December 18, 2025 • 43 mins
Please take our survey https://tinyurl.com/wyplistenersurvey about the show! Tell us what you want to hear, and what would make it better.
Tim Bucher is a farmer, and the founder and CEO of Agtonomy. Tim's problem is this: How do you build an autonomous tractor that can work for specialty crops like grapes, olives, apples, and almonds?
On today’s show, Tim explains what makes certain farming processes so difficult to automate, and how autonomy may soon change how we eat.
See omnystudio.com/listener for privacy information.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:15):
Pushkin. Just a quick note, today's episode is the last
one of twenty twenty five and for twenty twenty six.
We want to make the show better. We want to
make a show that you like more. To that end,
it would be extremely helpful to us and hopefully ultimately
also helpful to you if you could take a brief
(00:37):
survey that we have put together online. There's a link
to the survey in the show notes, and I'll give
you the url. It's TinyURL dot com slash wyp listener
survey again TinyURL dot com slash wyp like what's your
problem listener survey, or just click on the link in
(00:59):
the show notes. Thank you very much for listening to
the show, and thank you in advance for helping us
make the show better. Tim Booker grew up on his
fan family's dairy farm in rural northern California.
Speaker 2 (01:13):
You know how when you're a kid, you always wait
for summer vacation from school. My brother and I we
actually we hated summer vacation because that meant, you know,
eighty hours a week of work out in the open field.
So you know, our friends would always say, oh, I
can't wait until school's out. And we would go no,
I want to stay in school.
Speaker 1 (01:34):
Tim went off to college, started out studying agriculture, but
switched to computer science, and after he graduated he founded
a string of companies. One of them he sold to Microsoft,
now that he sold to Apple. But all along he
kept farming on the side, and eventually his farming life
and his tech life came together. As just kind of
(01:55):
a weekend side project on the farm he and a
few friends built on autonomous tractor. That weekend side project
has now turned into a company powering autonomous tractors across
the western US and part of Australia. I'm Jacob Goldstein
(02:17):
and this is What's Your Problem, the show where I
talk to people who are trying to make technological progress.
My guest today is Tim Booker. He's a longtime farmer
and the founder and CEO of Agtonomy. Tim's problem is this,
how do you bring autonomous vehicles to specialty crops, crops
like grapes and olives and almonds and apples. There's already
(02:41):
lots of autonomous equipment for row crops like corn and
wheat and soybeans, but as you'll hear, specialty crops present
a particularly tricky set of challenges. Tim and I talked
about the big picture about what autonomy will mean for
farming and for food. But to start, I asked him
about how he got from farming to computer science in
(03:03):
the first place. So you go off to college study
agricate culture Davis as a California farm kid does.
Speaker 2 (03:12):
Right.
Speaker 1 (03:13):
And did I hear you say in another interview that
you lost a bet and as a result of losing
the bet, had to take a computer science class.
Speaker 2 (03:23):
Yes, you heard right.
Speaker 1 (03:24):
What was the bet? Let's just say it was.
Speaker 2 (03:27):
At a fraternity party, so it might have involved you know,
it might have involved beer. I don't know.
Speaker 1 (03:32):
I like that it's a drunken fraternity party that leads
you into a career as a computer scientist. Like that
is an interesting combination.
Speaker 2 (03:40):
Yeah, it was the It was the greatest bet I
ever lost. But what was what was interesting? I all
my friends growing up, their last names were Sagatio, of
a Piano, Bach, Galupi, Gallo. These are incredible grape growing
in why making families back in the seventies, even and
I would always go over to their places to, you know,
(04:01):
help them, because that's what playdates were back then. Is
you basically went over to your friend's farm and worked
with them, and I would learn how to prune grapes
and even make wine. And I was fascinated by that industry,
like incredibly fascinated by the growing of grapes. So, you know,
I went to UC Davis for agriculture, and yes, I
lost a bet and took a class in a five
(04:24):
hundred person auditorium. Now keep in mind, I didn't grow
up with any technology. I mean like tractors. That was
about it. But the professor started talking and I'm in
the last row, and the professor started talking, and everything
he was saying I understood, and it just kind.
Speaker 1 (04:44):
Of hit me.
Speaker 2 (04:45):
And by the end of that quarter, you see Davis
on the quarter system. I was upfront teaching the class,
and so I knew something happened, like I found a passion,
like the passion I had for agriculture. And so I
had a decision to make, is do do I go
the agriculture route or do I go this computer science
(05:06):
route or this you know, high tech route. And I said,
you know what, I'm going to do both. I then
graduated from Davis and I went on to Stanford and
grad school and then started bouncing between Silicon Valley and
Sonoma County because when I was sixteen, I actually bought
my first vineyard, a small, little two acres, and I
(05:28):
loved farming that and that's kind of what got me to,
you know, get my own larger place, as I did
through the decades, which is now called Tratory Farms Tratory
which means tractor and Italian because I wanted to name
it after my childhood passion. And that's where that's where
(05:49):
kind of that whole thing started. The duality or the
dual path of agriculture and high tech.
Speaker 1 (05:57):
You have these sort of parallel lives going for a while, Right,
you're working for Steve Jobs, you're starting companies. You've got
a farm up in Sonoma. A little plane you're flying
back and forth, right, And so tell me about the
farm part of your life, Like that's you know, we're
(06:17):
going to bring back the technology in a minute. But
you're grown up. You have a real farm, Like what's
your farm?
Speaker 2 (06:25):
Well, it started as a small little vineyard, only two acres,
and then through the years I would do what's called
a ten thirty one tax free exchange and get a
bigger piece of land and then, you know, I don't know,
about twenty five years ago, I got a much bigger
piece of land.
Speaker 1 (06:43):
Yeah two acres, by the way, two acres isn't even
a farm, right, two acres is like a yard.
Speaker 2 (06:47):
Yeah, pretty much a yard. Good good call. But it
was French colmbard grapes and zinfandel grapes. So so, you know,
fast forward two years later, as I would kind of
get bigger and bigger places, I got this one place
which was amazing beautiful hillsides. I had a vision for it,
but it was just a forest, and I start clearing
(07:09):
it for you know, the intention of growing more grapes,
and I encountered these stumps, you know, equidistance throughout the land,
and I was like, what is this? And then finally
I uncovered some old olive trees from like the eighteen hundreds,
and so did some research and realized, well, this was
part of a big you know, Spanish land grant growing
(07:31):
olives back in the eighteen hundreds. So I decided to
bring that history back and that's what got me into olives.
So I grow olives for olive oil and grapes for wine,
and I have a winery operation public tasting room, so
kind of bouncing between two worlds for decades.
Speaker 1 (07:50):
So before you started etonomy, tell me about the way
you brought technology to your farm, Like, what did you automate?
Speaker 2 (08:00):
Yeah? I was unique in that I was a farmer,
but I was also an engineer. I had both, you know,
electrical engineering background as well as computer science background, so
I could do things like create you know, automated water
recycling plants that you know, had the flow meters everywhere
(08:21):
it knew, you know, how to inject different microbes to
process the waste water and then inject it back into
the irrigation automatically. I had, you know, obviously all kinds
of basic things like cameras and whatnot. But even in
the fermentation of grapes, the tanks are all computer controlled.
(08:44):
The computer controls whether to use you know, active refrigeration,
which costs energy, or if the outside air is cooler
than the inside, it automatically switches over and brings that
kind of air in. So you know, everything from fermentation
processes to water recycling to irrigation is radically automated. So
(09:04):
I had automated darn near everything I could, and I
didn't know I was actually doing ag tech. When I
was doing this, ag tech wasn't a word that existed
back then.
Speaker 1 (09:16):
You were just like puttering around on the weekends.
Speaker 2 (09:19):
I was puttering around on the weekends or night times whatever.
I just knew my farm needed technology to automate things,
because keep in mind, I was also bouncing between Sonoma
County and Silicon Valley, and so I needed technology to
also be able to monitor things. Right, But there was
one area I hadn't automated, and that was the actual
(09:41):
work in the orchards and the vineyards. And I was
sitting there going, hell, you know, I automated this. I
automated that, Hey, maybe I should try my hand at
automating the field work. There's so much money being invested
in autonomous passenger vehicles, and there's so much technology that's
evolved there. I wonder if I could just apply that
to this space and see if I can create an
(10:03):
autonomous vehicle that can do these kinds of repetitive tasks
like mowing, like spraying, you know, like underfining cultivation, like seating,
like disking, things of that nature where you're just going
back and forth up and down the rows.
Speaker 1 (10:18):
Right, Just to be clear, there had been a fair
bit of automation, right, But on the sort of big
egg row crop like corn and wheat side, as I
understand it, but not so much on the permanent crop
kind of fruit and nut side. Is that right? Was
that the sort of state of.
