►
From YouTube: CI WG demo: DesignSafe-CI
Description
Date: 1/5/18
Presenter: Ellen Rathje
Institution: UT Austin
South Big Data Hub
A
A
Here
at
the
University
of
Texas
at
Austin
and
as
the
senior
research
scientist
UT
bureau
of
economic
geology,
she
has
expertise
in
the
area
of
earthquake
engineering
and
nearing
seismology
and
remote
sensing
and
is
the
brittle
investigator
of
the
design
safety,
I
Ord,
cyber
infrastructure
for
the
NSF
funded
natural
hazards
and
engineering
research
infrastructure,
or
in
Harry,
which
has
been
going
on
here
at
UT
with
Ellen
for
two
years
now.
It.
A
B
Right
thanks
thanks
a
lot
Niles,
so
yeah
said:
I'll
talk
about
design
safe,
which
is
a
cyber
infrastructure
specifically
for
natural
hazards,
engineering
and
its
really
taken
a
great
it's.
It's
been
a
great
synergy
between
attack
and
the
earthquake,
engineering
and
natural
hazards,
engineering
communities.
B
There
we
go
so
just
to
give
you
a
broader
perspective
of
what
we're
trying
to
do
with
design
safe.
So
it's
a
web-based
research
platform
that
provides
not
only
access
to
data,
but
also
the
computational
tools
needed
to
manage
analyze
and
understand
that
data
that
we
use
to
understand
natural
hazards
and,
in
particular,
the
impact
of
natural
hazards
on
the
built
environment.
Our
vision
for
design
safe
is
to
provide
a
cyber
infrastructure
that
is
really
part
of
the
research
process.
The
research
discovery
process.
So
how
do
we
get
there?
B
We
were
supporting
end
to
end
research
workflows,
as
well
as
the
full
research
life
cycle,
all
the
way
from
when
you
just
start
thinking
about
some
research.
You
want
to
do
all
the
way
to
the
end
where
you
share
and
old.
We
published
that
data.
We
provide
cloud-based
tools
that
support
the
work
that
people
want
to
do
on
a
diverse.
B
Importantly,
we're
one
component
of
something
I'll
mention
with
it,
which
is
Neary
and
I'll
talk
just
a
little
bit
more
about
what
near
he
is,
but
our
second
part
of
our
vision
is
really
to
amplify
our
partners,
link
our
capabilities
and
not
only
the
Neary
partners,
but
really
natural
hazard
researchers
are
around
the
globe.
So
just
a
little
bit
about
Neary,
it's
NSF
funded
this
we're
providing
shared
use,
research
infrastructure
to
researchers
in
the
US
and
as
well
as
you
know,
some
of
it
can
be
accessed
abroad.
These
are
the
main
components.
B
Obviously
you
know
I'm
gonna
talk
about
the
CI
component,
but
importantly
they're.
You
know:
share
juiced
experimental
facilities
like
shaking
tables
tsunami,
wave
basins
that
people
are
generating
data
at
that
we
are
responsible
for
helping
curate
and
provide
access
to.
There
are
people
going
out
and
doing
rapid
reconnaissance
research
from
the
rapid
center
using
various
types
of
field
equipment.
B
We
need
to
be
able
to
deal
with
that
type
of
data
and
there's
also
computational
modeling
and
simulation
center,
so
we're
also
dealing
with
simulation
data,
so
lots
of
different
types
of
data
in
different
contexts
that
were
involved
in
so
we're
we're
really
a
big
part
of
linking
all
these
components
together.
So
moving
on
to
what
specifically
design
safety
provides.
If
you
go
to
our
website,
you
know
today,
you'll
see
these
four
main
headings
on
the
top
research
workbench
Learning
Center,
nearly
facilities
and
nearing
community.
B
The
research
workbench
is
really
the
place
to
do
the
research
it
involves
and
includes
our
data
Depot,
which
is
the
data
repository.
It
includes
the
discovery
workspace,
that's
where
our
tools
reside,
that
can
work
on
data,
that's
included
in
the
data
Depot
and
then
finally,
the
reconnaissance
portal,
which
is
specifically
to
provide
access
and
easy
searching
of
data.
That's
collected
during
field
recon.
