►
From YouTube: Open Source GPS Tractor
Description
With Matt Droter.
See Open Source Robotic Tractor - http://opensourceecology.org/wiki/Open_Source_Robotic_Tractor
What you see here at Open Source Ecology is an ambitious program based on a volunteer effort. To help us reach the goals - please consider joining as an OSE Developer in 2017-
http://opensourceecology.org/wiki/OSE_Developers
Take a minute to subscribe to our email newsletter (updates, workshops, etc): http://bit.ly/1LtcM44
B
Sorry
introduction
we're
here
with
Matt
rotor,
drone
and
Marcin
here
for
open
source,
ecology,
we're
talking
about
the
autonomous
tractor
and
how
to
go
about
it.
So
so
Matt
does
did
some
work
with
Ross
the
robotic
operating
system
we're
strategizing
on
how
to
implement
that
for
the
tractor.
So
let's
do
it.
What
is
currently
like?
What?
What
would
you
say?
Matt
would
be
currently
the
best
way
to
go
about
it
like
using
the
technology
that
maybe
that
you're
familiar
with
that?
A
B
You
know
some
people
that
are
doing
Ross
tractors.
Yep.
Oh
really!
Oh
nice!
Nice!
That's
that's
great,
because
if
we
have
case
examples
that
would
make
it
easy
for
us
yeah.
Are
you
in
contact
with
these
people.
B
See,
okay,
I
haven't
heard
about
any
of
this,
so
this
is.
This
is
good
news.
Are
you
able
to
reach
out?
Have
you
reached
out
to
them
yet,
or
were
you
planning
to
do
that?
B
A
B
B
B
A
A
Some
people
are
driving
cars
with
this
cell
phone
and
then
so
it
has
a
camera
already.
Built-In
has
the
communications
already
built
in
you
just
run
Linux
on
it,
and
then
you
can
interface
it
even
through
the
like
canvas.
There's
canvas
on
your
vehicle.
Some
inexpensive
methods
are
like
you're
saying
through
microcontrollers,
you
can
talk
to
it
and
then
you
can
have
multiple
microcontrollers
on
the
vehicle
and
then
maybe
run
brain
that
is
doing
the
visioning
process
or
navigation
and
controls.
A
A
B
A
That's
a
hard
part,
isn't
yeah,
because
once
you
can
get
it
under
computer
control
and
then
there's
already
auto
pilots
out
there,
so
you
can
look
at
the
picks
off
the
pixhawk.
Could
just
we
could
just
there's
software
already
done
for
it.
We
can
literally
download
the
software
and
start
driving
your
vehicle.
B
A
A
They
are
you've,
been
doing
it
for
a
number
of
years.
There's
a
large
community
so
basically
just
that
zero
to
five
controlling
the
RC
module
one
they
interfaces
directly
to
it.
So
we
could
practice
with
a
little
RC
car
and
drive
it
around
and
make
sure
everything's
okay
and
take
the
same
brain
out
of
that
and
put
it
on
your
tractor
and
drive
to
that.
Now
the
your
tractor
looks
like
it's
a
skid
steer
model
and
the
other
ones
control
just
a
regular
car,
but
they
have
something
secure
version
awesome.
They.
B
B
A
A
B
A
A
B
A
B
B
A
A
B
A
B
A
B
A
A
B
A
B
B
B
B
This
is
great
I
love
this.
We
can.
B
A
B
B
Okay,
I
see
you
in
there.
This
is
what
we
did
initially
I
don't
have
to
get
into
that,
but
we
were
talking
about
the
micro
track
here
and
following
lines
that
are
basically
not
lines
about
stakes
on
the
ground
using
computer
vision,
but
now
I
think
this
is.
We
can
return
to
this.
It's
just
some
design
points
here.
We're
gonna
look
at
putting
in
stakes
into
the
ground
like
little
PVC
stakes
every
15
feet
and
try
to
follow
them.
But
let's,
let's
start
on
what
we
can
do
here
so
our
OS.
B
B
And
the
idea
of
Ross
is
that
it
makes
it
easy
to
to
do
like
you
know
like,
for
example,
if
you're
starting
from
scratch,
you'd-
probably
look
at
raspberry,
PI's,
arduino
z',
so
ross
basically
has
a
bunch
of
libraries
that
enable
you
to
work
effectively
with
us.