Speaker 2 (10:34):
Play exactly, Jacob So, in row crops or in broad
acre farming, you know, the big million dollar tractors or
half million dollar tractors, there has been automation there for decades,
things called auto steer. You know, you're on these flat areas,
very big fields. You can see the sky. GPS had
(10:54):
evolved so it would be somewhat accurate, you know, a
meter or whatever it might be, and so you could
just control the vehicles based on that. When you think
about permanent crops, which is a huge percentage value wise
of our foods supply, the permanent crops, you're talking about
trees and vines that are very very expensive. They take
(11:17):
many years to develop, and so if you hit them,
it's kind of a big deal. It's kind of expensive
and you have to get really close to them. So
you're not talking about being an open field where you
can be you know, plus or minus half a meter.
You're talking about being like one inch away from you know,
the trunks of these trees, and if you accidentally move
(11:38):
you know, abruptly, or you know, don't steer correctly, you're
going to hit it. So you can't rely on kind
of the things that you can rely on in broad
acre farming and the rowe crop farming. So there needed
to be more advanced technology. And thankfully, due to all
the attention and investments that were being made in you know,
(11:58):
the autonomous passenger vehicle world, that technology started to become available.
You know, call it maybe less than a decade ago,
but when I said, you know, kind of seven years ago,
like hey, you know, my expenses are continuing to go
up at tratory. I've automated DAR near everything, but I
haven't automated the field work. Could we take some of
(12:21):
that technology and could we create a solution, an autonomous
solution that could do these repetitive tasks that you need
to do in permanent crops. And so with some friends
built a prototype five years ago and the prototype worked
really well. We literally built an all electric autonomous tractor.
Speaker 1 (12:44):
What did it look like and what did it do?
Speaker 2 (12:46):
Didn't look like a traditional tractor. We took out the
diesel engine. We put in electric motors and batteries, and
we put on cameras and we had you know, an
Nvidia compute system in it, and we started to create
this sensor based vehicle. It had eight wheels, it articulated,
(13:07):
and it could do really steep st and one of
the reason why we did that is because that tratory
the slopes are very very steep. By the way, one
of the main inspirations for doing this autonomous vehicle came
from a documentary I watched. I think it was a
National geographic. It was called Mission to Mars, The Story
of Spirit and Opportunity. And these were Mars rovers in
(13:28):
the nineties that NASA shot up and we're hoping would
operate for you know, fifteen days or you know, a
couple of months, and they ended up lasting almost fifteen years.
And I was watching this documentary before I actually built
this prototype, and it reminded me that, hey, there are
environments where you can take your time in making decisions.
(13:53):
So they had animations of how the rover would approach
a rock and it would just stop and then it
would radio JPL, you know, take some time to radio JPL.
They would take a week and write some new code.
They would then download it, and then the vehicle, the
rover would move to the left that whenever it encountered
a rock, it would move to the laughter to the right.
And I said to myself, my god, you know, they
(14:15):
had autonomous vehicles operating on Mars in the nineties and
they could do it because there wasn't a lot of
traffic on Mars, just like agriculture. And that's when my
head really exploded and I said, wait, there are these
industrial markets that need automation, and you don't necessarily have
to immediately, you know, be responsive to the vehicle if
(14:39):
it encounters something it can't figure out, like oh, there's
a tree in front of.
Speaker 1 (14:43):
Me, unlike on a road in traffic. Correct. Correct, So
when you build this prototype that you don't even know
is a prototype as you're telling it, right, this thing
that you're doing with your friends, I mean it sounds
like a lot of the sort of tex stack is
basically commodified by that point. I mean, is it a
sort of you're at this moment where you can be like, oh, well,
(15:05):
buy these sensors and we'll start with this piece of software,
and like you can almost build an autonomous vehicle from
off the shelf parts. Is that sort of the starting point.
Speaker 2 (15:14):
From a hardware perspective. Absolutely, from a software perspective, no,
at that time, so no, we were actually building a
perception stack, you know that could you know do all
the Well, first you had to localize, so you have
to figure out where in the world do you are,
And so we did need to see GPS at least once,
but then the computer vision technologies and the perception stack
(15:37):
could then align to the crop itself. One of the
big differences that in approach that we took that obviously
the row crops, the broad acre farms don't need to
do is we actually said, hey, what if we exploit
the structure the crop itself and utilize that structure to
navigate safely through it and farm precisely around it.
Speaker 1 (15:59):
Meaning like know what a tree is and know what
your relationship should be to a tree. Is that what
that means? Correct?
Speaker 2 (16:06):
And in the early days we said, hey, there's one
common characteristic and permanent crops. They all have a trunk,
So what if we could be the best detectors of
trunks in the world and utilize that for alignment? And
as I said, also you know, working around that plant.
Speaker 1 (16:25):
So that's the key sort of training piece, Like you
train really hard on trunks.
Speaker 2 (16:30):
That's right, that's right. We initially labeled trunks. We would
drive through, we would get you know, tons of photos
and we would manually label around the trunk area. And
then we would take those images and we would tweak
them and create synthetic data and we would label those
and once you get enough of a data set, you
(16:51):
can start to do some additional machine learning with that, right.
And what was interesting was, you know, we got so
good at detecting trunks. We even called it trunk vision.
We trademark trunk vision. Not that you know, we do
a lot more than trunk vision today, but that's how
that's how we got it to, you know, run through
the orchards and run through the vineyards. And what was
(17:12):
really interesting is we initially trained on grapes, on grapevines
and the rootstock right the trunk. And one day my CTOs,
you know, he said, hey, let's run it in the olives.
And I'm like, oh, but you know, you're sure it's
going to work. He's like, I don't know if it's
going to work. I don't know how well the training
will translate. And so we set it off in the
(17:34):
Olives and it just worked. Now, it wasn't perfect, but
it made us realize, Wow, there's a lot of leverage
between crops here in these permanent crop areas. Generalizes, it
generalizes really well. And of course he's continued to evolve that,
and you know, we now do all kinds of additional
training to make sure in a new crop type we
(17:56):
are very very accurate in exploiting the structure of the crop.
Speaker 1 (18:01):
Besides driving through trees without hitting them, which I recognize
as non trivial, what else did the tractor do? Did
it do work?
Speaker 2 (18:09):
Yes, it did work. It mowed initially. The very very
very first job we did was mowing because we have
cover crop. Sometimes the cover crop gets like six feet
tall and you chop it up with a flail mower
and then it composts in between the trees or the
vines and and even that, like you think mowing, Oh yeah,
(18:30):
you just set it at four miles an hour and
you let it go. You cannot do that, right, because
think about what a human does. The human's on a vehicle.
It's really funny. I asked my engineers you know, well,
did you ever use a push mower when you were
a kid, And they're like, what's a push mower? Anyway?
Speaker 1 (18:45):
Separate topic, but pushmore is really hard. It's harder than
you think that was th pushover means the more without
an engine. It doesn't just mean a gas mower that
you put.
Speaker 2 (18:56):
Right right right and then and then you know, we
got motorized mowers for lawns, right that you that you
know actually had wheels and had an engine on them.
But even with those, if you would go through the
and if it would bog down, as the person operating
that lawn mower, you would slow it down in order
(19:18):
for the blades to catch up, right. And so if
you think about it, you can't just set a tractor
going to mow. You need to get that feedback to
understand if it's bogging down, and therefore you need to
slow down or increase the RPM in order to do
the job properly.
Speaker 1 (19:37):
Right.
Speaker 2 (19:38):
So that's something we learned immediately, But we already thought
about that because we were all farmers. We had farmer
DNA in us, and we knew that's how a human operated.
So we just we started to think, Okay, how does
a human do this? And what are the things we're
gonna need feedback on like a human has in order
to get the job done correctly. So we started with
(19:59):
mowing and we became really really good at it.
Speaker 1 (20:02):
So you have your prototype and you decide to start
a company. What do you have to do to turn
your prototype into you know, something you can sell? Like, yeah,
what's the what's the leap?
Speaker 2 (20:18):
Yeah, this was a big leap. You know. My mechanical
engineering co founders was like, you know, whipping up all
kinds of cool ideas. And then I thought about it
as a farmer, and I said, wait a minute, as
a grower, would I trust a startup company, you know,
for a new tractor? Think about it, right, I'm a
(20:41):
farm I rely upon equipment to get my job done,
and that job is vital for the livelihood of the business,
for the livelihood of the family if it's a family farm.
So why would I trust a startup instead? What I
need as a grower is I need trusted brands. I
(21:04):
need their dealer networks. I need the parts, I need
the service. This this equipment needs to be operating all
the times that I need it for the sustainability of
my business. And so I said to my colleagues, I said,
you know what, I don't think we should build a tractor.
