
where
it's
it's
quite
central
component.
B
B
You
might
generate
load
of
false
positives
right
and
that's
that's
what
you
would
like
to
avoid
and
and
that's
one
aspect
and
another
aspect
is
even
if
let's
say
you
are
able
to
generate
an
direct
alert
that
okay,
something
has
gone
wrong,
but
then
you
will
never
know
what
has
gone
wrong
and
why
it
has
gone
wrong
right,
but
digital
twin,
being
a
central
component
here,
for
instance,
in
this
use
case
right,
it
allows
you
to
rectify
both
the
issues,
so
it
it
gives
you
all
the
refined
data
or
what
we
call
engineering
kpis
right,
these
very
different
information
that
I
use
in
machine
learning.
B
So
this
way
you
you
have
information,
also
on
why
and
what
has
gone
wrong
right
or
what
will
go
wrong.
Rather,
if
you're
talking
about
prediction,
that's
one
aspect.
The
second
aspect
is
you
avoid
false
false
positives,
because
then
all
the
additional
data
that
is
required
for
machine
learning
to
do
prediction
that
is
produced
here.
So
this
is
how
digital
twin
sits
as
a
central
component
in
in
overall
predictive
maintenance
picture.
C
Yeah
definitely,
okay.
So
a
follow-up
question
on
this
is
that
you
spoke
about
product
as
a
service.
How.
B
C
Digital
digital
twin
help
us
turn
a
project,
a
product
into
a
service.
B
Yeah,
okay,
perfect!
So
now,
let's
take
example.
Example
of
so
this
I
did
not
include
in
my
presentation
today,
so
this
is
for
a
compressed
company
that
we
are
working
with
based
out
of
germany,
so
they
they
build.
They
build
a
lot
of
compressors
right.
B
That
was
the
traditional
business,
so
they
would
sell
compressors,
but
now
they
were
looking
at
if
they
could,
instead
of
selling
compressors,
if
they
could
directly
sell
compressed
air
as
a
service
to
their
customer
and
then
customers
typically
are
in
oil
and
gas
industry
mining
industry
and
also
in
process
industry
right
who
use
compressed
air
who
needs
that
compressed
air.
So
now
they
are
delivering
compressed
air
as
a
service,
but
in
but
of
course,
in
order
to
do
that
right,
they
still
need
to
have
a
plant
at
this
compressed
air
plant
at
customer
site.
B
But
customer
is
not
owning
that
plant.
The
plant
is
still
owned
by
the
oem
by
the
by
the
pump
compressor
manufacturer,
but
they're
only,
but
the
operator
is
only
paying
for
the
amount
of
compressed
air.
That
is
that
they
are
using
right
and
in
order
to,
but
again
that
could
be
and
then
applied
operator
could
use
the
asset
in
really.
B
You
know
different
way
than
it
is
meant
to
be
so
in
order
to
do
a
good
monitoring
of
your
or
unless
in
order
to
charge
the
customer
the
right
way,
you
would
need
you
know
that
digital
twin,
so
that
you
can
monitor
that.
You
know
how
a
customer
is
using
that
compressed
layer.
How
is
that
affecting
your
asset,
health,
or
rather
compressor
health
right?
So
that's
that
is
one
example,
and
second
example
is:
is
that
crane
example
right
so
also?
B
A
lot
of
train
manufacturers
are
looking
into
solutions
where
you
know
they
are
not
selling
the
crane
itself,
but
they
are
selling.
They
are
renting
out
the
crane.
Basically,
so
then
they
are
leveraging
product
as
a
service
where
operator
would
pay
either
based
on
how
many
hours
they
have
used
the
crane
for
or
they
would
rent
out
it
rent
it
out
for
a
month.
Let's
say
right,
but
if,
but
irrespective
of
whichever
scenario
an
operator
or
an
operator,
is
you
know
renting
out
the
crane
for
as
an
oem?
Who
is
who?
B
Who
is
leveraging
that
that
that
way
to
do
it
right,
you'd
still
need
to
do
a
very
good
monitoring
right.
So,
for
instance,
if
crane
is
meant
to
build
100
kilos
right
and
if
operator
is
lifting
150
kilos
every
time,
so
then
the
wear
and
tear
of
your
crane
will
be
much
faster.
And
how
do
you
monitor
that
right
and
the
the
only
way
you
can
monitor
it
by
having
this?
B
You
know
real
time,
virtual,
twin
or
digital,
twin
technology,
where
you
know,
depending
on
your
depending
on
your
asset,
that
you
know
if
it's
operated
the
right
way.
This
is
how
my
wear
and
tear
would
be,
but
since
operator
is
not
operating
the
right
way,
this
is
how
vlan
tier
is
right
and
then
they
can
charge
the
operator
accordingly.
So
this
is
how
you
can
leverage
productivity
service.