Can
you
can
you
describe
what
ross
has
to
offer
offer.
B
B
A
B
B
B
B
B
A
B
A
A
A
B
A
A
B
A
A
B
And
there
they've
got
the
actual
schematics
online
for
the
actual
boards
does.
Does
anyone
build
them
them
themselves?
I?
Don't.
B
B
Because
something
like
on
the
order
of
you
know
like
for
the
initial
want
to
keep
everything
as
if
it's
the
first
prototype
say
we
want
to
do
a
chicken
tractor
across
an
acre
field.
You
know
we
can.
We've
got
a
field,
that's
fifty
feet
by
say
three
hundred
feet.
B
Yeah,
let's
take
that
as
a
use
case.
Let's
that's
design
for
that
that
we
can
do
I,
think
as
it
will
be
nice
to
show
that
we
can
do
a
chicken
tractor
with
chickens
in
there.
That's
we're
maybe
using
the
tractor
to
pull
pull
around
like
a
hundred
chickens
in
a
chicken
tractor
which
might
be
like
one
or
few
boxes
of
chickens.
How
does
that
sound
for
a
use
case?.
A
A
A
B
We
can
design
them
to
to
fit
so,
but
I
was
thinking
like
there's
a
limit
to
I
mean
it's
gonna,
be
easy
to
do
a
train,
the
chicken
train,
it's
harder
to
do
a
bigger
box.
So
it's
thinking
a
chicken
train
with
enough
room
to
turn
around
on
a
fifty
feet.
That's
kind
of
tight
you
can't
do.
The
train
can
be
too
long.
Yeah.
A
So
I've
seen
a
video
where
they
have
like
a
little
tractor.
That's
solar,
powered
and
they're
pulling
chicken
tractors
across
this
meadow
and
it's
they
have
a
straight
straight
bar
or
coming
off
of
the
back.
There's
a
long
straight
bar
coming
sideways
a
rope
pulling
the
bar
with
a
three
chicken
or
so
they're
parallel
coming
coming
down
the
field,
uh-huh.
A
A
B
Cuz,
it
would
be
nice
to
show
that
that's
actually
functional
and
for
real
I
was
thinking
a
watering
station
that
you
would
dock
to
a
like.
You
know,
like
the
trains
with
the
overhead
watering
thing,
the
steam
trains
you
dock,
to
a
watering
thing.
We
can
set
that
up
and
then
it
would
you
like
that.
That's
that's
awesome!
Yeah!
Let's
do
it!
So
that's
that
way
you
can
I
mean
I'm
talking
about
real
practical
use
case.
B
This
is
not
like
fooling
around
this
is
we're
gonna
actually
produce
chickens
and
have
them
free-range
by,
and
do
you
follow
the
idea
that
you
put
one
solar
panel
of
300
watts
on
it
to
drive
it
for
six
hours
during
the
day?
That's
what
we
would
like
to
do.
So
one
of
our
guys
is
doing
that
we're
doing
a
small
electric
power
cube
which
produces
a
tractor
that
moves
three
feet
per
minute.
B
B
B
What
else
with
3d
printed
case,
of
course,
of
course,
well
I
got
inspired
about
3d
printed
cases
cuz,
we
were
making
the
filament
maker
the
other
day
and
you
can
make
you
can
3d
print
big
cases
too.
If
you
glue
them
together,
we
glue
it
with
just
crazy
glue.
So
that
was
good.
Ok,
pencil
me
your
accuracy
for
1k.
Now
we
don't
do
that.
That's
that's
too
much
money
for
now.
B
B
A
It
just
depends
on
where
it
is
and
what
we're
doing
just
like
you
said
with
the
poles,
it's
a
known
environment
like
this
pin
then,
and
you
can
change
the
infrastructure.
Then
you
have
different
choices.
You
don't
need
to
know
your
global
position,
because
I'm
just
going
up
and
down
in
this
field.
So
if
you
can
use
different
things,
come.