I think that would be rather stupid. And I explained why,
and they said, oh, that makes sense, and they said, well,
(21:27):
how can we work with existing original equipment manufacturers? These
incredible equipment manufacturers who've been at it for over one
hundred years and they build incredible vehicles that really work
well in these harsh environments. And so you know, at
first investors that I was talking about, Hey, we're going
(21:47):
to do an OEM model and you know, original equipment manufacturer,
we're going to help accelerate their digital transformation because we
know where the puck is going, right. We know that
this industry too, just like other industries, needs to go
through a digital transformation. Every day there's less and less
skilled operators who go into agriculture. So the equipment manufacturer know,
(22:09):
you know, they see the writing on the wall. They're
not going to sell as much equipment because like there's
less people to operate them. So they know automation is
really important, and they know how to build incredible iron,
but do they know how to build an AI factory?
And that's what I said, you know we would focus
on is that we would build software to make their
(22:32):
incredible equipment incredibly smart and provide the solutions that make
them autonomous and give growers ultimately new technologies that can
help them save money and get an incredible ROI. It's
almost like building equipment that comes with a skilled operator
built in that you can turn on and off any
time you want.
Speaker 1 (22:53):
As a bonus, you don't have to have the crazy
capital outlay of building a tractor factory, right like, I mean,
that's business. It seems way better.
Speaker 2 (23:03):
It does, but there's one catch, and investors early on
would say, wow, I don't think you can get an OEM.
Speaker 1 (23:11):
And we did an OEM, an original equipment manufacturer attractor.
Speaker 2 (23:16):
Attractor company, and we did an incredible company called Bobcat.
Douce on.
Speaker 1 (23:21):
Bobcat.
Speaker 2 (23:22):
They make excavators large and small, They make skid steers,
they make trackloaders, they make tractors like normal looking tractors, right,
and they they make forklifts. They make all kinds of things.
So you know, really, what we're doing is exactly what
I said earlier, which is to accelerate the digital transformation
(23:46):
of these companies because they know they have to transform,
they know they have to bring autonomy into their equipment
and and and create you know, actonomy enabled vehicles. Right,
So we're not like creating tractors. We're just helping them
with the with the brains of the operation.
Speaker 1 (24:09):
We'll be back in just a minute. What's out there now?
Is your software out in the world now? Dive in
tractors around orchards and what are they doing and where
are they and how many of them are there?
Speaker 2 (24:29):
Yeah, we have tractors throughout the Western United States, and
I'm pleased to say that we also have vehicles in Australia.
So we're super excited about the capabilities we've enabled. In
the state of Washington, for example, where apples are number
one in the world, we're doing a lot of spraying
and a lot of mowing. And it's really it's very
(24:52):
impactful because you're not just having one vehicle, but our
customers have many vehicles. They have a fleet of vehicles,
and they're able to operate those with one of their
employees that they've upskilled to be able to supervise a fleet.
And so it's kind of neat to see, you know,
a whole fleet of vehicles leaving the main shop, driving
(25:15):
on the dirt roads out to the different areas where
they need to do work, do the work, and then
all come back right, So it's a very seamless operation
for them. Our software is also on mobile applications and
that's how the main user experience exists. So think of
you having the tablet or even just your smartphone, and
(25:38):
the site is mapped out, usually with aerial imagery, and
you can just point and say, I want to go
and mow this, you know, this block of the orchard,
and we basically can get close to where the crop is.
And then once we're close to the crop, that perception
stack I mentioned earlier about detecting you know this, or
(26:00):
exploiting the structure of the crop itself. Take trunk trunk
vision and it snaps the vehicle to grid, if you will. Right.
But now, once you have trunk vision and you see
in front of you, and you can see the terrain
to a very high degree of accuracy, you can take
mechanical weaters, for example, and manipulate underneath these trees, you know,
(26:25):
these robotic weaters, and navigate while you're doing that very precisely,
all under computer control. Whereas when a human does it,
you're you know, driving a vehicle six miles an hour
looking forward, you have your hand on six actuators on
the you know, the weater and behind you, and you're
trying to make it all work, and inevitably what happens
(26:46):
is you don't get the weeds so the efficacy is
not high, or you hit the tree and take it out,
and you know that again costs a lot of money.
So that's why herbicides have been used so much because
it's really easy. It's called strip spray and you just
spray it and you're done. But that's that's kind of
a thing that we didn't realize. You know, this kind
(27:07):
of autonomous technology would really open up is much more
offerings for growers so that they don't have to use
very expensive ag inputs or herbicides that they've been doing
in the past.
Speaker 1 (27:21):
All this work that you're describing, is it happening in
a like commercial way now, Like, are you out in
the world getting paid for this? Yes? And what's that
business model?
Speaker 2 (27:31):
On that note, Yeah, so we've been operating in paid
pilots in twenty twenty three and twenty twenty four. We
never intended to be paid for pilots. What happened was
we'd have these demo days. We'd invite growers and they
would say, I want to buy one. I said, well,
they're not for sale. Well I want to rent one.
(27:52):
I'm like, well maybe we can do that. Would you
be interested in doing a pilot? Yes, and I'll pay you.
And we're like okay. So in those original prototypes that
we built, we actually rented them and it was kind
of cool because it was a very small company. We
didn't actually think we were going to recognize revenue until
many years in. And I think our first year we did,
(28:13):
you know, it was actually close to a million dollars.
But then the second year we did several million dollars
and that was in twenty twenty four. And this year
we went big time and we have units operating from Washington, Oregon,
California and now Australia. And so our business model right
now is we work with the manufacturers. They manufacture everything,
(28:36):
they procure all the parts, they bring it into their
manufacturing facilities and they create these very high quality machines
that are factory fit. Right, A lot of people think, oh,
you do retrofit kits on tractors. We do not do that.
We believe it's important to work with the engineers of
these OEMs to make the equipment much more reliable, much safer,
(28:57):
and lower cost, much lower cost when you integrate it in.
Speaker 1 (29:01):
And then when the farmer buys the machine with your
software in it, with your autonomy package in it, like
is it are they paying you a subscription? Is the
price just embedded in the machine and you get some
of the money.
Speaker 2 (29:14):
So the answer is yes to all of those because
in some cases the software fees will be embedded in
the machine. In some cases you will pay a monthly
fee or an annual fee or a one time fee.
Speaker 1 (29:28):
Right.
Speaker 2 (29:29):
Think of serious XM. Serious XM creates technology, it gets
embedded by the OEMs and their manufacturing facilities.
Speaker 1 (29:38):
Satellite radio just to satellite radio, yeah yeah.
Speaker 2 (29:42):
And goes to the dealers, and then the dealers basically
sell to the end customer, and there's a free trial
that comes and after the free trial, Serious XM retains
those customers and charges a monthly fee.
Speaker 1 (29:57):
That's a good metaphor satellite radio, but for autonomous tractors,
for high value.
Speaker 2 (30:02):
Crops exactly, there's another twist. I'll tell you that. By
the way, the chairman of our company, the chairman of
the board is Jim Meyer. He's the former CEO of
Serious ExM.
Speaker 1 (30:12):
Ah okay, yeah.
Speaker 2 (30:14):
And by the way, he's brilliant at business and he
really helps shape this kind of business model.
Speaker 1 (30:19):
What's the technical thing you haven't figured out yet.
Speaker 2 (30:23):
I'll give you a small example, but it's kind of important.
So three point turns. If you think about permanent crops,
you're going down these rows and there's you know, a
person has a ranch it's called one thousand acres, and
there might be a fence around the entire thousand acres
right to keep deer out or whatever it might be.
And they try to maximize the land they have, so
(30:45):
they plant the crop pretty close to the edge of
the fence, let's say. And so when you turn, you
know in the headlands right, the headland turn, you might
not have a lot.
Speaker 1 (30:54):
Of space, so you can't just like make a kind
of U turn at the end of the row and
go down the next.
Speaker 2 (30:59):
Correct correct correct. So you know, right now we can
do about eighty percent of farmlands and permanent crops. But
imagine if we could do point turns, we can achieve
one hundred percent. But but there's a twist, right because
sometimes you not you don't have an implement, that's the
device you attached to the tractor that does the work.
(31:20):
So you have implements you can attach for mowing, for spraying,
for weeding, for disking, for all kinds of things. Some
of those implements are actually trailers.
Speaker 1 (31:30):
H huh.
Speaker 2 (31:30):
Okay, So imagine you know, being on a site slope
with rocky soil and everything, and you have a trailer
and you have to do a three point turn to
get into the next row.
Speaker 1 (31:41):
That's a high skill that's a high skill moment, right,
that's hi know what they're doing exactly exactly.
Speaker 2 (31:47):
And you know, ectonomy enabled vehicles have lots of cameras.
They have you know, minimum of eight cameras including rear
facing cameras and side facing cameras, and so you have
to really now start to you know, do very sophisticated
modeling of the you know, physics of that tractor as
well as that implement or trailer, which we which we
(32:08):
do today as we go forward, but now we have
to do it going backwards. So we haven't mastered that
yet but we're getting close, and I'm super excited about that,
both from a technological achievement perspective as well as a
business need.
