Building a Self-Driving Tractor to Change the Future of Food

0:00:01 This is an iHeart Podcast.
0:00:03 Guaranteed human.
0:00:06 Run a business and not thinking about podcasting?
0:00:07 Think again.
0:00:08 More Americans listen to podcasts,
0:00:11 then add supported streaming music from Spotify and Pandora.
0:00:12 And as the number one podcaster,
0:00:15 iHeart’s twice as large as the next two combined.
0:00:17 Learn how podcasting can help your business.
0:00:19 Call 844-844-iHeart.
0:00:21 Got a business problem?
0:00:23 There’s a TED Talk for that.
0:00:26 Stay updated on everything business on TED Business,
0:00:29 a podcast hosted by Columbia Business School professor
0:00:30 Modupe Akinola.
0:00:32 Every week, she’ll introduce you to leaders
0:00:34 with unique insights on work,
0:00:35 answering questions like,
0:00:38 how do four-day work weeks work?
0:00:40 Do will a machine ever take my job?
0:00:42 Get some surprising answers on TED Business
0:00:44 wherever you listen to podcasts.
0:00:52 Pushkin.
0:00:57 Just a quick note.
0:01:01 Today’s episode is the last one of 2025.
0:01:04 And for 2026, we want to make the show better.
0:01:06 We want to make a show that you like more.
0:01:10 To that end, it would be extremely helpful to us,
0:01:12 and hopefully ultimately also helpful to you,
0:01:15 if you could take a brief survey
0:01:17 that we have put together online.
0:01:19 There’s a link to the survey in the show notes,
0:01:21 and I’ll give you the URL.
0:01:27 It’s tinyurl.com slash WYP listener survey.
0:01:32 Again, tinyurl.com slash WYP,
0:01:34 like what’s your problem, listener survey.
0:01:36 Or just click on the link in the show notes.
0:01:39 Thank you very much for listening to the show,
0:01:42 and thank you in advance for helping us make the show better.
0:01:47 Tim Bucher grew up on his family’s dairy farm
0:01:49 in rural Northern California.
0:01:51 You know how when you’re a kid,
0:01:53 you always wait for summer vacation from school?
0:01:57 My brother and I, we actually,
0:02:00 we hated summer vacation because that meant,
0:02:03 you know, 80 hours a week of work out in the open field.
0:02:05 So, you know, our friends would always say,
0:02:07 oh, I can’t wait until school’s out.
0:02:09 And we would go, no, I don’t.
0:02:10 I want to stay in school.
0:02:12 Tim went off to college,
0:02:14 started out studying agriculture,
0:02:16 but switched to computer science.
0:02:17 And after he graduated,
0:02:20 he founded a string of companies.
0:02:21 One of them he sold to Microsoft,
0:02:23 another he sold to Apple.
0:02:27 But all along, he kept farming on the side.
0:02:29 And eventually, his farming life
0:02:31 and his tech life came together.
0:02:34 As just kind of a weekend side project on the farm,
0:02:38 he and a few friends built an autonomous tractor.
0:02:40 That weekend side project
0:02:42 has now turned into a company
0:02:44 powering autonomous tractors
0:02:46 across the Western U.S. and parts of Australia.
0:02:54 I’m Jacob Goldstein,
0:02:55 and this is What’s Your Problem,
0:02:57 the show where I talk to people
0:02:59 who are trying to make technological progress.
0:03:01 My guest today is Tim Bucher.
0:03:02 He’s a longtime farmer
0:03:05 and the founder and CEO of Agtonomy.
0:03:07 Tim’s problem is this.
0:03:09 How do you bring autonomous vehicles
0:03:11 to specialty crops?
0:03:14 Crops like grapes and olives
0:03:16 and almonds and apples.
0:03:19 There’s already lots of autonomous equipment
0:03:20 for row crops,
0:03:23 like corn and wheat and soybeans.
0:03:24 But as you’ll hear,
0:03:26 specialty crops present
0:03:28 a particularly tricky set of challenges.
0:03:31 Tim and I talked about the big picture,
0:03:34 about what autonomy will mean for farming
0:03:35 and for food.
0:03:36 But to start,
0:03:39 I asked him about how he got from farming
0:03:41 to computer science in the first place.
0:03:44 So you go off to college,
0:03:46 study agriculture,
0:03:47 Davis,
0:03:49 as a California farm kid does, right?
0:03:51 And did I hear you say
0:03:52 in another interview
0:03:55 that you lost a bet
0:03:57 and as a result of losing the bet
0:03:59 had to take a computer science class?
0:04:01 Yes, you heard right.
0:04:02 What was the bet?
0:04:05 Let’s just say it was at a fraternity party.
0:04:06 So it might have involved,
0:04:07 you know,
0:04:08 it might have involved beer.
0:04:09 I don’t know.
0:04:12 I like that it’s a drunken fraternity party
0:04:14 that leads you into a career
0:04:15 as a computer scientist.
0:04:17 Like that is an interesting combination.
0:04:20 Yeah, it was the greatest bet I ever lost.
0:04:23 But what was interesting though,
0:04:25 all my friends growing up,
0:04:27 their last names were Sagatio,
0:04:28 Fappiano, Bacchalupi, Gallo.
0:04:30 These were incredible grape growing
0:04:31 and winemaking families
0:04:33 back in the 70s even.
0:04:36 And I would always go over
0:04:37 to their places to,
0:04:38 you know,
0:04:38 help them
0:04:40 because that’s what playdates were back then
0:04:41 is you basically went over
0:04:42 to your friend’s farm
0:04:43 and worked with them.
