►
From YouTube: OpenShift at Thyssenkrupp Elevators with Joshua Bryant at OpenShift Commons Gathering 2019
Description
Case Study: OpenShift @ Thyssenkrupp Elevators
Billy Holmes (Red Hat)
Joshua Bryant (thyssenkrupp Elevator Americas)
at OpenShift Commons Gathering 2019
at Red Hat Summit
B
So
thyssen,
krupp
elevators:
this
is
a
quick
overview,
50,000
employees,
nine
hundred
branches
in
70
countries,
1.1
million
units-
and,
to
be
honest,
some
of
this
information
is
a
little
dated.
So
if
that
is
inaccurate
Joshua,
let
me
know
in
the
u.s.
eighty
six
hundred
employees
a
hundred
fifteen
branches,
two
hundred
twenty
thousand
units
under
Maine
as
an
units
or
not
just
elevators,
but
elevators.
B
Filled
service
organization,
twenty
five
hundred
again
install
base
about
22
K:
they
have
contractual
maintenance,
so
when
they
sign
up
with
a
building
or
a
contractor
or
another
company
for
their
elevators,
they
have
certain
contracts,
obligations
to
maintain
those
elevators
at
certain
cadence.
They
have
to
do
them
monthly
or
they
had
to
come
out.
You
know
every
six
months
or
whatever
that
is,
and
each
contract
is
unique.
B
It
is
specific
to
the
negotiations
and
then
they
also
have
brake
fix
right,
which
is
what
happens
when
you
don't
plan
for
something
when
the
elevator
just
doesn't
work
like
it
breaks
down
for
whatever
reason,
and
they
have
the
you
know.
How
do
you
get
your
maintenance
crew
to
the
correct
places
in
the
best
amount
of
time
in
order
to
reach
your
SLA
s
and
in
order
to,
in
addition,
you're
injecting
unplanned
outages
into
these
maintenance
plans?
So
it's
all
well
and
good
to
plan
what
you're
gonna
do
for
the
week
until
something
happens?
I.
B
B
Their
mechanics,
total
productive
time,
reduce
overtime,
reduce
the
miss
maintenance
and
increase
the
renewals
of
their
contracts.
Cuz.
If
you
can't
perform
what
you've
been
contracted
to
do,
then
no
one's
going
to
renew
your
contract
right
or
it
makes
renewal
difficult.
When
you
go
back
and
say,
hey
I
want
you
to
trusted
me
again
for
another
year
or
two,
and
you
haven't
done
what
you've
contracted
obligated
to
do.
B
It
becomes
a
very
difficult
negotiation
period,
so
they
wanted
to
clean
up
their
install
base,
ensure
they
meet
their
SLA,
ensure
resource
availability,
which
is
awful
resource
ability
in
this
instance
meant
actual
mechanics
and
actual
individuals
that
were
performing
the
job
functions
being
able
to
group
locations.
So
you
didn't
have
people
driving
unnatural
distances
to
meet
the
contractual
obligations,
so
that
would
require
overtime
because
they're
dry
driving
longer
distances
in
order
to
reach
these
locations
and
there's
other
issues
around.
B
B
You
know
a
twelve
by
twelve
by
twelve
elevator
and
get
it
put
in
your
building.
Every
elevator
is
uniquely
designed
and
custom-built
for
each
building
and
thyssen
krupp,
elevators
adopts
other
elevators
that
were
built
by
other
companies.
So
even
if
they
did
not
build
that
elevator
in
a
building,
they
can
do
the
maintenance
on
that
elevator,
if
you
so
require
them
to,
and
that
is
something
I
did
not.
B
Those
two
things
are
I
did
not
know,
which
kind
of
also
increases
the
complexity
of
this
problem
a
little
bit,
because
now
I
have
to
train
my
technicians
to
not
only
to
repair
my
stuff
right,
the
stuff
that
but
I've
built
and
I've
created,
but
as
well
as
you
know,
potentially
my
competitors,
people
that
have
or
legacy
maybe
the
company's
not
around
anymore.