Why
are
digital
twins.
D
Okay,
even
in
terms
of
like
one
aspect
is
like
okay,
you
can
charge
the
operator.
The
second
aspect
of
what
ankit
already
explained
is
like
okay.
When
we
have
the
digital
twins,
we
have
insights
and
information
which
we
were
previously
not
having
right
so
now
you
know,
when
is
the
asset
likely
to
break
right?
That
is
one
of
the
things
right,
so
you
can.
D
You
know
when
to
service
that
asset
right,
and
that
is
that
is
the
insights
and
you
that
will
help
you
to
ensure
the
uptime
of
that
asset,
because
that's
where
one
of
the
models
is
like:
okay,
you're
charging
product
as
a
service,
but
you
need
to
ensure
that
the
product
is
up
and
running
so
so
this
is
where
digital
term
twin
actually
helps.
In
addition
to
what
ankit
already
mentioned,.
C
So
so
another
question
that
has
come
into
the
chat
on
one
to
one
is
how
so
we
have.
We
have
been
you
know
in
earlier
this
thing
we
have
been
talking
about
intelligent
assets,
which
means
the
production
assets,
and
now
we
are
talking
about
intelligent
products.
Now,
what
he's
asking
is?
How
are
these
two
connected?
Is
there
a
connection
between
them,
or
these
are
two
separate
tracks
that
run
parallely.
B
No
not
at
all,
rather
so
I
would
say
when
we
talk
about
intelligence
in
technology
right,
so
everything
is
connected,
and
this
is
how,
as
for
mentioned
earlier
right,
we
if
I
may
go
back
to
go
back
to
point
right
and
that
slide
how
we
enable
digital
thread
right.
So
basically,
let
me
quickly
go
back,
go
here,
yeah.
So
when
we
talk
about
this
overall
intelligence
right,
everything
is
connected.
B
So,
basically-
and
now,
if
you
talk
about
intelligent
products,
so
as
an
oem,
you
are
building
intelligent
products
right,
so
basically
you're
designing
your
you're
collaborating
with
different
parties.
You
are
installing
and
then
you
are
selling
it
to
your
operator
and
now
operator
is
operating
that
asset
in
a
smarter
way
and
then
smart.
As
assured,
there
are
different
ways
to
do
it
right.
B
So
one
of
them
to
do
it
is
via
digital
twins
based
on
the
insights
and
that
digital
twin
is
coming
from
oem
right
as
a
digital
product,
for
instance,
and
then
there's
one
example
and
of
course
digital
twin
could
also
come
from
other.
You
could
develop
it
yourself
if
you
don't
want
to
buy
from
oem
really
that
doesn't
matter,
but
at
the
end
of
the
day
and
then
once
you
are,
then
you
are
operating
your
asset.
B
You
are
maintaining
it,
then,
for
instance,
if
you,
if
you
need
servicing
right,
you
might
also
have
some
sort
of
agreement
with
oem
that
they
take
care
of
the
servicing
or
you
you
could
generate
and
sell
service
yourself
or
how
your
asset
is
performing.
You
would
always
give
you
know
feedback
back
to
the
back
to
your
oem
so
that
they
can
improve
the
design
for
the
next
time
right
or
they
can
improve
your
product
basically.
So
this
is
how
intelligent
product
and
intelligent
asset
management
are
are
connected.
B
So
as
an
oem
you're,
building
up
intelligent
product
and
as
an
operator,
you
are
operating
those
products
in
an
intelligent
way.
So
what
we
call
is
an
intelligent
asset
management,
so
everything
is
linked.
C
Okay,
thanks
thanks
for
that
answer,
and
the
other
answers
again,
we'll
wait
for
anybody
else
who
needs
to
ask
a
question.
C
B
B
Yeah,
since
we
did
not
have
time
today
so
I
ran
out,
I
ran
over
time
during
the
use
cases,
so
I
did
not
have
time
to
talk
about
it,
but
yeah.
So,
basically,
how
we
integrate
is
with
siemens
is
that
is
it's
wired
where
team
center,
so
we
have,
for
instance,
this
is,
if
you
have,
if
you're
using
siemens
team
center
right
plm
team
center
and
then,
if
you
want
to
bring
all
the
data
from
team
center
to
sap's
erp
system.
So
then
we
have
built
up
integration
with
s4.
B
B
A
Hi,
so
I'm
sorry,
I
had
to
step
out
for
a
couple
of
minutes
for
something.
So
thank
you
so
much
everyone
for
for
your
time,
insights.
I
have
one
question.
I
do
not
know
whether
this
was
asked
and
answered
before
I
kind
of
whenever
stepped
out.
A
The
question
is
the
following,
which
is:
there
is
a
whole
suite
of
possibility
when
you
are
able
to
provide
your
product
not
just
as
a
product
but
as
a
service.