0:04:46 And I would learn how to prune grapes
0:04:48 and even make wine.
0:04:50 And I was fascinated by that industry,
0:04:52 like incredibly fascinated
0:04:53 by the growing of grapes.
0:04:54 So, you know,
0:04:55 I went to UC Davis for agriculture
0:04:58 and yes,
0:04:59 I lost a bet
0:05:01 and took a class
0:05:02 in a 500 person auditorium.
0:05:04 Now, keep in mind,
0:05:06 I didn’t grow up with any technology.
0:05:07 I mean, like tractors,
0:05:08 that was about it.
0:05:12 But the professor started talking
0:05:14 and I’m in the last row
0:05:16 and the professor started talking
0:05:18 and everything he was saying,
0:05:19 I understood.
0:05:21 And it just kind of hit me.
0:05:23 And by the end of that quarter,
0:05:25 UC Davis was on the quarter system,
0:05:28 I was up front teaching the class.
0:05:30 And so I knew something happened.
0:05:33 Like I found a passion,
0:05:35 like the passion I had for agriculture.
0:05:39 And so I had a decision to make
0:05:41 is do I go the agriculture route
0:05:43 or do I go this computer science route
0:05:44 or this, you know,
0:05:45 high tech route?
0:05:47 And I said, you know what?
0:05:48 I’m going to do both.
0:05:50 I then graduated from Davis
0:05:52 and I went on to Stanford in grad school
0:05:55 and then started bouncing
0:05:56 between Silicon Valley
0:05:58 and Sonoma County
0:06:00 because when I was 16,
0:06:02 I actually bought my first vineyard.
0:06:04 A small little two acres
0:06:06 and I love farming that.
0:06:08 and that’s kind of what got me
0:06:10 to, you know,
0:06:12 get my own larger place
0:06:13 as I did through the decades,
0:06:16 which is now called Trattori Farms.
0:06:19 Trattori, which means tractor in Italian
0:06:21 because I wanted to name it
0:06:23 after my childhood passion.
0:06:25 And that’s where,
0:06:27 that’s where kind of
0:06:28 that whole thing started.
0:06:31 The duality or the dual path
0:06:33 of agriculture and high tech.
0:06:36 You have these sort of parallel lives
0:06:37 going for a while, right?
0:06:40 You’re working for Steve Jobs.
0:06:41 You’re starting companies.
0:06:44 You’ve got a farm up in Sonoma.
0:06:45 You got a little plane.
0:06:47 You’re flying back and forth, right?
0:06:51 And tell me,
0:06:53 so tell me about the farm part of your life.
0:06:54 Like that’s, you know,
0:06:55 we’re going to bring back
0:06:56 the technology in a minute,
0:06:57 but you’re grown up.
0:07:00 you have a real farm.
0:07:01 Like what’s your farm?
0:07:05 Well, it started as a small little vineyard,
0:07:06 only two acres.
0:07:08 And then through the years,
0:07:10 I would do what’s called
0:07:11 a 1031 tax-free exchange
0:07:12 and get a bigger,
0:07:13 bigger piece of land.
0:07:16 And then, you know,
0:07:16 I don’t know,
0:07:18 about 25 years ago,
0:07:20 I got a much bigger piece of land.
0:07:21 Yeah, two acres, by the way,
0:07:23 two acres isn’t even a farm, right?
0:07:24 Two acres is like a yard.
0:07:26 Yeah, pretty much a yard.
0:07:28 Good call.
0:07:30 But it was French colombard grapes
0:07:32 and Zinfandel grapes.
0:07:34 So, you know,
0:07:37 fast forward to years later
0:07:38 as I would kind of get bigger
0:07:39 and bigger places,
0:07:40 I got this one place,
0:07:41 which was amazing,
0:07:43 beautiful hillsides.
0:07:44 I had a vision for it,
0:07:45 but it was just a forest.
0:07:47 And I started clearing it
0:07:47 for, you know,
0:07:51 the intention of growing more grapes.
0:07:53 And I encountered these stumps,
0:07:54 you know,
0:07:56 equidistant throughout the land.
0:07:56 And I was like,
0:07:57 what is this?
0:07:58 And then finally,
0:08:00 I uncovered some old olive trees
0:08:02 from like the 1800s.
0:08:04 And so did some research
0:08:05 and realized,
0:08:06 well, this was part of a big,
0:08:06 you know,
0:08:07 Spanish land grant
0:08:08 growing olives
0:08:10 back in the 1800s.
0:08:11 So I decided to bring that
0:08:13 history back.
0:08:14 And that’s what got me
0:08:15 into olives.
0:08:16 So I grow olives
0:08:17 for olive oil
0:08:19 and grapes for wine.
0:08:21 And I have a winery operation,
0:08:22 public tasting room.
0:08:24 So kind of bouncing
0:08:25 between two worlds
0:08:26 for decades.
0:08:30 So before you started
0:08:30 Ectonomy,
0:08:33 tell me about the way
0:08:34 you brought technology
0:08:35 to your farm.
0:08:36 Like, what did you automate?
0:08:40 Yeah, I was unique
0:08:41 in that I was a farmer,
0:08:43 but I was also an engineer.
0:08:46 I had both, you know,
0:08:47 electrical engineering background
0:08:48 as well as computer science
0:08:49 background.
0:08:51 So I could do things
0:08:53 like create, you know,
0:08:55 automated water recycling plants
0:08:57 that, you know,
0:08:58 had the flow meters everywhere.