Speaker 1 (32:25):
It's a nice when they go together. Yeah, so let's
so madch for a minute, like I'm you know, there
are other people working on other agricultural technology projects, other
ag tech projects, even other autonomous projects. Like when you
zoom out and think about farming and technology more generally
(32:45):
in the medium term, you know whatever that means, five years,
ten years, like, how's the world going to change?
Speaker 2 (32:55):
Well, so, first of all, there are a lot of
others working on similar problems. We all have a slight twist,
either in business model or in technology approach. I think
that in the future is all going to be autonomous capable,
and we see the writing on the wall right when
(33:16):
it comes to labor challenges that we're having, we have
a severe labor gap. If I were to ask you
how many people do you know who you know whose
kids go into agriculture, the answer is probably going to
be not that many. I have three children, Jacob, and
none of them are taking over the farm. And then
you know, of course we have things like you know,
(33:38):
some immigration policy changes that have occurred that are only
exacerbating the problem.
Speaker 1 (33:42):
Even in the absence of a labor shortage, that some margin,
automation wins, right, Like that's what happened with the tractor
and like, whatever's right. So yes, I'm fully prepared to believe.
And yes, I can see how the current politics might
be accelerating the shift to automation. But yes, so okay,
automation is going to win. I stipulated, like what's that
(34:02):
going to mean? What's it going to look like? You know,
tell me something about the future based on that fact.
Speaker 2 (34:07):
Think about ro crops, right, I believe row crops are
going to radically change. So here is one big inflection
point that I think is going to happen. Right now,
we have these large, large tractors in row crops. It
cost a million dollars. Fully autonomous ones cost well over that.
And why are there big tractors. It's because these these farms,
(34:29):
you know, want as many acres done per hour by
one person. But now if automation comes in, you don't
necessarily need these large vehicles that have huge ground compassion.
Speaker 1 (34:40):
This is your trillion dollar play. Now you're going to
disrupt John Deere. Is that where this is.
Speaker 2 (34:45):
Going yes, So I mean, but think about it, right, if.
Speaker 1 (34:50):
You're the only reason you need to make the vehicle
as big as possible is if you have to have
a dude on every one of them. That's what you're saying.
It's not necessarily it could be like like the way
it's like a drone swarm, you could do a drone swarm,
but for tractors.
Speaker 2 (35:02):
Broad acre swarming exactly right, Yeah, And that's there's some
additional benefits, right, So instead of spending and a million dollars,
you can do more work with five tractors that cost
less than that with their implements as well, and now
you have some additional benefits, so you actually get more
work done faster, and if one of the tractors breaks down,
(35:24):
you have redundancy. Right, If your big million dollar tractor
breaks down, you got issues. It's like thousands of dollars
a minute that you know that it's not running. So
I think there's a lot of opportunity in agriculture that
we hadn't thought about once automation is achieved. I mean,
I describe the herbicide challenge in permanent crops. I believe
(35:46):
that is completely can be completely eliminated with automation, and
you know computer controls, computer vision, like you.
Speaker 1 (35:55):
Just pick the weeds instead. If you have essentially very
cheap robots that are very dexterous, you just pick the
weeds instead of spraying them.
Speaker 2 (36:02):
Absolutely, it's already happening in row crops, you know. You
look at carbon robotics, they're using lasers to kill weeds,
you know, in lettuce fields and other kinds of crops
that are called specialty crops. That's huge, right because before
humans would would go in there, or you'd spray and
use bag inputs or herbicides. So I think there's a
(36:23):
huge there is a huge inflection point. But I just
I want to make it clear to your listeners that
just the core base of you know, being able to
produce enough food for the population, we need automation. Four
there are going to be these other benefits, some of
which we don't even know yet. So I'm such a
big believer in automation. It's why I brought my two
(36:44):
worlds together, and it's why I've really dedicated my life
to this mission. And I just I've never been as
excited as I as I am now in terms of
the impact that this company can have our share of impact,
but all of us together, there's so much opportunity here.
And you know, I think the incumbents are the ones
(37:04):
that the incumbent equipment manufacturers are the ones that are
really going to make it happen in terms of distribution
and support and you know, getting it out there to
the world. So couldn't be more excited.
Speaker 1 (37:20):
We'll be back in a minute with the lightning round.
Let's finish with the lightning round. I heard you say
on another interview that you are not a big fan
of IPOs and that you generally prefer acquisitions. Why what
(37:42):
do you got against IPOs.
Speaker 2 (37:45):
Well, that's from the old days. I've went through a
couple of IPOs and just about killed me. So a
lot of people think, you know, IPOs are glamorous and everything,
but you know, the road shows are pretty intense and
the pressures are really intense, and you know, I'm not
just a fan of acquisitions. I mean, I think you're
seeing a lot of differences in the private market than
(38:06):
you have in the past. Companies like open Ai who
are raising billions of dollars in the private market and
not having to go public.
Speaker 1 (38:15):
You could just stay private forever. You could just be
a company that makes more money than it's spends. You
could do that for as long as.
Speaker 2 (38:21):
You want, exactly exactly. So I think the IPO, you know,
I mean, just look at the last few years of IPOs.
There just haven't been that many. And so there are
other ways to build incredibly valuable companies. I get asked
all the time, what's your exit strategy? And I always
have the same answer. I only focus on building value
(38:43):
opportunities come, and I think entrepreneurs who build companies for
an exit aren't building the company for the right reasons
and aren't building really valuable companies either.
Speaker 1 (38:56):
What's one thing that Steve Jobs told you that stuck
with you.
Speaker 2 (39:02):
I had the pleasure of working for some incredible entrepreneurs,
Steve Jobs being one of them. And what I did
learn is that these incredible entrepreneurs, they each had like
a singular mantra. And for Steve, it was all about design.
It's all about design, and there were things he would
push for that we'd sit there and scratch our heads
(39:22):
and go, wait, that's going to add like one hundred
dollars cost to this, and and we would do it
and the product would be successful, and we're like, Okay,
he was right. You know. I learned from from Bill
Gates that it's all about software, and you know what,
he was right. I learned from Michael Dell that it's
(39:43):
all about costs, and guess what, he was right. So
I think, you know, the every entrepreneur that makes serious impact,
you know, has these kind of core mantras. So what's
yours I viewed as a combination. But there is one
thing that I will tell you very very clearly in
your listeners, and I tell my teammates probably every day,
(40:07):
if not every week. You know, it's all about show me,
don't tell me. And that particularly applies to these industrial markets.
The best way to sell industrial equipment, and this is
what dealers do is, you know, hear Susie farmer take
(40:28):
this piece of equipment and use it for a day,
and she'll take that piece of equipment, use it on
her ranch, on her farm, whatever, and inevitably she will
buy it because she gets to try, you know before.
So I think I think the show me mentality is
really really important in this era, and in ag tech
in particular, because there've been some companies that have been
(40:49):
developed by pure tech people and pure tech people. You know,
telling farmers that you know, we can farm better than
you is not a good recipe for success, right, And
so by focusing on show me, it's really what resonates
with our customers, it's really what our customers need, and
(41:11):
it's just how business is done in these industrial worlds.
So that's my mantra, show me, don't tell me.
Speaker 1 (41:21):
What's one ridiculous word somebody once used to describe the
wine that you grow? Hm hmm.
Speaker 2 (41:33):
I didn't even know what this. I still don't know
what this word means. But ethereal, I'm like, like, what
does what does that mean?
Speaker 1 (41:42):
It doesn't stuck around, It's like I don't even know
I drank.
Speaker 2 (41:45):
Yet, Like yeah, I don't know, right, So you get
all kinds of feedback.
Speaker 1 (41:50):
Is that a fancy way of saying easy drinking sounds
like easy drinking? Probably?
Speaker 2 (41:54):
Yeah? Yeah, yeah, But you get all kinds of great feedback.
And I mean that's the cool thing about sort of
the show me mentality is when you develop products that
end users you know, consume or use, and you see
their reaction and you know, there you you please them,
You give them something of value That's that's what gets
(42:16):
me motivated all the time. And that's why you know,
this space is so exciting for me, because that that
impact has felt like really really quickly. When when you
do have a solutions at work, or when you do
make great wine, or when you do make great olive oil.
Speaker 1 (42:32):
Yeah, I don't think autonomous tractors as ethereal, I please don't.
Tim Booker is a farmer and the founder and CEO
of Actonomy. Please email us at problem at pushkin dot fm.
(42:52):
We are always looking for new guests for the show.
Today's show was produced by Trinamanino and Gabriel Hunter Chang,
who was edited by Alexander Garrettson and engineered by Sarah Brugueret.
I'm Jacob Goldstein. We'll be off for the next few
weeks for the holidays. I want to thank you very
much for listening to the show this year. It really
(43:13):
means a tremendous amount to all of us. I hope
you have a great holiday, happy New Year, and we
will be back with more episodes of What's Your Problem
in twenty twenty six.
Pushkin. Just a quick note, today's episode is the last
one of twenty twenty five and for twenty twenty six.