A
A
B
A
Lasers
are
a
really
expensive
technology
to
use,
so
the
lowest
cost
is
cameras,
but
if
you
can
interact
with
your
environment,
putting
up
little
fiducials
putting
up
if
the
camera
is
having
a
hard
time.
Reading
the
environment
or
distinguishing
the
environment
we
can
cheat
just
by
putting
the
posters
of
you
know,
blocks
the
different
colors.
It
makes
it
easier
for
the
camera
to
see
you
know
where
it
is.
Can.
A
B
A
A
B
Cuz,
let's
see
this,
we
always
talk
about
scalability.
So
can
such
a
solution,
the
GPS
solution
can
definitely
scale
to
any
area.
In
fact,
it's
positively
scalable
meaning
it's
easier
to
operate
at
larger
scale
for
the
field
size
right.
A
B
B
B
B
A
B
A
B
B
B
B
A
B
A
B
B
B
A
The
discussion
forum
and
start
a
sub-project,
maybe
if
there's
enough
people
better
interested
so
right
now,
I've
asked
a
few
of
the
developers
that
I
know:
iroh,
Boston
and
Missouri,
and
one
of
the
guys
committed
to
come
and
we're
gonna
have
a
couple
other
ones
so
and
then
so
that's
the
way
it
is
now.
Okay,.
B
Excellent,
so,
let's
see,
did
you
fill
out
the
response
form
on
the
for
the
workshop
did
I.
Send
you
that
oh
I
haven't
sent
you
that
so
I
have
a
I
need
to
send
you
that
every
just
a
response,
so
you
can
provide
some
logistics
information,
uh-huh
I'll.
Do
that
and
the
forum
is,
you
said,
that's
a:
is
it
a
forum
or
what
is
it
you
said?
Sorry?
Can
you
put
a
link
into
the
document
for
where
that
discussion
takes
place.
B
A
B
B
A
B
B
A
A
B
A
B
What
what
would
you
say
would
be
the
best
way
to
I
mean
RC
controller?
Yes,
I
can
understand
that
it
would
talk
to
the
Arduino.
Somehow
cuz
I
want
to
just
kind
of
sculpt
it,
because
we
we
don't
have
a
lot
of
time.
We
should
kind
of
minimize
the
discussion.
If
we
know
a
ready
route
of
implementation,
like
our
dwee
knows,
are
acceptable
because
we
use
like
for
us.
We,
we
use
a
finite
amount
of
components
in
the
global
village.
Construction
set
all
together,
such
that
it's
manageable,
like
on
the
community
scale.
B
B
A
B
Yeah
we
want
to
be
a
whole
open
source
yep.
That
sounds
good.
So,
let's
list
catch,
try
to
figure
this
out
open
source,
GPS,
Raspberry,
Pi
Arduino,
with
solenoids
driving
the
motors
RC
controller
that
can
talk
directly
to
the
Arduino.
Would
it
be
talking
directly
to
the
Arduino
or
to
the
through
the
Raspberry
Pi?
We
should
probably
set
it
up.
So
it's
talking
directly
without
the
Raspberry
Pi,
so
the
Raspberry
Pi.
B
B
The
only
thing
is
this
concerns
parallel
development
and
also
the
simplicity,
like
the
way
reason
why
we
want
to
do
the
Arduino
with
RC
controller.
Is
that
we
don't
have.
If
we
just
have
remote
control
application,
we
don't
want
to
involve
the
Raspberry
Pi
and
besides
it
allows
for
a
parallel
development
path
where
a
person
is
developing
the
RC
controller
with
Arduino
independently.
So
we
can
say
that
is
one
parallel
development
path
by
some
developer,
which
is
independent
totally
of
the
Raspberry
Pi.
And
then
we
can
say
when
we
invoke
the
Raspberry
Pi.
B
B
A
B
Right
well,
we'll
develop
a
system
here
where
we
are
using
that
for
the
GPS
right,
so
in
this
instance,
we
do
or
as
Barry
pie
with
GPS
in
parallel
with
the
plain
Arduino
module
and
we're
using
one
Arduino.
So
we're
not
gonna
do
like
to
Arduino
score
this
because
we
can
just
connect
to
that
same
one.