0:08:59 It knew, you know,
0:09:02 how to inject different microbes
0:09:04 to process the wastewater
0:09:06 and then inject it back
0:09:07 into the irrigation automatically.
0:09:10 I had, you know,
0:09:12 obviously all kinds of basic things
0:09:13 like cameras and whatnot.
0:09:16 But even in the fermentation
0:09:17 of grapes,
0:09:19 the tanks are all
0:09:20 computer controlled.
0:09:22 The computer controls
0:09:23 whether to use, you know,
0:09:25 active refrigeration,
0:09:26 which costs energy,
0:09:28 or if the outside air
0:09:30 is cooler than the inside,
0:09:31 it automatically switches over
0:09:33 and brings that kind of air in.
0:09:34 So, you know,
0:09:35 everything from fermentation
0:09:37 processes to water recycling
0:09:38 to irrigation
0:09:40 is radically automated.
0:09:42 So I had automated
0:09:44 darn near everything I could
0:09:46 and I didn’t know
0:09:48 I was actually doing ag tech
0:09:50 when I was doing this.
0:09:52 Ag tech wasn’t a word
0:09:53 that existed back then.
0:09:54 You were just like
0:09:54 puttering around
0:09:55 on the weekends?
0:09:57 I was puttering around
0:09:58 on the weekends
0:09:59 or nighttimes, whatever.
0:10:01 I just knew my farm
0:10:02 needed technology
0:10:04 to automate things
0:10:05 because, keep in mind,
0:10:06 I was also bouncing
0:10:08 between Sonoma County
0:10:09 and Silicon Valley
0:10:12 and so I needed technology
0:10:13 to also be able
0:10:14 to monitor things.
0:10:15 Right.
0:10:16 But there was one area
0:10:17 I hadn’t automated
0:10:18 and that was the actual work
0:10:19 in the orchards
0:10:20 and the vineyards.
0:10:22 And I was sitting there going,
0:10:23 well, you know,
0:10:23 I automated this,
0:10:24 I automated that.
0:10:26 Hey, maybe I should try my hand
0:10:27 at, you know,
0:10:28 automating the field work.
0:10:30 There’s so much money
0:10:31 being invested
0:10:33 in autonomous passenger vehicles
0:10:34 and there’s so much technology
0:10:35 that’s evolved there.
0:10:37 I wonder if I could just
0:10:38 apply that to this space
0:10:40 and see if I can create
0:10:42 an autonomous vehicle
0:10:43 that can do these kinds
0:10:44 of repetitive tasks
0:10:45 like mowing,
0:10:46 like spraying,
0:10:47 you know,
0:10:49 like undervine cultivation,
0:10:50 like seeding,
0:10:51 like disking,
0:10:52 things of that nature
0:10:53 where you’re just going
0:10:53 back and forth,
0:10:55 up and down the rows,
0:10:55 right?
0:10:57 Just to be clear,
0:10:59 there had been
0:10:59 a fair bit of automation,
0:11:00 right?
0:11:01 But on the sort of
0:11:03 big ag row crop
0:11:04 like corn and wheat side,
0:11:06 as I understand it,
0:11:07 but not so much
0:11:08 on the permanent crop
0:11:09 kind of fruit and nut side.
0:11:10 Is that right?
0:11:11 Was that the sort of
0:11:11 state of play?
0:11:13 Exactly, Jacob.
0:11:14 So in row crops
0:11:16 or in broad acre farming,
0:11:16 you know,
0:11:18 the big million dollar tractors
0:11:19 or half a million dollar tractors,
0:11:21 there has been automation
0:11:22 there for decades.
0:11:25 Things called auto steer.
0:11:25 You know,
0:11:27 you’re on these flat areas,
0:11:29 very big fields.
0:11:30 You can see the sky.
0:11:32 GPS had evolved,
0:11:33 so it would be
0:11:34 somewhat accurate,
0:11:35 you know,
0:11:35 a meter,
0:11:36 whatever it might be.
0:11:38 And so you could just
0:11:39 control the vehicles
0:11:40 based on that.
0:11:41 When you think about
0:11:42 permanent crops,
0:11:44 which is a huge percentage
0:11:45 value-wise
0:11:46 of our food supply,
0:11:49 the permanent crops,
0:11:50 you’re talking about
0:11:52 trees and vines
0:11:52 that are very,
0:11:53 very expensive.
0:11:55 They take many years
0:11:55 to develop.
0:11:57 And so if you hit them,
0:11:59 it’s kind of a big deal.
0:12:00 It’s kind of expensive.
0:12:02 And you have to get
0:12:03 really close to them.
0:12:04 So you’re not talking about
0:12:05 being in an open field
0:12:06 where you can be,
0:12:06 you know,
0:12:07 plus or minus half a meter.
0:12:08 You’re talking about
0:12:10 being like one inch away
0:12:11 from, you know,
0:12:12 the trunks of these trees.
0:12:14 and if you accidentally
0:12:15 move, you know,
0:12:17 abruptly or, you know,
0:12:17 don’t steer correctly,
0:12:19 you’re going to hit it.
0:12:21 So you can’t rely on
0:12:22 kind of the things
0:12:23 that you can rely on
0:12:25 in broad acre farming,
0:12:26 in the row crop farming.
0:12:28 So there needed to be
0:12:29 more advanced technology.
0:12:30 And thankfully,
0:12:32 due to all the attention
0:12:33 and investments
0:12:34 that were being made
0:12:36 in, you know,
0:12:37 the autonomous passenger
0:12:39 vehicle world,
0:12:40 that technology started
0:12:42 to become available,
0:12:42 you know,
0:12:43 call it maybe
0:12:45 less than a decade ago.