We want to make the show better. We want to
make a show that you like more. To that end,
it would be extremely helpful to us and hopefully ultimately
also helpful to you if you could take a brief
(00:37):
survey that we have put together online. There's a link
to the survey in the show notes, and I'll give
you the url. It's TinyURL dot com slash wyp listener
survey again TinyURL dot com slash wyp like what's your
problem listener survey, or just click on the link in
(00:59):
the show notes. Thank you very much for listening to
the show, and thank you in advance for helping us
make the show better. Tim Booker grew up on his
fan family's dairy farm in rural northern California.
Speaker 2 (01:13):
You know how when you're a kid, you always wait
for summer vacation from school. My brother and I we
actually we hated summer vacation because that meant, you know,
eighty hours a week of work out in the open field.
So you know, our friends would always say, oh, I
can't wait until school's out. And we would go no,
I want to stay in school.
Speaker 1 (01:34):
Tim went off to college, started out studying agriculture, but
switched to computer science, and after he graduated he founded
a string of companies. One of them he sold to Microsoft,
now that he sold to Apple. But all along he
kept farming on the side, and eventually his farming life
and his tech life came together. As just kind of
(01:55):
a weekend side project on the farm he and a
few friends built on autonomous tractor. That weekend side project
has now turned into a company powering autonomous tractors across
the western US and part of Australia. I'm Jacob Goldstein
(02:17):
and this is What's Your Problem, the show where I
talk to people who are trying to make technological progress.
My guest today is Tim Booker. He's a longtime farmer
and the founder and CEO of Agtonomy. Tim's problem is this,
how do you bring autonomous vehicles to specialty crops, crops
like grapes and olives and almonds and apples. There's already
(02:41):
lots of autonomous equipment for row crops like corn and
wheat and soybeans, but as you'll hear, specialty crops present
a particularly tricky set of challenges. Tim and I talked
about the big picture about what autonomy will mean for
farming and for food. But to start, I asked him
about how he got from farming to computer science in
(03:03):
the first place. So you go off to college study
agricate culture Davis as a California farm kid does.
Speaker 2 (03:12):
Right.
Speaker 1 (03:13):
And did I hear you say in another interview that
you lost a bet and as a result of losing
the bet, had to take a computer science class.
Speaker 2 (03:23):
Yes, you heard right.
Speaker 1 (03:24):
What was the bet? Let's just say it was.
Speaker 2 (03:27):
At a fraternity party, so it might have involved you know,
it might have involved beer. I don't know.
Speaker 1 (03:32):
I like that it's a drunken fraternity party that leads
you into a career as a computer scientist. Like that
is an interesting combination.
Speaker 2 (03:40):
Yeah, it was the It was the greatest bet I
ever lost. But what was what was interesting? I all
my friends growing up, their last names were Sagatio, of
a Piano, Bach, Galupi, Gallo. These are incredible grape growing
in why making families back in the seventies, even and
I would always go over to their places to, you know,
(04:01):
help them, because that's what playdates were back then. Is
you basically went over to your friend's farm and worked
with them, and I would learn how to prune grapes
and even make wine. And I was fascinated by that industry,
like incredibly fascinated by the growing of grapes. So, you know,
I went to UC Davis for agriculture, and yes, I
lost a bet and took a class in a five
(04:24):
hundred person auditorium. Now keep in mind, I didn't grow
up with any technology. I mean like tractors. That was
about it. But the professor started talking and I'm in
the last row, and the professor started talking, and everything
he was saying I understood, and it just kind.
Speaker 1 (04:44):
Of hit me.
Speaker 2 (04:45):
And by the end of that quarter, you see Davis
on the quarter system. I was upfront teaching the class,
and so I knew something happened, like I found a passion,
like the passion I had for agriculture. And so I
had a decision to make, is do do I go
the agriculture route or do I go this computer science
(05:06):
route or this you know, high tech route. And I said,
you know what, I'm going to do both. I then
graduated from Davis and I went on to Stanford and
grad school and then started bouncing between Silicon Valley and
Sonoma County because when I was sixteen, I actually bought
my first vineyard, a small, little two acres, and I
(05:28):
loved farming that and that's kind of what got me to,
you know, get my own larger place, as I did
through the decades, which is now called Tratory Farms Tratory
which means tractor and Italian because I wanted to name
it after my childhood passion. And that's where that's where
(05:49):
kind of that whole thing started. The duality or the
dual path of agriculture and high tech.
Speaker 1 (05:57):
You have these sort of parallel lives going for a while, Right,
you're working for Steve Jobs, you're starting companies. You've got
a farm up in Sonoma. A little plane you're flying
back and forth, right, And so tell me about the
farm part of your life, Like that's you know, we're
(06:17):
going to bring back the technology in a minute. But
you're grown up. You have a real farm, Like what's
your farm?
Speaker 2 (06:25):
Well, it started as a small little vineyard, only two acres,
and then through the years I would do what's called
a ten thirty one tax free exchange and get a
bigger piece of land and then, you know, I don't know,
about twenty five years ago, I got a much bigger
piece of land.
Speaker 1 (06:43):
Yeah two acres, by the way, two acres isn't even
a farm, right, two acres is like a yard.
Speaker 2 (06:47):
Yeah, pretty much a yard. Good good call. But it
was French colmbard grapes and zinfandel grapes. So so, you know,
fast forward two years later, as I would kind of
get bigger and bigger places, I got this one place
which was amazing beautiful hillsides. I had a vision for it,
but it was just a forest, and I start clearing
(07:09):
it for you know, the intention of growing more grapes,
and I encountered these stumps, you know, equidistance throughout the land,
and I was like, what is this? And then finally
I uncovered some old olive trees from like the eighteen hundreds,
and so did some research and realized, well, this was
part of a big you know, Spanish land grant growing
(07:31):
olives back in the eighteen hundreds. So I decided to
bring that history back and that's what got me into olives.
So I grow olives for olive oil and grapes for wine,
and I have a winery operation public tasting room, so
kind of bouncing between two worlds for decades.
Speaker 1 (07:50):
So before you started etonomy, tell me about the way
you brought technology to your farm, Like, what did you automate?
Speaker 2 (08:00):
Yeah? I was unique in that I was a farmer,
but I was also an engineer. I had both, you know,
electrical engineering background as well as computer science background, so
I could do things like create you know, automated water
recycling plants that you know, had the flow meters everywhere
(08:21):
it knew, you know, how to inject different microbes to
process the waste water and then inject it back into
the irrigation automatically. I had, you know, obviously all kinds
of basic things like cameras and whatnot. But even in
the fermentation of grapes, the tanks are all computer controlled.
(08:44):
The computer controls whether to use you know, active refrigeration,
which costs energy, or if the outside air is cooler
than the inside, it automatically switches over and brings that
kind of air in. So you know, everything from fermentation
processes to water recycling to irrigation is radically automated. So
(09:04):
I had automated darn near everything I could, and I
didn't know I was actually doing ag tech. When I
was doing this, ag tech wasn't a word that existed
back then.
Speaker 1 (09:16):
You were just like puttering around on the weekends.
Speaker 2 (09:19):
I was puttering around on the weekends or night times whatever.
I just knew my farm needed technology to automate things,
because keep in mind, I was also bouncing between Sonoma
County and Silicon Valley, and so I needed technology to
also be able to monitor things. Right, But there was
one area I hadn't automated, and that was the actual
(09:41):
work in the orchards and the vineyards. And I was
sitting there going, hell, you know, I automated this. I
automated that, Hey, maybe I should try my hand at
automating the field work. There's so much money being invested
in autonomous passenger vehicles, and there's so much technology that's
evolved there. I wonder if I could just apply that
to this space and see if I can create an
(10:03):
autonomous vehicle that can do these kinds of repetitive tasks
like mowing, like spraying, you know, like underfining cultivation, like seating,
like disking, things of that nature where you're just going
back and forth up and down the rows.
Speaker 1 (10:18):
Right, Just to be clear, there had been a fair
bit of automation, right, But on the sort of big
egg row crop like corn and wheat side, as I
understand it, but not so much on the permanent crop
kind of fruit and nut side. Is that right? Was
that the sort of state of.
Speaker 2 (10:34):
Play exactly, Jacob So, in row crops or in broad
acre farming, you know, the big million dollar tractors or
half million dollar tractors, there has been automation there for decades,
things called auto steer. You know, you're on these flat areas,
very big fields. You can see the sky. GPS had
(10:54):
evolved so it would be somewhat accurate, you know, a
meter or whatever it might be, and so you could
just control the vehicles based on that. When you think
about permanent crops, which is a huge percentage value wise
of our foods supply, the permanent crops, you're talking about
trees and vines that are very very expensive. They take
(11:17):
many years to develop, and so if you hit them,
it's kind of a big deal. It's kind of expensive
and you have to get really close to them. So
you're not talking about being an open field where you
can be you know, plus or minus half a meter.