There
is
this
entire
concept
of
mobility
as
a
service,
there
is
a
concept
of
hot
air
as
a
service
rather
than
boilers,
that
is
the
service
of
mileage
as
a
service
rather
than
selling
a
vehicle.
A
So
how?
How
does
this
entire
concept
can
be
realized
using
what
what
you're
trying
to
do
by
the
technology
that
you
talked
about.
B
A
Okay,
and
so
one
of
the
questions
that
also
comes
in
for
me,
is
you
know,
is
this
only
relevant
for
a
heavy
machinery
or
heavy
big
assets
industry,
or
can
it
also
be
deployed
in
terms
of
you
know,
consumer
grade,
cpg
kind
of
situations
as
well.
B
Yes,
absolutely
absolutely
so
so
today,
I,
since
the
theme
was
this-
you
know
in
this
intelligent
how
it
goes
intelligent.
You
know
maintenance,
but
we
also
have
a
lot
of
customers
in
cpg
and
and
also
in
chemical
and
process
industry
as
well,
where
they
are,
you
know
using
specific
epd
technology
in
their
r
d
and
define
phase
in
order
to
improve
the
r
d
process.
B
So
specifically,
if
I
let
me
just
quickly
go
into
so,
then
they
are
specifically
focusing
on
these
aspects,
so
that
is
supplier
collaboration,
specification
collaboration,
product
validation,
document
collaboration.
So
all
of
these
aspect
is
something
you
know
where
cpg
industry
is
is
using
our
solution
and
also
so
that
is
like
the
r
d
aspect
and
and
anyway,
in
order
to
produce
all
of
those
products
right,
you
have
some
sort
of
assets,
some
machines
that
are
producing
those
assets
right.
B
It
could
be
anything
so
and
then
we
talk
about
the
maintenance
aspect,
so
then
for
those
industry.
If
we
talk
about
from
maintenance
aspect,
then
we
have
to
think
of
maintenance
of
those
those
assets,
so
that
I
mean
the
assets.
Could
those
maintenance
assets
could
be
as
simple
as
a
motor
right?
It
doesn't
need
to
be
very
heavy
it.
We
have
customers
doing
it
for
as
simple
as
a
motor,
but
also
at
the
same
time
as
big,
as
you
know,
wind
turbine,
for
example,
but
and
in
the
r
d
r
d
phase.
B
They
are
typically
using
it
for
these
specific
pieces
of
the
business
process,
that
is,
collaboration,
r
d
requirements,
engineering,
product
validation,
etc,
and
when
it
comes
to
maintenance,
it
really
does
not
matter.
If
the
asset
is
big
or
small,
we
can
cover
up
too
small,
also
and
also
heavy
assets.
D
And
to
your
question
on
the
like:
okay,
just
another
part
or
another
aspect
to
this
is
yes,
we
do
look
at
the
the
critical
assets.
That's
the
I
would
say
that's
the
positioning,
but
it
also
depends
on
what
is
the
the
the
roi
right
in
terms
of
like
okay,
you
cannot
whether
the
customer
wants
to
monitor
their
a
small
asset
right.
It
depends
how
critical
that
is
like
and
if
the
the
asset
is,
if
the
asset
fails
right,
what
is
the
opportunity
cost
for
that?
D
When
we
talk
about
digital
twins,
we
are
only
looking
at
very
critical
assets
like
okay,
take
an
example
of
oil
and
gas
if
the
oil
and
gas
plant
shut
down
for
a
day,
and
there
are
millions
of
millions
and
millions
of
revenue
loss
for
them
on
a
daily
basis.
D
So
that's
the
customer
who
has
to
decide
okay,
whether
this
asset
is
critical
for
them.
For
the,
I
would
say,
the
process,
industries,
digital
twin
story
only
fits
in
the
maintenance
aspects
like
ankit
mentioned,
for
the
process
industry
for
the
discrete
industry.
It
could
go
into
their
assets
as
well,
depending
what
they
want
to
look
at.
A
If
you're
speaking-
oh
I'm
sorry
even
after
two
and
a
half
years
of
doing
this,
your.
B
C
A
That
I
said
if,
if
there
are
no
further
questions,
I
think
I
am
done
with
my
questions.
A
I
would
like
to
thank
the
presenters
and
team
for
the
entire
insights,
and
I
know
it
was
you
know
well
received.
Would
you
like
to
close
out
the
session.
D
Maybe
one
point
before
we
close
from
our
side:
thanks
everyone
for
joining
this,
just
in
the
interest
of
the
time
we
just
pulled
out,
I
would
say
a
handful
of
the
use
cases
but
have,
but
we
have
used
cases
across
all
industries,
whether
it's
process
or
discrete.
So
if
you
are
interested
to
know
more
just
feel
free
to
just
ping,
us
email,
us
and
we'll
get,
we
can
have
those
one-on-one
conversations.