0:12:47 But when I said,
0:12:47 you know,
0:12:48 kind of seven years ago,
0:12:49 like,
0:12:50 hey, you know,
0:12:51 my expenses are continuing
0:12:52 to go up at Trattori.
0:12:53 I’ve automated
0:12:55 darn near everything,
0:12:56 but I haven’t automated
0:12:57 the field work.
0:12:58 Could we take
0:12:59 some of that technology
0:13:00 and could we create
0:13:01 a solution,
0:13:02 an autonomous solution,
0:13:03 that could do
0:13:05 these repetitive tasks
0:13:06 that you need to do
0:13:08 in permanent crops?
0:13:10 And so with some friends,
0:13:11 built a prototype
0:13:12 five years ago.
0:13:15 And the prototype
0:13:16 worked really well.
0:13:17 We literally built
0:13:18 an all-electric
0:13:20 autonomous tractor.
0:13:22 What did it look like
0:13:23 and what did it do?
0:13:24 Didn’t look like
0:13:25 a traditional tractor.
0:13:26 We took out
0:13:27 the diesel engine,
0:13:29 we put in electric motors
0:13:30 and batteries,
0:13:33 and we put on cameras,
0:13:34 and we had,
0:13:35 you know,
0:13:36 an NVIDIA compute system
0:13:36 in it.
0:13:39 And we started to create
0:13:41 this sensor-based vehicle.
0:13:43 It had eight wheels,
0:13:44 it articulated,
0:13:45 and it could do
0:13:46 really steep slopes.
0:13:47 And one of the reasons
0:13:48 why we did that
0:13:49 is because at Trattori,
0:13:50 the slopes are very,
0:13:50 very steep.
0:13:51 By the way,
0:13:53 one of the main inspirations
0:13:54 for doing this
0:13:55 autonomous vehicle
0:13:56 came from
0:13:57 a documentary I watched.
0:13:59 I think it was
0:14:00 in National Geographic.
0:14:01 It was called
0:14:02 Mission to Mars,
0:14:03 The Story of Spirit
0:14:04 and Opportunity.
0:14:04 And these were
0:14:06 Mars rovers in the 90s
0:14:08 that NASA shot up
0:14:09 and were hoping
0:14:10 would operate
0:14:11 for, you know,
0:14:12 15 days
0:14:13 or, you know,
0:14:14 a couple of months.
0:14:15 And they ended up
0:14:16 lasting almost 15 years.
0:14:18 And I was watching
0:14:19 this documentary
0:14:20 before I actually
0:14:21 built this prototype,
0:14:23 and it reminded me
0:14:24 that, hey,
0:14:25 there are environments
0:14:26 where you can
0:14:28 take your time
0:14:29 in making decisions.
0:14:30 So,
0:14:32 they had animations
0:14:32 of how the rover
0:14:34 would approach a rock
0:14:35 and it would just stop.
0:14:35 And then it would
0:14:36 radio JPL,
0:14:37 you know,
0:14:38 take some time
0:14:39 to radio JPL.
0:14:40 They would take a week
0:14:41 and write some new code.
0:14:42 They would then download it.
0:14:43 And then the vehicle,
0:14:44 the rover,
0:14:45 would move to the left.
0:14:46 And after that,
0:14:47 whenever it encountered a rock,
0:14:48 it would move to the left
0:14:49 or to the right.
0:14:50 And I said to myself,
0:14:51 my God,
0:14:51 you know,
0:14:53 they had autonomous vehicles
0:14:54 operating on Mars
0:14:54 in the 90s.
0:14:56 And they could do it
0:14:57 because there wasn’t
0:14:58 a lot of traffic on Mars,
0:15:00 just like agriculture.
0:15:01 And that’s when,
0:15:02 you know,
0:15:03 my head really exploded.
0:15:04 And I said,
0:15:04 wait,
0:15:06 there are these industrial markets
0:15:07 that need automation
0:15:10 and you don’t necessarily
0:15:12 have to immediately,
0:15:13 you know,
0:15:14 be responsive
0:15:16 to the vehicle
0:15:17 if it encounters something
0:15:18 it can’t figure out.
0:15:19 Like,
0:15:19 oh,
0:15:20 there’s a tree
0:15:20 in front of me.
0:15:22 Unlike on a road
0:15:23 in traffic.
0:15:24 Correct.
0:15:24 Correct.
0:15:26 So when you build
0:15:28 this prototype
0:15:29 that you don’t even know
0:15:29 is a prototype
0:15:30 as you’re telling it,
0:15:30 right?
0:15:31 This thing that you’re
0:15:32 doing with your friends.
0:15:34 I mean,
0:15:35 it sounds like
0:15:36 a lot of the sort of
0:15:36 tech stack
0:15:38 is basically commodified
0:15:38 by that point.
0:15:39 I mean,
0:15:40 is it a sort of,
0:15:41 you’re at this moment
0:15:41 where you can be like,
0:15:42 oh,
0:15:43 we’ll buy these sensors
0:15:44 and we’ll start with this
0:15:45 piece of software
0:15:46 and like,
0:15:47 you can almost build
0:15:48 an autonomous vehicle
0:15:50 from off-the-shelf parts?
0:15:50 Is that sort of
0:15:51 the starting point?
0:15:52 From a hardware perspective,
0:15:53 absolutely.
0:15:55 From a software perspective,
0:15:55 no.