So
that's
again
the
main
research
areas.
B
Of
course,
we
need
to
be
able
to
teach
people
how
to
use
the
cyber
infrastructure,
so
the
Learning
Center
provides
those
training
resources,
as
well
as
broader
engagement
with
students
and
natural
hazards,
engineers
doing
research,
the
Nereid
facilities,
as
I
mentioned,
we're
part
of
this
Neri
network,
and
so
that's
where
you
can
find
information
about
the
all
the
naree
facilities
and
then
in
our
efforts
to
bring
the
community
together
and
provide
some.
You
know
virtual
interactions.
B
We
have
the
Neary
community
section,
which
involves
not
only
things
like
news
and
announcements,
but
we've
we're
leveraging
slack
to
be
to
provide
kind
of
an
online
community
where
we
have
a
design,
safe,
slack
channel
or
slack
team,
and
when
you
join
design
safe
and
get
an
account,
you
get.
You
get
an
invitation
to
join
the
slack
team
where
you
can
interact
with
other
folks
about
design
safe
I
am
gonna
focus
mostly
on
the
research
workbench,
given
the
audience
I'm
speaking
to
so
let
me
first
talk
about
our
data.
B
Depot
importantly,
we
have
these
different
levels
of
access.
We
have
a
private
area
that,
as
a
user
you'll
see,
you'll
have
my
data.
We
also
have
an
area
where
you
can
create
a
project,
and
these
projects
are
semi-private,
where
you
can
control
access
to
selected
team
members.
So
it's
really
a
collaborative
space.
B
We
have
an
area
for
published
data
and
published
works
which
is
publicly
accessible
as
well
as
curated,
with
respect
to
our
data
models
and
then,
finally,
we
have
an
area
of
called
community
data
which
is
also
publicly
accessible,
but,
as
my
colleague
Dan
Stanzione
likes
to
say
it's
a
little
bit
of
the
wild
wild
west.
So
it's
uncured
and
you
can
certainly
search
and
see
what
you
can
find,
but
it
doesn't
follow
the
detailed
data
models.
Also
within
the
data
Depot.
B
You
know
we,
we
we
support,
upload
and
download
via
various
mechanisms,
everything
from
as
simple
as
a
web
browser
we
integrate
with
different
cloud
service
providers,
box,
Dropbox,
Google
Drive,
so
your
your
cloud
service
providers
will
show
up
within
the
data
Depot
and
you
can
my
data
our
project
and
then
for
those
bigger
bowls
transfers.
We
use
what
you
all
were
just
talking
about,
like
locusts,
so
we've
integrated
all
these
different
types
of
mechanisms
to
move
data
around
when
you're
in
the
data
Depot.
B
You
can
also
manage
and
preview
files
and
then,
ultimately,
through
your
projects,
you
can
organize
curate
and
ultimately,
then
publish
your
data.
So
let's
just
answer
to
give
you
a
little
look
on
what
the
data
Depot
looks
like
and
in
particular
what
a
project
looks
like
when
it's
published
so
the
first
thing,
you'll
notice.
Here
on
the
left-hand
side,
you
can
see
my
data,
my
projects,
box
comm.
B
So
this
is
where
you
can
navigate
to
all
those
different
levels
within
the
data
Depot,
but
I
can
navigate
into
the
published
area
and
look
at
this
very
specific
project
similar
to
a
journal
paper.
We've
got
our.
You
know
our
P
I,
which
would
be
equivalent
to
authors
date
of
publication,
some
keywords,
a
description
and
then
within
this
project
we
can
have
multiple
experiments
that
have
been
published.
So
this
experiment
was
about
large-scale
laboratory
experiments
of
wave
impacts
on
vertical
cylinders.
B
We
can
scroll
down
to
this
experiment
and
take
a
closer
look
and
very
quickly
by
scrolling
down
the
page.
You
can
find
the
data
in
some
of
our
previous
instances,
with
feeling
with
earthquake
engineering
data
we've
developed
data
repositories
that
were
not
quite
as
easy
to
find
the
data.
So
one
of
our
goals
here
was
to
easily
find
the
data
and-
and
that's
shown
here
under
the
various
events
now
you
may
not
fully
appreciate
the
organization
of
the
data.