B
B
A
B
Yeah,
whatever
don't
don't
confuse
people
with
the
last
four
pages
which
were
on
the
plain
computer
vision
route,
so
we're
not
really
getting
there
at
this
point
unless
there's
enough
interest
that
somebody
wants
to
ride
on
top
of
that.
But
if
we
do
go
with
computer
vision,
we
would
have
to
design
that
as
a
parallel
process,
so
that
it's
not
interfering
with
this.
B
A
B
Okay,
yeah,
that
might
be
a-
we
can
definitely
invite
people
to
for
further
development
on
that.
For
now
we
should
focus
on
on
this.
For
the
instance
of
October
25
26,
it's
coming
up.
It's.
We
don't
have
time
to
do
that
right
now,
but
we
can
definitely
get
people
started
on
just
I
mean
will
will
Dec
will
then
to
that
after
the
workshop,
but
for
now
I
think
we're
out
of
time
for
the
morn
Bish's
development
do.
A
B
B
Time
we'll
get
them,
we
go
to
the
first
replication
workshop
once
we
give
this
perfected
well,
I
mean
by
all
means,
I
mean
this
is
a
constant
evolution
process.
So
we
do
the
first
prototype
here.
If
it
works,
we
can
talk
about
running
a
workshop
relatively
quickly
after
that,
like
if
you
want,
like
a
month
after
that
or
two
months
or
something
I'd
love
to
see
that
happen.
Yeah,
okay,
yeah,
and
by
that
time
we
connect.
A
B
B
A
B
A
B
B
Yeah,
the
open-source
robotic
tractor
page
doesn't
have
much
in
there,
except
for
the
initial
discussion
on
what
we
were
talking
about
for
the
there's,
a
video
there
regarding
the
discussion
on
the
computer
vision
tractor.
So
but
it's
that's
related,
but
yeah
yeah
there
you
go
so
there's
your
log
and.
B
A
A
A
B
So
and
yeah
are
you
in
there
in
the
working
doc?
So
if
you
can
paste
the
Twitter
link,
which
is
the
second
bullet
point-
oh
yeah
you'd
you
jumped
in
there
again,
okay.
B
B
B
B
A
B
B
A
B
Yeah
there's
we
have
internet
in
a
hab
lab
there
with
a
building
look
at
where
we
can
do
some
of
the
work
yeah
yeah.
We
could
do
that.
Okay,.
A
B
Got
what
we
have
that's
available
right
now
is:
is
that
it's
the
hab
lab
it's
our.
It's
got
like
12
enough
space
for
12
people,
but
that
is
just
multi.
It's
got
it's
a
dorm
dorms
with
a
couple
of
people
per
room.
You
people
for
room.
So
if
you're
comfortable
with
that,
you
can
do
that.
If
you're
not,
you
can
get
a
hotel.
B
B
A
A
A
B
Man,
that's
funny:
I,
yeah,
okay,
well,
I
didn't
recognize
the
name
and
it's
I'll
see
it
in
my
emails,
cuz
I
switch
to
emails,
I
thank
2012
or
so
Wow.
Okay.
What
so
you
know
all
this?
Okay,
I
didn't
I,
didn't
totally
didn't
connect
good
to
see
you
again
then
welcome
back
okay,
okay,
Wow,
no
I!
You
must
have
changed
cuz,
I,
I,
totally,
don't
recognize
you
at
all.
I
mean
I.
Remember
a
bunch
of
those
guys
I
totally.
A
B
A
B
So
far,
the
only
thing
we
have
going
on
the
regular
basis,
our
3d
printer
workshops
and
we've
done
a
bunch
of
CB
press
builds.
The
largest
revenue
we
ever
got
was
from
the
Eko
home
workshop
where
that
model
works.
But
the
idea
is
that
it's
sporadic,
we
can't
do
it
regularly,
so
we
started
the
first
thing
that
we
started.
B
That's
regular
for
regular
revenue
is
the
3d
printer
builds
which
we're
trying
to
run
just
about
every
month
and
we're
trying
to
get
a
stable
revenue
model
from
that
that
we
can
actually
get
other
people
to
do
that
in
a
really
replicable
way.