0:15:57 At that time.
0:15:58 So no,
0:15:58 we were actually
0:16:00 building a perception stack
0:16:03 that could do all the,
0:16:03 well,
0:16:04 first you had to localize
0:16:05 so you had to figure out
0:16:05 where in the world
0:16:06 you are.
0:16:08 And so we did need
0:16:08 to see GPS
0:16:09 at least once.
0:16:10 But then
0:16:12 the computer vision
0:16:12 technologies
0:16:14 and the perception stack
0:16:14 could then
0:16:15 align
0:16:17 to the crop itself.
0:16:18 One of the big differences
0:16:19 in approach
0:16:20 that we took
0:16:21 that obviously
0:16:22 the row crops,
0:16:23 the broad acre farms
0:16:24 don’t need to do,
0:16:26 is we actually said,
0:16:26 hey,
0:16:28 what if we exploit
0:16:28 the structure
0:16:29 of the crop itself
0:16:32 and utilize that structure
0:16:33 to navigate safely
0:16:33 through it
0:16:35 and farm precisely
0:16:36 around it?
0:16:37 meaning like know
0:16:38 what a tree is
0:16:39 and know what
0:16:40 your relationship
0:16:41 should be to a tree?
0:16:42 Is that what that means?
0:16:44 And in the early days
0:16:45 we said,
0:16:45 hey,
0:16:46 there’s one common
0:16:47 characteristic
0:16:48 in permanent crops.
0:16:50 They all have a trunk.
0:16:52 So what if we could be
0:16:53 the best
0:16:55 detectors of trunks
0:16:56 in the world
0:16:58 and utilize that
0:16:58 and utilize that
0:16:59 for alignment
0:17:00 and as I said,
0:17:00 also,
0:17:00 you know,
0:17:01 working around
0:17:02 that plant?
0:17:03 So that’s the key
0:17:04 sort of training piece?
0:17:05 Like you train
0:17:06 really hard
0:17:07 on trunks?
0:17:07 That’s right.
0:17:08 That’s right.
0:17:10 We initially labeled
0:17:10 trunks.
0:17:11 We would drive through.
0:17:12 We would get,
0:17:12 you know,
0:17:13 tons of photos
0:17:14 and we would
0:17:15 manually label
0:17:17 around the trunk area
0:17:18 and then we would
0:17:20 take those images
0:17:21 and we would tweak them
0:17:21 and create
0:17:22 synthetic data
0:17:24 and we would
0:17:24 label those
0:17:26 and once you get
0:17:27 enough of a data set
0:17:29 you can start to do
0:17:30 some additional
0:17:30 machine learning
0:17:31 with that,
0:17:31 right?
0:17:33 And what was
0:17:34 interesting was,
0:17:34 you know,
0:17:35 we got so good
0:17:36 at detecting trunks
0:17:37 we even called it
0:17:37 trunk vision
0:17:38 and we trademarked
0:17:39 trunk vision.
0:17:40 Not that,
0:17:41 you know,
0:17:41 we do a lot more
0:17:42 than trunk vision today
0:17:44 but that’s how
0:17:46 we got it to,
0:17:46 you know,
0:17:47 run through the orchards
0:17:49 and run through the vineyards
0:17:50 and what was really
0:17:50 interesting is we
0:17:51 initially trained
0:17:52 on grapes
0:17:53 on grapevines
0:17:55 and the rootstock,
0:17:55 right?
0:17:55 The trunk
0:17:57 and one day
0:17:59 my CTO,
0:17:59 you know,
0:18:00 he said,
0:18:00 hey,
0:18:01 let’s run it
0:18:01 in the olives
0:18:03 and I’m like,
0:18:03 oh,
0:18:03 but,
0:18:03 you know,
0:18:04 you sure it’s going to work?
0:18:04 He’s like,
0:18:06 I don’t know if it’s going to work.
0:18:07 I don’t know how well
0:18:09 the training will translate
0:18:11 and so we set it off
0:18:12 in the olives
0:18:14 and it just worked.
0:18:15 Now,
0:18:16 it wasn’t perfect
0:18:18 but it made us realize,
0:18:18 wow,
0:18:19 there’s a lot of leverage
0:18:20 between crops here
0:18:21 in these permanent crop
0:18:22 areas
0:18:23 It generalizes.
0:18:25 It generalizes really well
0:18:26 and of course
0:18:27 he’s continued
0:18:28 to evolve that
0:18:28 and you know,
0:18:29 we now do all kinds
0:18:31 of additional training
0:18:31 to make sure
0:18:33 in a new crop type
0:18:34 we are very,
0:18:34 very accurate
0:18:36 in exploiting
0:18:36 the structure
0:18:37 of the crop.
0:18:40 besides driving through trees
0:18:41 without hitting them,
0:18:42 which I recognize
0:18:42 is non-trivial,
0:18:44 what else did the tractor do?
0:18:46 Did it do work?
0:18:47 Yes,
0:18:47 it did work.
0:18:48 It mowed.
0:18:50 Initially,
0:18:50 the very,
0:18:51 very,
0:18:51 very first job
0:18:52 we did was mowing
0:18:54 because we have cover crops.
0:18:54 Sometimes the cover crop
0:18:56 gets like six feet tall
0:18:57 and you chop it up
0:18:59 with a flail mower
0:19:01 and then it composts
0:19:02 in between the trees
0:19:03 or the vines
0:19:05 and even that,
0:19:06 like you think mowing,
0:19:07 oh yeah,
0:19:07 you just set it
0:19:08 at four miles an hour
0:19:09 and you let it go.