B
So
if
you
click
on
relations
here,
you
can
see
basically
a
treaty
that
explains
how
the
different
elements
of
the
data
are
linked
together.
So,
for
instance,
for
this
we
have
the
model
configuration
in
blue
the
sensor
configurations
in
green,
and
then
these
events
listed
that
are
associated
with
this
sensor
and
the
model
the
beauty
here
is.
This
is
flexible.
B
Not
every
user
is
going
to
organize
their
data
the
same
way,
although
we
you
do
need
to
have
a
model,
a
sense
or
an
event,
but
it
allows
the
users
to
organize
the
data
that
makes
sense
to
them
and
document
how
it's
organized.
So,
ultimately,
a
new
user
can
come
in
and
and
understand
the
data
and
use
it
importantly.
Both
the
project
and
the
experiment
have
a
DOI,
and
so
that
excitable
with
authors
and
publishers,
etc,
and
so
importantly,
we
can
find
that
data
easily
moving
forward.
B
I
thought
I'd
take
a
second
and
mention
our
data
management
philosophy,
which
I've
kind
of
alluded
to
in
my
previous
couple
of
slides.
Is
we
really
want
to
make
this
as
easy
as
possible
for
users
to
store,
share
document
and
publish
that
data
and
we're
focused
on
achieving
their
goals
rather
than
setting
up
a
lot
of
rules?
I
had
a
colleague
who
said
I
want
tools,
not
rules
because
in
the
but
in
the
previous
iterations,
again
lots
of
rules
and
nobody
likes
to
put
a
lot
of
metadata
together.
B
So
we
want
to
make
sure
things
are
documented
well
enough
that
people
can
reuse
the
data
but
not
be
too
onerous
on
the
data
generator.
So,
as
I
mentioned,
flexible
data
models
that
support
how
organizers
want
to
research
or
organize
their
data,
and
importantly,
we
have
developed
different
data
models
for
kind
of
the
major
different
types
of
projects.
So
I
showed
you
an
experimental
project,
but
if
you
have
a
simulation
project,
the
organization
is
a
little
bit
different.
B
B
Now,
on
the
other
side,
we
have
the
discovery
workspace.
So
this
is
where
we
provide
various
types
of
tools
for
researchers.
We
have
a
lot
of
computational
tools,
I'll
focus
on
the
data
analysis
tools
that
we
have
available,
one
that
we're
really
promoting
is
is
Jupiter,
notebooks
and
I
know
you
guys
have
heard
about
Jupiter,
I
guess
in
one
of
your
previous
meetings.
We
also
have
MATLAB,
of
course,
you're
all
familiar
with
MATLAB,
but
we
have
a
cloud-based
version,
so
you
can
work
again
on
those
files
that
are
located
in
our
data.
B
So
let
me
just
give
you
some
examples
of
how
we're
using
Jupiter
some
of
our
users
are
developing
electronic
data
reports
and
it
really
provides
an
interactive
interface
with
the
data
that
they're
publishing.
So
this
data
report
in
the
past,
but
would
be
300
page
PDF
plotting
all
of
the
sensor
data,
for
you
know
a
large
amount
of
the
experiments
it
could
be.
As
I
said,
hundreds
of
pages
now
this
experiment,
this
electronic
data
report.
B
So
our
reconnaissance
portal
is
another
part
of
the
design
safe,
which
provides
an
easy
way
to
identify,
publish
and
archive
datasets
that
are
in
the
data
Depot,
but
they're
from
reconnaissance
events.
So
this
fall
was
an
amazing
time
for
natural
hazards.
Reconnaissance
I
must
say
so.
Within
about
a
6-week
timeframe.
B
We
had
four
hurricanes
in
an
earthquake,
so
we
know
all
the
Hurricanes
and
the
in
the
Gulf
and
the
Caribbean,
and
then
we
had
the
Mexico
City
earthquake,
and
so
what
you
can
see
here
is
when
we're
usually
looking
for
data
from
a
specific
event.
B
We
can
more
easily
identify
them
just
from
a
simple
map
like
this,
and
so
we
can
then
click
on,
for
instance,
and
I'm
gonna
go
to
a
different
event,
not
one
shown
here,
but
we
did
a
reconnaissance
for
the
Kaikoura
earthquake
in
New
Zealand
in
2016
when
you
click
and
look
at
the
details.