So
you
have
elevators
out
there
that
they
did
not
build
that
they
still
have
to
support
and
as
we'll
learn
a
little
later.
These
are
not
just
mechanical
situations.
B
So
we
have
a
managed
OpenShift
product
on
overshift
dedicated,
which
removed
a
lot
of
the
operational
burden
on
thyssen
krupp,
to
manage
the
environment
and
allowed
them
to
build
a
solution
set
that
resolved
some
of
the
issues
that
they
were
facing
around
how
to
get
their
mechanics
to
the
right
place
at
the
right
time,
reducing
their
costs
and,
as
we'll
find
out
a
little
later,
it
actually
created
other
benefits
while
they
had
on
their
roadmap.
As
I
started,
learn
about
the
solution,
I've
only
known
account
for
about
a
year.
B
B
The
best
route
within
that
map
on
how
to
best
reach
those
places
right
is
kind
of
like
the,
if
you
think,
of
a
an
autonomous
robot
who
needs
to
travel
from
one
side
of
the
room
to
the
other,
with
a
bunch
of
legos
in
the
way
or
blocks
in
the
way.
How
does
it
reach
the
other
side
of
the
room
with
the
least
amount
of
energy
or
the
least
amount
of
time
right?
B
Whatever
your
metric
is
of
success,
and
so
in
this
case
they
have
inputted
several
rules
that
describe
what
success
means
to
thyssen
krupp,
and
that
might
mean
SLA
s.
It
might
mean
risk
of
contract.
It
I
mean
availability
of
mechanics
right.
It
might
mean
how
much
overtime
a
particular
mechanic
has
already
had,
or
what
how
many
hours
they've
already
worked
in
in
the
you
know,
two-week
period.
B
So
all
this
data
goes
into
off
the
planner
and
it
calculates
the
best
route
of
the
fleet
that
they
have
to
reach
these
destinations
and
then
at
the
end
they
merge
that
data
into
a
larger
set
and
then
at
the
end
you
see
we
have
to
you
know.
You
know
this
is
simplified
right,
but
you
have
two
different
outcomes.
One
has
a
worse
score
and
one
has
a
better
score,
and
so
you,
you
obtain
a
the
most
optimal
solution.
Now
out.
The
planner
can't
run.
A
If
you
think
about
it,
because
before
we
would
give
the
mechanic
30
days
worth
of
tickets-
and
we
would
just
say
here-
you
go
here's
your
workload
go
to
this
for
the
month
before
we
implemented
up
the
planner,
that's
how
they
would
work,
but
with
the
implementation
about
define
our
net
planning
the
best
route.
Now
we
can
give
the
mechanic
an
achievable
schedule
every
week
so
and
we
knows
what
buildings
that
he's
needs
to
go
to,
and
it's
not
just
here's
a
load
of
work,
and
this
is
what
you
need
to
do.
B
And
the
other
interesting
thing
about
the
planner
is
that
the
longer
you
let
it
run
the
better
outcome
at
the
cheeves,
but
it's
called
it's
something
called
diminishing
returns
and
I
like
to
describe
diminishing
returns
as
eating
Oreos.
The
first
Oreo
is
great,
the
second
one's
pretty
good.
The
third
is
all
right
and
after
the
hundredth
one
you're
like
I,
don't
really
want
an
Oriole
anymore.
B
Where
at
some
point
you
reach
a
point,
I
can
run
it
for
twenty
minutes
as
an
example
and
I
save
you
know
twenty
five
minutes
or
you
know
whatever
of
time,
but
if
I
let
it
run
for
30
minutes,
then
I
only
save.
You
know
two
more
minutes.
You
know
of
my
route
right,
so
you
know
getting
diminished
returns
at
some
point.
B
So,
with
this
project
that
we've
done
with
Thyssen
crop,
it's
been
a
two-year
pilot
he's
been
live
for
six
months,
and
so
the
media
business
impacts
they've
seen
is
there
completed
maintenance
rows
from
50%
to
75%
and
their
goal
is
within
a
hundred
percent
within
a
year
which,
because
of
these
initial
success,
they
have
a
lot
of
confidence
that
they're
going
to
attain
their
goal.