0:19:10 You cannot do that,
0:19:11 right?
0:19:11 Because think about
0:19:12 what a human does.
0:19:13 A human’s on a vehicle.
0:19:15 It’s really funny.
0:19:16 I asked my engineers,
0:19:17 you know,
0:19:18 well,
0:19:18 did you ever use
0:19:19 a push mower
0:19:19 when you were a kid?
0:19:20 And they’re like,
0:19:21 what’s a push mower?
0:19:22 Anyway,
0:19:23 separate topic.
0:19:25 Push mower is really hard.
0:19:26 It’s harder than you think.
0:19:28 That was my experience.
0:19:29 Push mower means
0:19:30 a mower without an engine.
0:19:32 It doesn’t just mean
0:19:32 a gas mower
0:19:33 that you push.
0:19:33 Right,
0:19:34 right.
0:19:34 Just to be clear.
0:19:35 Right.
0:19:35 And then,
0:19:36 and then,
0:19:36 you know,
0:19:37 we got motorized
0:19:40 mowers for lawns,
0:19:40 right?
0:19:40 That you,
0:19:41 that,
0:19:41 you know,
0:19:42 actually had wheels
0:19:43 and had an engine on them.
0:19:45 But even with those,
0:19:46 if you would go through the grass
0:19:48 and if it would bog down
0:19:50 as the person
0:19:52 operating that lawn mower,
0:19:54 you would slow it down
0:19:55 in order for the blades
0:19:56 to catch up,
0:19:56 right?
0:19:58 And so,
0:19:59 if you think about it,
0:19:59 you can’t just
0:20:01 set a tractor going
0:20:02 to mow.
0:20:03 you need to
0:20:05 get that feedback
0:20:06 to understand
0:20:07 if it’s bogging down
0:20:08 and therefore
0:20:10 you need to slow down
0:20:11 or increase the RPM
0:20:13 in order to do
0:20:14 the job properly,
0:20:14 right?
0:20:15 So,
0:20:16 that’s something
0:20:18 we learned immediately.
0:20:19 But we already
0:20:20 thought about that
0:20:21 because we were
0:20:21 all farmers.
0:20:23 We had farmer DNA
0:20:23 in us
0:20:24 and we knew
0:20:25 that’s how a human
0:20:25 operated.
0:20:26 so we just,
0:20:27 we started to think,
0:20:27 okay,
0:20:28 how does a human
0:20:29 do this
0:20:30 and what are the things
0:20:30 we’re going to need
0:20:31 feedback on
0:20:32 like a human has
0:20:33 in order to get
0:20:35 the job done correctly?
0:20:36 So,
0:20:37 we started with mowing
0:20:38 and we became
0:20:38 really,
0:20:39 really good at it.
0:20:41 so you have your
0:20:41 prototype
0:20:42 and you decide
0:20:43 to start a company.
0:20:47 What do you have to do
0:20:48 to turn your prototype
0:20:49 into,
0:20:50 you know,
0:20:51 something you can sell?
0:20:52 Like,
0:20:53 what’s the,
0:20:54 what’s the leap?
0:20:55 yeah,
0:20:57 this was a big leap.
0:20:58 You know,
0:21:00 my mechanical engineering
0:21:01 co-founders was like,
0:21:01 you know,
0:21:02 whipping up all kinds
0:21:04 of cool ideas.
0:21:06 And then I thought about it
0:21:07 as a farmer
0:21:07 and I said,
0:21:08 wait a minute,
0:21:09 as a grower,
0:21:11 would I trust
0:21:12 a startup company,
0:21:13 you know,
0:21:14 for a new tractor?
0:21:16 Think about it,
0:21:17 right?
0:21:18 I’m a farm.
0:21:20 I rely upon
0:21:21 equipment
0:21:23 to get my job done
0:21:25 and that job
0:21:27 is vital
0:21:28 for the livelihood
0:21:28 of the business,
0:21:29 for the livelihood
0:21:30 of the family,
0:21:31 if it’s a family farm.
0:21:32 So,
0:21:33 why would I trust
0:21:34 a,
0:21:35 a startup?
0:21:36 Instead,
0:21:38 what I need
0:21:38 as a grower
0:21:39 is I need
0:21:40 trusted brands,
0:21:41 I need their
0:21:42 dealer networks,
0:21:43 I need the parts,
0:21:44 I need the service,
0:21:45 this,
0:21:46 this equipment
0:21:47 needs to be operating
0:21:49 all the times
0:21:49 that I need it
0:21:51 for the sustainability
0:21:52 of my business.
0:21:54 and so I said
0:21:54 to my colleagues,
0:21:55 I said,
0:21:55 you know what,
0:21:56 I don’t think
0:21:56 we should build
0:21:57 a tractor.
0:21:58 I think that would
0:21:59 be rather stupid
0:22:00 and I explained
0:22:01 why and they said,
0:22:01 oh,
0:22:02 that makes sense
0:22:03 and I said,
0:22:03 well,
0:22:04 how can we
0:22:06 work with
0:22:07 existing original
0:22:08 equipment manufacturers,
0:22:09 these incredible
0:22:10 equipment manufacturers
0:22:11 who’ve been at it
0:22:12 for over 100 years
0:22:13 and they build
0:22:15 incredible vehicles
0:22:17 that really work
0:22:18 well in these
0:22:18 harsh environments
0:22:20 and so,
0:22:21 you know,
0:22:22 at first,
0:22:23 investors that I
0:22:23 was talking about,
0:22:24 hey,
0:22:24 we’re going to do
0:22:25 an OEM model
0:22:25 and,
0:22:26 you know,
0:22:26 original equipment
0:22:27 manufacturer,
0:22:28 we’re going to help
0:22:30 accelerate their
0:22:31 digital transformation
0:22:32 because we know
0:22:33 where the puck is going,
0:22:33 right?