For
that
event,
the
available
datasets
are
listed.
Some
of
these
datasets
will
be
in
design
safe.
B
Some
of
them
may
be
outside
of
the
design
safe,
but
now
you
can
easily
find
them
and
then,
for
instance,
if
we
go
to
the
landslides
inventory,
oh,
we
can
open
one
of
the
files
in
has
mapper.
So
this
is
an
interactive
map
viewer
and
in
this
case,
what
I'm,
showing
here
is
the
red
polygons
represent
the
landslides
that
were
generated
by
that
Kaikoura
earthquake
and
the
blue
and
the
pink
lines
represent
our
GPS
tracks.
B
While
we're
in
the
field,
I've
got
some
geo
coded
photographs
that
have
been,
you
know,
we
just
drag
and
drop
and
they're
put
into
the
appropriate
location.
So
this
really
allows
us
to
assemble
that
type
of
geo
spatial
data
to
share
with
our
teams
and
share
with
other
researchers,
and
we've
made
use
of
this
significantly
more
recently
in
the
in
the
hurricane
events
in
terms
of
people
looking
at
damage
to
specific
structures
and
having
a
way
to
archive
those
results,
as
opposed
to
a
bunch
of
photographs
that
are
annotated,
so
I'm
gonna.
A
C
B
Haven't
gone
through
that
use
case
yet,
but
we're
certainly
open
to
identifying
which
datasets
you
may
be
interested
in
operating
with.
So
we
haven't
gone
down
that
path,
yet
we're
doing
a
little
work,
we're
considering
like
open
topography,
for
instance,
we
just
started
some
discussions
with
them
and
seeing
how
we
can
interact
with
them.
So
that's
something
that's
in
our
in
our
interest,
but
we
haven't
done
anything
specifically
yet
so.
C
D
C
So
I
think
the
answer
probably,
is
that
there
is
no
similarity.
Assumer
is
a
hydrologic
modeling
capability
for
I
guess
it's
called
structure
for
unified,
multiple
modeling
alternatives,
and
it's
really
a
systematic
way
to
go
through
different,
representing
different
processes
in
models
to
try
and
evaluate
which,
which
work
best
so
I
think
yeah
I
mean
I,
don't
think
it
has
very
much
in
common
with
what
you've
shown
now.
C
B
B
Yeah
I
should
mention
I'm,
not
sure
I
specifically
said
it,
I
mean
through
design
safe.
We
provide
users
with
those
computational
tools
have
the
ones
that
require
it,
have
access
to
HPC
and
so
we're
trying
to
lower
the
bar
or
lower
the
barrier
to
getting
people
to
you
to
take
advantage
of
high
performance
computing
for
various
types
of
tools,
and-
and
so
you
know
that
sounds
like
at
least
on
the
summa
side.
That
might
be
kind.
You
know
something
that
you
might
consider
as
well.
Yeah.
C
C
B
And
I
should
mention
that
our
mandate
from
NSF
is
earthquake
in
wind
storms,
which
then
hurt
you
know.
Hurricane
storm
surges
and
water
comes
into
play,
but
the
hydrologic,
like
flooding
side
of
things,
does
not
specifically
fall
into
our
purview,
but
we're
very
happy
to
collaborate
on
with
folks
in
any
way
that
relates
to
natural
hazards.
So
we'd
be
very
interested
to
hear
more
about
the
the
flooding
side
of
things
right.
C
A
So
yeah
I
think
it'd
be
interesting
to
start
to
look
in
in
the
out-years
and
so
how
the
Federation
between
these
projects
could
take
place,
so
wrap
it
back
to
what
we
were
talking
about
with
the
with
the
irods
portion
as
well.
So
I'll
just
try
mine,
real
quick
and
say
that
a
lot
of
the
design
safe
stuff
is
based
on
the
agave
frameworks
that
we've
developed
here
attack
which
power
the
now
cybers,
formerly
I,
plant,
where
our
federation
capabilities
there.
It's
just
you
know
at
hours
in
the
day
and
money's
in
the
bank.