You're talking about being like one inch away from you know,
the trunks of these trees, and if you accidentally move
(11:38):
you know, abruptly, or you know, don't steer correctly, you're
going to hit it. So you can't rely on kind
of the things that you can rely on in broad
acre farming and the rowe crop farming. So there needed
to be more advanced technology. And thankfully, due to all
the attention and investments that were being made in you know,
(11:58):
the autonomous passenger vehicle world, that technology started to become available.
You know, call it maybe less than a decade ago,
but when I said, you know, kind of seven years ago,
like hey, you know, my expenses are continuing to go
up at tratory. I've automated DAR near everything, but I
haven't automated the field work. Could we take some of
(12:21):
that technology and could we create a solution, an autonomous
solution that could do these repetitive tasks that you need
to do in permanent crops. And so with some friends
built a prototype five years ago and the prototype worked
really well. We literally built an all electric autonomous tractor.
Speaker 1 (12:44):
What did it look like and what did it do?
Speaker 2 (12:46):
Didn't look like a traditional tractor. We took out the
diesel engine. We put in electric motors and batteries, and
we put on cameras and we had you know, an
Nvidia compute system in it, and we started to create
this sensor based vehicle. It had eight wheels, it articulated,
(13:07):
and it could do really steep st and one of
the reason why we did that is because that tratory
the slopes are very very steep. By the way, one
of the main inspirations for doing this autonomous vehicle came
from a documentary I watched. I think it was a
National geographic. It was called Mission to Mars, The Story
of Spirit and Opportunity. And these were Mars rovers in
(13:28):
the nineties that NASA shot up and we're hoping would
operate for you know, fifteen days or you know, a
couple of months, and they ended up lasting almost fifteen years.
And I was watching this documentary before I actually built
this prototype, and it reminded me that, hey, there are
environments where you can take your time in making decisions.
(13:53):
So they had animations of how the rover would approach
a rock and it would just stop and then it
would radio JPL, you know, take some time to radio JPL.
They would take a week and write some new code.
They would then download it, and then the vehicle, the
rover would move to the left that whenever it encountered
a rock, it would move to the laughter to the right.
And I said to myself, my god, you know, they
(14:15):
had autonomous vehicles operating on Mars in the nineties and
they could do it because there wasn't a lot of
traffic on Mars, just like agriculture. And that's when my
head really exploded and I said, wait, there are these
industrial markets that need automation, and you don't necessarily have
to immediately, you know, be responsive to the vehicle if
(14:39):
it encounters something it can't figure out, like oh, there's
a tree in front of.
Speaker 1 (14:43):
Me, unlike on a road in traffic. Correct. Correct, So
when you build this prototype that you don't even know
is a prototype as you're telling it, right, this thing
that you're doing with your friends, I mean it sounds
like a lot of the sort of tex stack is
basically commodified by that point. I mean, is it a
sort of you're at this moment where you can be like, oh, well,
(15:05):
buy these sensors and we'll start with this piece of software,
and like you can almost build an autonomous vehicle from
off the shelf parts. Is that sort of the starting point.
Speaker 2 (15:14):
From a hardware perspective. Absolutely, from a software perspective, no,
at that time, so no, we were actually building a
perception stack, you know that could you know do all
the Well, first you had to localize, so you have
to figure out where in the world do you are,
And so we did need to see GPS at least once,
but then the computer vision technologies and the perception stack
(15:37):
could then align to the crop itself. One of the
big differences that in approach that we took that obviously
the row crops, the broad acre farms don't need to
do is we actually said, hey, what if we exploit
the structure the crop itself and utilize that structure to
navigate safely through it and farm precisely around it.
Speaker 1 (15:59):
Meaning like know what a tree is and know what
your relationship should be to a tree. Is that what
that means? Correct?
Speaker 2 (16:06):
And in the early days we said, hey, there's one
common characteristic and permanent crops. They all have a trunk,
So what if we could be the best detectors of
trunks in the world and utilize that for alignment? And
as I said, also you know, working around that plant.
Speaker 1 (16:25):
So that's the key sort of training piece, Like you
train really hard on trunks.
Speaker 2 (16:30):
That's right, that's right. We initially labeled trunks. We would
drive through, we would get you know, tons of photos
and we would manually label around the trunk area. And
then we would take those images and we would tweak
them and create synthetic data and we would label those
and once you get enough of a data set, you
(16:51):
can start to do some additional machine learning with that, right.
And what was interesting was, you know, we got so
good at detecting trunks. We even called it trunk vision.
We trademark trunk vision. Not that you know, we do
a lot more than trunk vision today, but that's how
that's how we got it to, you know, run through
the orchards and run through the vineyards. And what was
(17:12):
really interesting is we initially trained on grapes, on grapevines
and the rootstock right the trunk. And one day my CTOs,
you know, he said, hey, let's run it in the olives.
And I'm like, oh, but you know, you're sure it's
going to work. He's like, I don't know if it's
going to work. I don't know how well the training
will translate. And so we set it off in the
(17:34):
Olives and it just worked. Now, it wasn't perfect, but
it made us realize, Wow, there's a lot of leverage
between crops here in these permanent crop areas. Generalizes, it
generalizes really well. And of course he's continued to evolve that,
and you know, we now do all kinds of additional
training to make sure in a new crop type we
(17:56):
are very very accurate in exploiting the structure of the crop.
Speaker 1 (18:01):
Besides driving through trees without hitting them, which I recognize
as non trivial, what else did the tractor do? Did
it do work?
Speaker 2 (18:09):
Yes, it did work. It mowed initially. The very very
very first job we did was mowing because we have
cover crop. Sometimes the cover crop gets like six feet
tall and you chop it up with a flail mower
and then it composts in between the trees or the
vines and and even that, like you think mowing, Oh yeah,
(18:30):
you just set it at four miles an hour and
you let it go. You cannot do that, right, because
think about what a human does. The human's on a vehicle.
It's really funny. I asked my engineers you know, well,
did you ever use a push mower when you were
a kid, And they're like, what's a push mower? Anyway?
Speaker 1 (18:45):
Separate topic, but pushmore is really hard. It's harder than
you think that was th pushover means the more without
an engine. It doesn't just mean a gas mower that
you put.
Speaker 2 (18:56):
Right right right and then and then you know, we
got motorized mowers for lawns, right that you that you
know actually had wheels and had an engine on them.
But even with those, if you would go through the
and if it would bog down, as the person operating
that lawn mower, you would slow it down in order
(19:18):
for the blades to catch up, right. And so if
you think about it, you can't just set a tractor
going to mow. You need to get that feedback to
understand if it's bogging down, and therefore you need to
slow down or increase the RPM in order to do
the job properly.
Speaker 1 (19:37):
Right.
Speaker 2 (19:38):
So that's something we learned immediately, But we already thought
about that because we were all farmers. We had farmer
DNA in us, and we knew that's how a human operated.
So we just we started to think, Okay, how does
a human do this? And what are the things we're
gonna need feedback on like a human has in order
to get the job done correctly. So we started with
(19:59):
mowing and we became really really good at it.
Speaker 1 (20:02):
So you have your prototype and you decide to start
a company. What do you have to do to turn
your prototype into you know, something you can sell? Like, yeah,
what's the what's the leap?
Speaker 2 (20:18):
Yeah, this was a big leap. You know. My mechanical
engineering co founders was like, you know, whipping up all
kinds of cool ideas. And then I thought about it
as a farmer, and I said, wait a minute, as
a grower, would I trust a startup company, you know,
for a new tractor? Think about it, right, I'm a
(20:41):
farm I rely upon equipment to get my job done,
and that job is vital for the livelihood of the business,
for the livelihood of the family if it's a family farm.
So why would I trust a startup instead? What I
need as a grower is I need trusted brands. I
(21:04):
need their dealer networks. I need the parts, I need
the service. This this equipment needs to be operating all
the times that I need it for the sustainability of
my business. And so I said to my colleagues, I said,
you know what, I don't think we should build a tractor.
I think that would be rather stupid. And I explained why,
and they said, oh, that makes sense, and they said, well,
(21:27):
how can we work with existing original equipment manufacturers? These
incredible equipment manufacturers who've been at it for over one
hundred years and they build incredible vehicles that really work
well in these harsh environments. And so you know, at
first investors that I was talking about, Hey, we're going
(21:47):
to do an OEM model and you know, original equipment manufacturer,
we're going to help accelerate their digital transformation because we
know where the puck is going, right. We know that
this industry too, just like other industries, needs to go
through a digital transformation. Every day there's less and less
skilled operators who go into agriculture. So the equipment manufacturer know,
(22:09):
you know, they see the writing on the wall. They're
not going to sell as much equipment because like there's
less people to operate them. So they know automation is
really important, and they know how to build incredible iron,
but do they know how to build an AI factory?
And that's what I said, you know we would focus
on is that we would build software to make their
(22:32):
incredible equipment incredibly smart and provide the solutions that make
them autonomous and give growers ultimately new technologies that can
help them save money and get an incredible ROI. It's
almost like building equipment that comes with a skilled operator
built in that you can turn on and off any
time you want.