0:22:34 We know that
0:22:36 this industry too,
0:22:36 just like other
0:22:37 industries,
0:22:38 needs to go through
0:22:39 a digital transformation.
0:22:41 Every day,
0:22:41 there’s less and less
0:22:42 skilled operators
0:22:43 who go into
0:22:44 agriculture
0:22:45 so the equipment
0:22:46 manufacturers know,
0:22:47 you know,
0:22:47 they see the writing
0:22:48 on the wall.
0:22:49 They’re not going to
0:22:50 sell as much equipment
0:22:51 because,
0:22:51 like,
0:22:52 there’s less people
0:22:52 to operate them.
0:22:53 So,
0:22:54 they know automation
0:22:55 is really important
0:22:57 and they know
0:22:57 how to build
0:22:58 incredible iron
0:22:59 but do they know
0:22:59 how to build
0:23:00 an AI factory?
0:23:02 And that’s what I said,
0:23:02 you know,
0:23:03 we would focus on
0:23:04 is that we would
0:23:06 build software
0:23:08 to make
0:23:09 their incredible
0:23:10 equipment
0:23:11 incredibly smart
0:23:13 and provide
0:23:14 the solutions
0:23:15 that make
0:23:16 them autonomous
0:23:16 and give
0:23:17 growers
0:23:18 ultimately
0:23:20 new technologies
0:23:21 that can help
0:23:21 them save money
0:23:22 and get an
0:23:23 incredible ROI.
0:23:24 It’s almost like
0:23:25 building equipment
0:23:26 that comes with
0:23:26 a skilled operator
0:23:27 built in
0:23:27 that you can
0:23:28 turn on and off
0:23:29 anytime you want.
0:23:31 And as a bonus,
0:23:32 you don’t have to
0:23:32 have the crazy
0:23:33 capital outlay
0:23:34 of building
0:23:36 a tractor factory,
0:23:36 right?
0:23:36 Like,
0:23:37 I mean,
0:23:37 as a business,
0:23:39 it seems way better.
0:23:41 It does,
0:23:42 but there’s one catch.
0:23:44 and investors
0:23:45 early on
0:23:45 would say,
0:23:46 well,
0:23:46 I don’t think
0:23:46 you can get
0:23:47 an OEM.
0:23:49 And we did.
0:23:50 An OEM,
0:23:51 an original
0:23:52 equipment manufacturer,
0:23:52 a tractor company.
0:23:53 Yeah,
0:23:53 a tractor company.
0:23:55 And we did
0:23:56 an incredible
0:23:57 company called
0:23:57 Bobcat,
0:23:59 Doosan Bobcat.
0:24:00 They make
0:24:01 excavators,
0:24:02 large and small.
0:24:02 They make
0:24:03 skid steers.
0:24:04 They make
0:24:05 track loaders.
0:24:05 They make
0:24:06 tractors,
0:24:09 like normal
0:24:10 looking tractors,
0:24:10 right?
0:24:12 And they
0:24:13 make
0:24:14 forklifts.
0:24:14 They make
0:24:15 all kinds of
0:24:15 things.
0:24:15 So,
0:24:16 you know,
0:24:17 really what we’re
0:24:18 doing is
0:24:19 exactly what I said
0:24:19 earlier,
0:24:20 which is to
0:24:21 accelerate the
0:24:22 digital transformation
0:24:23 of these companies
0:24:24 because they know
0:24:24 they have to
0:24:25 transform.
0:24:26 They know
0:24:26 they have to
0:24:27 bring autonomy
0:24:27 into their
0:24:29 equipment
0:24:32 and create,
0:24:32 you know,
0:24:33 autonomy-enabled
0:24:34 vehicles,
0:24:34 right?
0:24:35 So,
0:24:35 we’re not like
0:24:37 creating tractors.
0:24:38 We’re just
0:24:39 helping them
0:24:40 with the brains
0:24:42 of the operation.
0:24:46 We’ll be back
0:24:47 in just a minute.
0:24:57 Run a business
0:24:58 and not thinking
0:24:58 about podcasting?
0:24:59 Think again.
0:25:00 More Americans
0:25:01 listen to podcasts
0:25:02 than ad-supported
0:25:03 streaming music
0:25:03 from Spotify
0:25:04 and Pandora.
0:25:05 And as the
0:25:05 number one
0:25:06 podcaster,
0:25:07 iHeart’s twice
0:25:07 as large as
0:25:08 the next two
0:25:08 combined.
0:25:09 So,
0:25:09 whatever your
0:25:10 customers listen
0:25:10 to,
0:25:11 they’ll hear
0:25:11 your message.
0:25:12 Plus,
0:25:12 only iHeart
0:25:13 can extend
0:25:13 your message
0:25:14 to audiences
0:25:15 across broadcast
0:25:15 radio.
0:25:17 Think podcasting
0:25:17 can help your
0:25:17 business?
0:25:18 Think iHeart.
0:25:19 Streaming,
0:25:20 radio,
0:25:21 and podcasting.
0:25:23 Call 844-844-IHEART
0:25:24 to get started.
0:25:26 That’s 844-844-IHEART.
0:25:28 Got a business
0:25:29 problem?
