►
Description
The folks behind Koda, the world’s first decentralized dog, explain how IPFS + AI is a match made in heaven for robotics.
For more information on IPFS
- visit the project website: https://ipfs.io
- or follow IPFS on Twitter: https://twitter.com/IPFS
Sign up to get IPFS news, including releases, ecosystem updates, and community announcements in your inbox, each Tuesday: http://eepurl.com/gL2Pi5
A
A
So
I
wasn't
particularly
interested
in
coda.
Coda
utilizes
go
ipfs
with
the
hd,
you
know
at
the
core
and
is
a
a
pretty
amazing,
robotic
dog.
So
let's
talk
about
it
so
way
back
in
2018,
a
group
of
engineers
and
designers
at
coda
really
wanted
to
build
a
decentralized
computing,
ai
platform,
and
they
had
experience
with
with
different
form
factors
and
what
really
came
to
light
was
if
you
could
come
up
with
something
that
was
entirely
useful.
A
A
A
You
know
things
that
are
coming
out
in
six
and
seven
joining
five
right
now,
but
it's
really
set
up
so
that
it
can
allow
a
user
to
select
what
projects,
for
example,
for
decentralized
computing,
but
also
even
in
the
default
state,
allow
it
to
be
a
node
and
that's
that's.
The
idea
to
you
know
proliferate
this
idea
of
decentralized
computing
without
the
hindrances
and
restraints
of
you
know,
using
you
know,
kind
of
the
traditional
traditional
vendors,
let's
say.
A
A
A
little
bit
overview
and
again
this
is
really
just
meant
to
be
an
intro,
I'm
not
going
too
deep
into
this
I'll,
go
a
little
bit
into
the
architecture
for
those
that
are
interested,
but
essentially
it's
very
cute.
One
of
the
first
things
I
had
come
to
mind
is
I've.
Seen
just
like
I'm
sure
a
lot
of
you
on
youtube.
You
know
kind
of
what
robotic
dogs
can
look
like
and
they're,
certainly
very
capable
things
out
there.
This
one
has
a
head,
which
I
thought
was
a
great
start.
A
A
It
will
arc
its
head,
so
kind
of
does
a
cant
you
know
in
order
to,
in
order
to
to
hear
you
know
better
and
that's
exactly
how
a
real
dog
works
and
because
of
the
placement
of
these
sensors
and
how
the
motor
you
know,
they've
gone
through
a
lot
of
effort
and
putting
the
motors
kind
of
how
a
real
dog
works
with
the
you
know,
obviously
the
gate,
but
also
how
it
utilizes
the
sensors.
It's
got
a
very
cool
effect
in
the
system.
A
For
my
part
of
this,
I'm
particularly
interested
in
obviously
doing
my
background
in
cyber
security.
In
the
you
know,
protection
of
not
just
private
data
and
personal
data,
but
also
the
data
for
when
coda's
time
is
being
volunteered
and
we're
particularly
sensitive
to
that.
Just
thought.
I'd
mention
that.
A
So
this
this
I'm
going
to
go
into
much
detail,
but
it
does
use
a
a
fusion
of
cpus.
There
are
several,
as
you
might
imagine
it
has
14
motors.
It
has
actually
five
3d
cameras
this.
This
is
a
little
dated,
sorry
guys.
This
uses
v
slam
some
of
your
newer.
You
know
rumba
vacuum
cleaners,
roomba
vacuum
cleaners
uses
v
slam
to
navigate
and
to
judge
distance
so
does
tesla.
I
believe
mercedes
does
as
well,
I'm
not
totally
sure
on
that
guy.
A
Audibly,
one
of
the
first
one
of
the
first
ideas
for
using
coders
in
general
are
for
visually
impaired
people,
and
so,
if
you
have
someone
who
would
benefit
from
a
you
know,
living
and
breathing
seeing
eye
dog
you
know,
coda
can
be
particularly
advantageous
as
a
supplement
and
and
sometimes
a
replacement
depending
on,
what's
needed.
But
it
needs
to
be
able
to
act
and
understand
its
user,
just
like
a
living
and
breathing
animal
would
and
to
that
end
it
needs
to
understand,
for
example,
if
it's
owners
under
duress.
A
So
you
know,
if
you
can
imagine
you're
at
a
traffic
intersection
and
the
light
turns
green,
but
the
owner,
you
know,
is
having
an
issue
and
doesn't
necessarily
want
to
cross
the
street
at
that
point.
Coda
can
understand
this
as
a
duress
event,
depending
on
the
tone
and
act
accordingly
and
provide
additional
information
and
perhaps
on
route
or
what
have
you
navigations
built
into
the
into
the
system.
But
the
idea
here
is
not
just
to
kind
of
use
the
traditional
sensors
in
the
traditional
ways.
A
So
this
is
just
a
little
quick
summary.
You
know
it's,
you
know
smarter,
safer,
faster.
There
have
been
other
things
out
on
the
market.
This
is
really
its
core
was
built
to
be
a
decisionalized
computing
platform.
You
know,
codas
are,
are
you
know,
destined
to
be
able
to
teach
each
other
with
their
owner's
permission?
What
they
learn?
You
know
if
one
code
learns
how
to
navigate
a
particularly
difficult
type
of
stairs,
it
can
share
that
information
with
other
codas,
so
other
coders
when
they
encounter
similar
geometry,
can
navigate
them.
A
You
know
obviously
much
better,
and
so
that's
just
obviously
a
very
small
example,
but
that
that
really
translates
across
what
these
devices
are
able
to
learn
and
share
with
each
other.
There
is
full
control
over
what
is
shared
and
received
in
terms
of
instruction,
but
the
idea
here
is
it's
going
to
not
just
be
a
platform
for
decentralized
computing
and,
and
you
and
ai,
it's
it's
going
to
use
it
itself
in
order
to
in
order
to
learn
and
learn
quickly,
a
little
bit
about
the
company.
A
A
A
little
bit
on
timeline,
a
lot
of
people
ask
it.
Whenever
we
talk
about
this,
we
haven't
really
launched
yet
so
for
the
few
that
have
seen
what
you're,
seeing
today
a
little
bit
of
background
us
beginning
in
2018,
first
lab-based
prototype
was
in
january,
headquartered
in
silicon
valley
opened
up
in
february
here
in
silicon
valley.