Speaker 1 (22:53):
As a bonus, you don't have to have the crazy
capital outlay of building a tractor factory, right like, I mean,
that's business. It seems way better.
Speaker 2 (23:03):
It does, but there's one catch, and investors early on
would say, wow, I don't think you can get an OEM.
Speaker 1 (23:11):
And we did an OEM, an original equipment manufacturer attractor.
Speaker 2 (23:16):
Attractor company, and we did an incredible company called Bobcat.
Douce on.
Speaker 1 (23:21):
Bobcat.
Speaker 2 (23:22):
They make excavators large and small, They make skid steers,
they make trackloaders, they make tractors like normal looking tractors, right,
and they they make forklifts. They make all kinds of things.
So you know, really, what we're doing is exactly what
I said earlier, which is to accelerate the digital transformation
(23:46):
of these companies because they know they have to transform,
they know they have to bring autonomy into their equipment
and and and create you know, actonomy enabled vehicles. Right,
So we're not like creating tractors. We're just helping them
with the with the brains of the operation.
Speaker 1 (24:09):
We'll be back in just a minute. What's out there now?
Is your software out in the world now? Dive in
tractors around orchards and what are they doing and where
are they and how many of them are there?
Speaker 2 (24:29):
Yeah, we have tractors throughout the Western United States, and
I'm pleased to say that we also have vehicles in Australia.
So we're super excited about the capabilities we've enabled. In
the state of Washington, for example, where apples are number
one in the world, we're doing a lot of spraying
and a lot of mowing. And it's really it's very
(24:52):
impactful because you're not just having one vehicle, but our
customers have many vehicles. They have a fleet of vehicles,
and they're able to operate those with one of their
employees that they've upskilled to be able to supervise a fleet.
And so it's kind of neat to see, you know,
a whole fleet of vehicles leaving the main shop, driving
(25:15):
on the dirt roads out to the different areas where
they need to do work, do the work, and then
all come back right, So it's a very seamless operation
for them. Our software is also on mobile applications and
that's how the main user experience exists. So think of
you having the tablet or even just your smartphone, and
(25:38):
the site is mapped out, usually with aerial imagery, and
you can just point and say, I want to go
and mow this, you know, this block of the orchard,
and we basically can get close to where the crop is.
And then once we're close to the crop, that perception
stack I mentioned earlier about detecting you know this, or
(26:00):
exploiting the structure of the crop itself. Take trunk trunk
vision and it snaps the vehicle to grid, if you will. Right.
But now, once you have trunk vision and you see
in front of you, and you can see the terrain
to a very high degree of accuracy, you can take
mechanical weaters, for example, and manipulate underneath these trees, you know,
(26:25):
these robotic weaters, and navigate while you're doing that very precisely,
all under computer control. Whereas when a human does it,
you're you know, driving a vehicle six miles an hour
looking forward, you have your hand on six actuators on
the you know, the weater and behind you, and you're
trying to make it all work, and inevitably what happens
(26:46):
is you don't get the weeds so the efficacy is
not high, or you hit the tree and take it out,
and you know that again costs a lot of money.
So that's why herbicides have been used so much because
it's really easy. It's called strip spray and you just
spray it and you're done. But that's that's kind of
a thing that we didn't realize. You know, this kind
(27:07):
of autonomous technology would really open up is much more
offerings for growers so that they don't have to use
very expensive ag inputs or herbicides that they've been doing
in the past.
Speaker 1 (27:21):
All this work that you're describing, is it happening in
a like commercial way now, Like, are you out in
the world getting paid for this? Yes? And what's that
business model?
Speaker 2 (27:31):
On that note, Yeah, so we've been operating in paid
pilots in twenty twenty three and twenty twenty four. We
never intended to be paid for pilots. What happened was
we'd have these demo days. We'd invite growers and they
would say, I want to buy one. I said, well,
they're not for sale. Well I want to rent one.
(27:52):
I'm like, well maybe we can do that. Would you
be interested in doing a pilot? Yes, and I'll pay you.
And we're like okay. So in those original prototypes that
we built, we actually rented them and it was kind
of cool because it was a very small company. We
didn't actually think we were going to recognize revenue until
many years in. And I think our first year we did,
(28:13):
you know, it was actually close to a million dollars.
But then the second year we did several million dollars
and that was in twenty twenty four. And this year
we went big time and we have units operating from Washington, Oregon,
California and now Australia. And so our business model right
now is we work with the manufacturers. They manufacture everything,
(28:36):
they procure all the parts, they bring it into their
manufacturing facilities and they create these very high quality machines
that are factory fit. Right, A lot of people think, oh,
you do retrofit kits on tractors. We do not do that.
We believe it's important to work with the engineers of
these OEMs to make the equipment much more reliable, much safer,
(28:57):
and lower cost, much lower cost when you integrate it in.
Speaker 1 (29:01):
And then when the farmer buys the machine with your
software in it, with your autonomy package in it, like
is it are they paying you a subscription? Is the
price just embedded in the machine and you get some
of the money.
Speaker 2 (29:14):
So the answer is yes to all of those because
in some cases the software fees will be embedded in
the machine. In some cases you will pay a monthly
fee or an annual fee or a one time fee.
Speaker 1 (29:28):
Right.
Speaker 2 (29:29):
Think of serious XM. Serious XM creates technology, it gets
embedded by the OEMs and their manufacturing facilities.
Speaker 1 (29:38):
Satellite radio just to satellite radio, yeah yeah.
Speaker 2 (29:42):
And goes to the dealers, and then the dealers basically
sell to the end customer, and there's a free trial
that comes and after the free trial, Serious XM retains
those customers and charges a monthly fee.
Speaker 1 (29:57):
That's a good metaphor satellite radio, but for autonomous tractors,
for high value.
Speaker 2 (30:02):
Crops exactly, there's another twist. I'll tell you that. By
the way, the chairman of our company, the chairman of
the board is Jim Meyer. He's the former CEO of
Serious ExM.
Speaker 1 (30:12):
Ah okay, yeah.
Speaker 2 (30:14):
And by the way, he's brilliant at business and he
really helps shape this kind of business model.
Speaker 1 (30:19):
What's the technical thing you haven't figured out yet.
Speaker 2 (30:23):
I'll give you a small example, but it's kind of important.
So three point turns. If you think about permanent crops,
you're going down these rows and there's you know, a
person has a ranch it's called one thousand acres, and
there might be a fence around the entire thousand acres
right to keep deer out or whatever it might be.
And they try to maximize the land they have, so
(30:45):
they plant the crop pretty close to the edge of
the fence, let's say. And so when you turn, you
know in the headlands right, the headland turn, you might
not have a lot.
Speaker 1 (30:54):
Of space, so you can't just like make a kind
of U turn at the end of the row and
go down the next.
Speaker 2 (30:59):
Correct correct correct. So you know, right now we can
do about eighty percent of farmlands and permanent crops. But
imagine if we could do point turns, we can achieve
one hundred percent. But but there's a twist, right because
sometimes you not you don't have an implement, that's the
device you attached to the tractor that does the work.
(31:20):
So you have implements you can attach for mowing, for spraying,
for weeding, for disking, for all kinds of things. Some
of those implements are actually trailers.
Speaker 1 (31:30):
H huh.
Speaker 2 (31:30):
Okay, So imagine you know, being on a site slope
with rocky soil and everything, and you have a trailer
and you have to do a three point turn to
get into the next row.
Speaker 1 (31:41):
That's a high skill that's a high skill moment, right,
that's hi know what they're doing exactly exactly.
Speaker 2 (31:47):
And you know, ectonomy enabled vehicles have lots of cameras.
They have you know, minimum of eight cameras including rear
facing cameras and side facing cameras, and so you have
to really now start to you know, do very sophisticated
modeling of the you know, physics of that tractor as
well as that implement or trailer, which we which we
(32:08):
do today as we go forward, but now we have
to do it going backwards. So we haven't mastered that
yet but we're getting close, and I'm super excited about that,
both from a technological achievement perspective as well as a
business need.
Speaker 1 (32:25):
It's a nice when they go together. Yeah, so let's
so madch for a minute, like I'm you know, there
are other people working on other agricultural technology projects, other
ag tech projects, even other autonomous projects. Like when you
zoom out and think about farming and technology more generally
(32:45):
in the medium term, you know whatever that means, five years,
ten years, like, how's the world going to change?
Speaker 2 (32:55):
Well, so, first of all, there are a lot of
others working on similar problems. We all have a slight twist,
either in business model or in technology approach. I think
that in the future is all going to be autonomous capable,
and we see the writing on the wall right when
(33:16):
it comes to labor challenges that we're having, we have
a severe labor gap. If I were to ask you
how many people do you know who you know whose
kids go into agriculture, the answer is probably going to
be not that many. I have three children, Jacob, and
none of them are taking over the farm. And then
you know, of course we have things like you know,
(33:38):
some immigration policy changes that have occurred that are only
exacerbating the problem.