They've
increased
their
maintenance
planning
and
their
scheduling,
reduced
cancellations.
B
Well,
I
guess
here:
I
made
a
mistake:
they've.
Instead
of
reducing
their
billing
compliance,
there
they've
increased
their
billing
compliance
and
they've
met
their
SLA
x'
reduce
the
number
of
devices
that
the
mechanics
had
to
carry.
They
divided
the
how
many
devices
were
mechanics
caring.
Well,
they
were
having.
B
So
not
only
is
it,
you
know:
increased
cost
savings
from
having
to
purchase
more
mobile
devices,
it's
a
less
of
a
burden
on
the
actual
mechanics
themselves,
because
there's
less
devices
are
having
to
keep
track
of
and
having
to
update
and
having
a
doing
SS.
Also,
the
reduced
capex
costs
by
reducing
by
utilizing
cloud
they're
using
our
managed
red-hats
managed
to
offering
the
they
can
also
see
the
availability
and
the
mechanic
overall
over
allocation.
So
they
could
see
how
their
mechanics
are
able
to
achieve
their
goals
or,
as
I
think
before.
B
Like
Josh
said
they
would
have
this
30-day
window
of
here's
all
this
stuff.
He
have
to
do
and
what
was
happening
is
not
only
were
there
a
lot
of
overtime,
but
there
was
a
lot
of
you
know
so,
I
guess
each
mechanic
is
measured
on
some
key
performance
indicators
and
that
determines
whether
you
know
they're
still
employed
or
they
get
bonuses
or
they
get.
B
So
from
a
is
what
I
call
it
non
measurable
impacts
like
I?
Don't
have
like
metric
information
on
this,
but
it
was.
These
are
definitely
still
impacts
for
the
business
they're
eliminated
skill
gaps
from
their
infrastructure,
so
they
don't
have
to
have
as
much
knowledge
in
their
internal
company
I
team.
In
order
to
manage
these
servers
and
I
manage
the
network
order,
managed
SLA
in
order
manage,
you
know,
outages,
we
have
three
regions
of
availability.
B
B
A
lot
of
this
comes
into
their
machine
learning.
They
call
it
max
integrated
with
the
mobile
technology,
allows
better
source
on
site
using
a
database
of
the
fault
codes,
so
not
only
the
NA
as
they're
there.
It
could
is
using
this
analytics
on
the
backend
to
determine
okay.
You've
had
you
know
so
many
calls
at
the
site
or
you
you're,
about
to
go
to
this
break
fixed
situation.
Here's
the
historic
data
on
the
here's,
some
analytical
and
inference
that
we
can
make
from
this.
B
These
are
some
tickets
that
are
similar,
so
it
gives
the
technicians
more
tools,
so
they
could
resolve
the
issues
better
and
when
the
technician
is
on
site
and
is
able
to
gain
access
to
information
to
help
them
solve
the
issue
while
they're
there,
it
means
you
don't
have
to
call
a
technician
back
again,
the
next
day
or
the
next
week
with
more
information,
because
they
didn't
have
to
go
back
and
research.
They
had
it
all
at
their
fingertips
and
then
security
was
also
for
the
whole
solution
set.
A
Yeah
because
I
mean
there
was
one
thing
that
we
look
for,
we
were
trying
to
move
things
more
off
ground
and
we
wanted
something
in
the
managed
environment
and
the
partnership
that
well,
you
guys
had
everything
that
we
needed.
Plus
you
had
the
expertise
with
the
consulting
side
at
us
to
help
us
bridge
that
skill
gap
that
we
didn't
have
in
the
beginning
and
now
I
mean
as
you've.
Seen
from
the
slots,
I
mean
we've.
We've
gained
tremendous
performance,
encryption
improvements
based
on
the
applications,
so
we
got.