So
we're
working
on
that,
but
I
mean
the
amazing
thing
just
to
reminisce
about
some
of
these
things.
It's.
It
has
not
happened
where
people
just
come
in,
get
together
for
a
little
bit
of
time
and
then
come
out
with
a
product.
B
It's
like
everyone,
quits,
the
the
work
doesn't
get
done
to
the
finish
line.
So
we've
got
the
last
mile
problem
on
just
about
everything
and
we're
just
really
nailing
I.
Think
we're
at
the
cusp
of
nailing,
the
3d
printer
CB
press
power
cube
power
cube
is
like
ready
for
it,
ready
for
action
totally
and
tractor.
B
Of
people,
it's
financial
pressure,
some
it's
like
they
don't
see
enough
of
the
vision,
so
they
just
go
off
on
their
own
project.
Thinking
that
see,
because
we're
committed
to
a
very
certain
way
of
doing
things
because
it's
a
set
and
they
say-
oh
well,
it's
easier
to
do
it
this
way
and
I
just
go
off
on
their
separate
Fork
instead
of
saying.
Okay,
we're
gonna
still
keep
together,
because
we
are
still
gonna
get
a
good
product.
They're
thinking,
I'm
just
gonna
defect,
because
it's
gonna
be
easier.
So
a
lot
of
times.
B
B
B
Yeah
I
think
so
I
mean
cuz.
I
mean
we're
definitely
running
into
some
of
that.
You
know
like
when
I
had
the
brick
press
back
in
20,
the
second
prototype,
the
automated
version
I
thought
wow
like
once
we
get
it.
I
was
at
that
point.
I
went
through
my
open
source
kind
of
rites
of
passage
the
open
source,
rites
of
passage
when
I
thought.
Okay,
now
we've
got
this
brick
press.
Let's
see,
can
we
open
source
this
or
can
we
not
because
this
is
so
good
and
it's
gonna
just
spread
all
over
the
place.
B
So
at
that
point,
I
had
the
rights
of
passage,
saying
no
we're
absolutely
committed
to
open
source,
but
I
thought
that
back
in
this
was
probably
like
twenty
ten
to
twenty
two
thousand
nine,
where
I
went
through
those
rites
of
passage
where
I
thought
it
was
finished,
it's
done
now.
Everyone's
gonna
copy
it
and
it
didn't
happen.
Twelve
there's
twelve
are
applications
around
the
world
right
now,
as
far
as
the
CV
press
goes,
but
far
from
viral.
B
So
the
point
is
that
that
last
night
I
think
it's
the
last
mile
problem
or
if
we
develop
it
to
the
enough
of
a
finely
refined
state,
then
it's
going
to
spread
more
like,
like
everything,
documentation,
build
process.
Everything
like
so
we're
finding
out
that
to
get
that
hundred
percent
is
a
lot
of
work.
It's
more
than
anyone
wants
to
wants
to
spend,
because
part
of
that
is
documentation
and
that's
you
know
just
twice
as
much
work
as
the
design
itself
and
stuff.
So
I
think
it's.
The
bottom
line
is
just
not
easy.
A
A
B
Definitely
got
a
lot
of
trucks,
I
mean
the
brick
press.
I
mean
you
can
build
it
for
that.
You
can
build
a
problem
between
three
and
five
thousand
dollars
the
competitors
sell
for,
like,
like
fifty
thousand
for
the
level
of
a
machine,
but
given
that
nobody
went
out
there
still
to
just
take
that
and
make
a
business
out
of
it,
I
mean
that's
the
thing
that
just
blows
my
mind,
which
says
that
I
mean
people
are
slight:
we're
slaves,
we
we
are
not
entrepreneurial
enough
as
a
society.
B
You
know
because
I
still
think
I
mean
that
business
opportunity
there
is
huge.
But
if
you
look
at
us
too,
we're
not
pursuing
it
actively
because
we're
doing
all
these
different
projects,
but
I
would
say
somebody
out
there
that
just
you
know
just
takes
that
they
just
run
with
it
and
make
a
good
life
out
of
out
of
self
building
and
selling
those.
But
nobody
has
so
that's.
That's
the
part
that
kind
of
blows
my
mind.