Speaker 1 (33:42):
Even in the absence of a labor shortage, that some margin,
automation wins, right, Like that's what happened with the tractor
and like, whatever's right. So yes, I'm fully prepared to believe.
And yes, I can see how the current politics might
be accelerating the shift to automation. But yes, so okay,
automation is going to win. I stipulated, like what's that
(34:02):
going to mean? What's it going to look like? You know,
tell me something about the future based on that fact.
Speaker 2 (34:07):
Think about ro crops, right, I believe row crops are
going to radically change. So here is one big inflection
point that I think is going to happen. Right now,
we have these large, large tractors in row crops. It
cost a million dollars. Fully autonomous ones cost well over that.
And why are there big tractors. It's because these these farms,
(34:29):
you know, want as many acres done per hour by
one person. But now if automation comes in, you don't
necessarily need these large vehicles that have huge ground compassion.
Speaker 1 (34:40):
This is your trillion dollar play. Now you're going to
disrupt John Deere. Is that where this is.
Speaker 2 (34:45):
Going yes, So I mean, but think about it, right, if.
Speaker 1 (34:50):
You're the only reason you need to make the vehicle
as big as possible is if you have to have
a dude on every one of them. That's what you're saying.
It's not necessarily it could be like like the way
it's like a drone swarm, you could do a drone swarm,
but for tractors.
Speaker 2 (35:02):
Broad acre swarming exactly right, Yeah, And that's there's some
additional benefits, right, So instead of spending and a million dollars,
you can do more work with five tractors that cost
less than that with their implements as well, and now
you have some additional benefits, so you actually get more
work done faster, and if one of the tractors breaks down,
(35:24):
you have redundancy. Right, If your big million dollar tractor
breaks down, you got issues. It's like thousands of dollars
a minute that you know that it's not running. So
I think there's a lot of opportunity in agriculture that
we hadn't thought about once automation is achieved. I mean,
I describe the herbicide challenge in permanent crops. I believe
(35:46):
that is completely can be completely eliminated with automation, and
you know computer controls, computer vision, like you.
Speaker 1 (35:55):
Just pick the weeds instead. If you have essentially very
cheap robots that are very dexterous, you just pick the
weeds instead of spraying them.
Speaker 2 (36:02):
Absolutely, it's already happening in row crops, you know. You
look at carbon robotics, they're using lasers to kill weeds,
you know, in lettuce fields and other kinds of crops
that are called specialty crops. That's huge, right because before
humans would would go in there, or you'd spray and
use bag inputs or herbicides. So I think there's a
(36:23):
huge there is a huge inflection point. But I just
I want to make it clear to your listeners that
just the core base of you know, being able to
produce enough food for the population, we need automation. Four
there are going to be these other benefits, some of
which we don't even know yet. So I'm such a
big believer in automation. It's why I brought my two
(36:44):
worlds together, and it's why I've really dedicated my life
to this mission. And I just I've never been as
excited as I as I am now in terms of
the impact that this company can have our share of impact,
but all of us together, there's so much opportunity here.
And you know, I think the incumbents are the ones
(37:04):
that the incumbent equipment manufacturers are the ones that are
really going to make it happen in terms of distribution
and support and you know, getting it out there to
the world. So couldn't be more excited.
Speaker 1 (37:20):
We'll be back in a minute with the lightning round.
Let's finish with the lightning round. I heard you say
on another interview that you are not a big fan
of IPOs and that you generally prefer acquisitions. Why what
(37:42):
do you got against IPOs.
Speaker 2 (37:45):
Well, that's from the old days. I've went through a
couple of IPOs and just about killed me. So a
lot of people think, you know, IPOs are glamorous and everything,
but you know, the road shows are pretty intense and
the pressures are really intense, and you know, I'm not
just a fan of acquisitions. I mean, I think you're
seeing a lot of differences in the private market than
(38:06):
you have in the past. Companies like open Ai who
are raising billions of dollars in the private market and
not having to go public.
Speaker 1 (38:15):
You could just stay private forever. You could just be
a company that makes more money than it's spends. You
could do that for as long as.
Speaker 2 (38:21):
You want, exactly exactly. So I think the IPO, you know,
I mean, just look at the last few years of IPOs.
There just haven't been that many. And so there are
other ways to build incredibly valuable companies. I get asked
all the time, what's your exit strategy? And I always
have the same answer. I only focus on building value
(38:43):
opportunities come, and I think entrepreneurs who build companies for
an exit aren't building the company for the right reasons
and aren't building really valuable companies either.
Speaker 1 (38:56):
What's one thing that Steve Jobs told you that stuck
with you.
Speaker 2 (39:02):
I had the pleasure of working for some incredible entrepreneurs,
Steve Jobs being one of them. And what I did
learn is that these incredible entrepreneurs, they each had like
a singular mantra. And for Steve, it was all about design.
It's all about design, and there were things he would
push for that we'd sit there and scratch our heads
(39:22):
and go, wait, that's going to add like one hundred
dollars cost to this, and and we would do it
and the product would be successful, and we're like, Okay,
he was right. You know. I learned from from Bill
Gates that it's all about software, and you know what,
he was right. I learned from Michael Dell that it's
(39:43):
all about costs, and guess what, he was right. So
I think, you know, the every entrepreneur that makes serious impact,
you know, has these kind of core mantras. So what's
yours I viewed as a combination. But there is one
thing that I will tell you very very clearly in
your listeners, and I tell my teammates probably every day,
(40:07):
if not every week. You know, it's all about show me,
don't tell me. And that particularly applies to these industrial markets.
The best way to sell industrial equipment, and this is
what dealers do is, you know, hear Susie farmer take
(40:28):
this piece of equipment and use it for a day,
and she'll take that piece of equipment, use it on
her ranch, on her farm, whatever, and inevitably she will
buy it because she gets to try, you know before.
So I think I think the show me mentality is
really really important in this era, and in ag tech
in particular, because there've been some companies that have been
(40:49):
developed by pure tech people and pure tech people. You know,
telling farmers that you know, we can farm better than
you is not a good recipe for success, right, And
so by focusing on show me, it's really what resonates
with our customers, it's really what our customers need, and
(41:11):
it's just how business is done in these industrial worlds.
So that's my mantra, show me, don't tell me.
Speaker 1 (41:21):
What's one ridiculous word somebody once used to describe the
wine that you grow? Hm hmm.
Speaker 2 (41:33):
I didn't even know what this. I still don't know
what this word means. But ethereal, I'm like, like, what
does what does that mean?
Speaker 1 (41:42):
It doesn't stuck around, It's like I don't even know
I drank.
Speaker 2 (41:45):
Yet, Like yeah, I don't know, right, So you get
all kinds of feedback.
Speaker 1 (41:50):
Is that a fancy way of saying easy drinking sounds
like easy drinking? Probably?
Speaker 2 (41:54):
Yeah? Yeah, yeah, But you get all kinds of great feedback.
And I mean that's the cool thing about sort of
the show me mentality is when you develop products that
end users you know, consume or use, and you see
their reaction and you know, there you you please them,
You give them something of value That's that's what gets
(42:16):
me motivated all the time. And that's why you know,
this space is so exciting for me, because that that
impact has felt like really really quickly. When when you
do have a solutions at work, or when you do
make great wine, or when you do make great olive oil.
Speaker 1 (42:32):
Yeah, I don't think autonomous tractors as ethereal, I please don't.
Tim Booker is a farmer and the founder and CEO
of Actonomy. Please email us at problem at pushkin dot fm.
(42:52):
We are always looking for new guests for the show.
Today's show was produced by Trinamanino and Gabriel Hunter Chang,
who was edited by Alexander Garrettson and engineered by Sarah Brugueret.
I'm Jacob Goldstein. We'll be off for the next few
weeks for the holidays. I want to thank you very
much for listening to the show this year. It really
(43:13):
means a tremendous amount to all of us. I hope
you have a great holiday, happy New Year, and we
will be back with more episodes of What's Your Problem
in twenty twenty six.
Popular Podcasts
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
The Joe Rogan Experience
The official podcast of comedian Joe Rogan.
On Purpose with Jay Shetty
I’m Jay Shetty host of On Purpose the worlds #1 Mental Health podcast and I’m so grateful you found us. I started this podcast 5 years ago to invite you into conversations and workshops that are designed to help make you happier, healthier and more healed. I believe that when you (yes you) feel seen, heard and understood you’re able to deal with relationship struggles, work challenges and life’s ups and downs with more ease and grace. I interview experts, celebrities, thought leaders and athletes so that we can grow our mindset, build better habits and uncover a side of them we’ve never seen before. New episodes every Monday and Friday. Your support means the world to me and I don’t take it for granted — click the follow button and leave a review to help us spread the love with On Purpose. I can’t wait for you to listen to your first or 500th episode!