B
It
just
shows
that
people
are
not
as
an
entrepreneurial
or
free
to
do
different
things
as
we
would
like,
and
definitely
the
entrepreneur
types
they're
gonna
go
for
the
low-hanging
fruit
they're.
Just
you
don't
get
funded,
get
some.
You
know
get
some
funding
get
investors
do
some
stuff
that
doesn't
really
change
the
world
and
they
go
away
with
that
happy
because
these
things
are
more
like
an
the
regenerative
more
like
kind
of
mundane,
boring
stuff
that
we
do
and
I
think
that's
why
we
don't
attract.
B
Maybe
the
pure
entrepreneurs
that
just
simply
find
an
easier
way
to
make
entrepreneurship
happen.
You
know
so
yeah.
It's
it's
a
very
interesting
story,
but
I'm
definitely
surprised
I
still
remain
surprised
that
nobody
just
took
it
and
ran
with
it
yeah
yeah
cuz.
That
was
ready
for
the
taking
like
since
five
years.
You
know,
since
2006
we
got
the
first,
automated
machine
was
which
was
like
2011.
At
that
point,
I
mean
it's
good
enough.
B
It's
got
enough
of
an
integration,
and
everything
worked
out
that
someone,
someone
who
wants
to
take
it
from
you'll
say
the
ninety
ninety
five
percent
to
a
hundred
they
they
can
building
upon
all
the
work
that
we
have,
but
also
granted
it
it
did
take.
It
would
take
like
I,
would
say
that
to
get
to
ninety
percent
is
probably
one-fifth
the
work
it
takes
to
go
from
ninety
to
a
hundred.
You
know
it's
actually
much
more
to
do.
That's
the
last
mile
problem.
That's
probably
why
this
hasn't
taken
off.
B
It's
it's
definitely
the
last
mile
problem
and
and
and
we're
doing
everything
like
if
we're
doing
it,
open-source
like
everything
matters
like
from
the
open-source
CAD,
you
know
everything
that
and
just
just
now
have
we
picked
up
open-source
CAD
that
we're
really
developing
robust
CAD
capacity.
You
know
before
I
was
like
someone
would
do
it
and
you
know
an
auto,
cad
and
then
drop
off
and
we
never
see
any
of
those
files
again
cuz,
nobody
can
open
them.
You
know
things
like
that.
That's
why.
A
B
A
B
So
the
thing
is
I
think
the
critical
some
of
the
critical
elements
to
making
this
succeed
is.
You
have
to
have
the
whole
package
and
I
was
adamant
on
keeping
it
honest
throughout
meaning
no
you're
gonna
use
open
source
software.
You
know
we're
gonna
do
all
of
it
right
and
that's
why
I
think
it
took
so
many
years
to
get
to
that
place
and
I
think
we're
getting
there
and
it's
it's
it's
the
next
trillion
dollar
in
economy,
but
I
think
there's
enough.
B
You
know
difficulty
in
the
process
because
we're
trying
to
do
the
whole
thing
in
an
honest
way,
all
open
source,
so
we're
adding
these
fundamental
levels
of
difficulty
like
you
asked
me
about
on
shape.
Well,
no
we're
gonna
do
freak
out
because
it's
scalable
to
the
world,
that's
gonna,
last
them
soap
and
so
I
have
a
bigger
impact
and
we're
seeing
that
we're
seeing
the
fruits
of
that
like
with
free,
CAD
and
other
things,
I'm
just
really
getting
that
process
down
right
so
that
we
can
scale
yeah
thanks
for
coming
back
to
it.
B
A
B
A
A
B
B
Let's
see,
let's
see
where
it
goes
so:
okay,
well,
Matt!
Thank
you
so
much
for
the
time
and
we'll
continue,
yup
and
log
on
your
log.
The
thing
is:
make
sure
you
log
anything
that
you
do
on
your
log,
so
we
can
follow
what
you're
doing
and
that
whenever
we
invite
any
other
people
will
in
will
show
me
your
log
shown
this
video
I'll
post
this
video
on
your
log
as
well
as
well
as
on
YouTube,
so
yeah.
Definitely
please
to
keep
logging
mm-hmm.