0:25:29 There’s a TED Talk
0:25:30 for that.
0:25:31 Stay updated
0:25:31 on everything
0:25:32 business on
0:25:33 TED Business,
0:25:34 a podcast hosted
0:25:35 by Columbia Business
0:25:36 School professor
0:25:37 Modupe Akinola.
0:25:38 Every week,
0:25:39 she’ll introduce you
0:25:40 to leaders with
0:25:41 unique insights
0:25:41 on work,
0:25:42 answering questions
0:25:42 like,
0:25:44 how do four-day
0:25:45 work weeks work?
0:25:46 Do will a machine
0:25:47 ever take my job?
0:25:48 Get some surprising
0:25:50 answers on TED Business
0:25:51 wherever you listen
0:25:51 to podcasts.
0:25:55 What’s out there
0:25:56 now?
0:25:57 Is your software
0:25:57 out in the world
0:25:58 now,
0:25:59 driving tractors
0:26:00 around orchards?
0:26:01 What are they doing?
0:26:02 Where are they?
0:26:02 How many of them
0:26:03 are there?
0:26:05 We have tractors
0:26:05 throughout the
0:26:06 western United States
0:26:07 and I’m pleased
0:26:08 to say that
0:26:10 we also have
0:26:11 vehicles in Australia.
0:26:12 We’re super excited
0:26:14 about the capabilities
0:26:15 we’ve enabled.
0:26:17 In the state of
0:26:17 Washington,
0:26:18 for example,
0:26:19 where apples
0:26:19 are number one
0:26:20 in the world,
0:26:22 we’re doing a lot
0:26:22 of spraying
0:26:24 and a lot of mowing.
0:26:25 And it’s really,
0:26:27 it’s very impactful
0:26:28 because you’re not
0:26:30 just having one vehicle,
0:26:31 but our customers
0:26:32 have many vehicles.
0:26:33 They have a fleet
0:26:33 of vehicles
0:26:34 and they’re able
0:26:35 to operate those
0:26:37 with one of their
0:26:37 employees
0:26:39 that they’ve upskilled
0:26:40 to be able
0:26:41 to supervise a fleet.
0:26:43 And so it’s kind of
0:26:44 neat to see,
0:26:45 you know,
0:26:46 a whole fleet
0:26:47 of vehicles
0:26:48 leaving the main shop,
0:26:50 driving on the dirt roads
0:26:51 out to the different
0:26:52 areas where they
0:26:53 need to do work,
0:26:54 do the work,
0:26:56 and then all come back,
0:26:56 right?
0:26:58 So it’s a very
0:27:00 seamless operation
0:27:00 for them.
0:27:01 Our software is also
0:27:03 on mobile applications
0:27:04 and that’s how
0:27:05 the main user
0:27:06 experience exists.
0:27:08 So think of you
0:27:09 having a tablet
0:27:10 or even just
0:27:10 your smartphone
0:27:13 and the site
0:27:14 is mapped out
0:27:15 usually with
0:27:15 aerial imagery
0:27:17 and you can just
0:27:18 point and say,
0:27:18 I want to go
0:27:20 and mow this,
0:27:20 you know,
0:27:21 this block
0:27:22 of the orchard
0:27:24 and we basically
0:27:25 can get close
0:27:26 to where the
0:27:27 crop is
0:27:28 and then once
0:27:29 we’re close
0:27:29 to the crop,
0:27:30 that perception
0:27:31 stack I mentioned
0:27:32 earlier about
0:27:32 detecting,
0:27:33 you know,
0:27:34 or exploiting
0:27:34 the structure
0:27:35 of the crop
0:27:36 itself takes over.
0:27:36 Trunk vision.
0:27:37 Trunk vision.
0:27:39 And it snaps
0:27:40 the vehicle to grid,
0:27:40 if you will.
0:27:41 Right?
0:27:42 But now,
0:27:43 once you have
0:27:44 trunk vision
0:27:45 and you see
0:27:46 in front of you
0:27:47 and you can see
0:27:48 the terrain
0:27:49 to a very
0:27:50 high degree
0:27:51 of accuracy,
0:27:53 you can take
0:27:54 mechanical weeders,
0:27:55 for example,
0:27:56 and manipulate
0:27:57 underneath these
0:27:58 trees,
0:27:59 you know,
0:27:59 these robotic
0:28:02 weeders
0:28:03 and navigate
0:28:04 while you’re
0:28:05 doing that
0:28:06 very precisely
0:28:07 all under
0:28:07 computer control.
0:28:09 Whereas when a
0:28:09 human does it,
0:28:10 you’re,
0:28:10 you know,
0:28:11 driving a vehicle
0:28:12 six miles an hour
0:28:12 looking forward,
0:28:14 you have your hand
0:28:15 on six actuators
0:28:16 on the,
0:28:16 you know,
0:28:16 the weeder
0:28:17 behind you
0:28:18 and you’re trying
0:28:18 to make it
0:28:19 all work
0:28:19 and inevitably
0:28:20 what happens
0:28:21 is you don’t
0:28:22 get the weeds
0:28:23 so the efficacy
0:28:24 is not high
0:28:25 or you hit the tree
0:28:26 and take it out
0:28:27 and, you know,
0:28:28 that again costs
0:28:29 a lot of money.
0:28:31 So that’s why
0:28:31 herbicides have
0:28:32 been used so much
0:28:33 because it’s really easy.
0:28:33 It’s called strip
0:28:34 spraying.
0:28:35 You just spray it
0:28:35 and you’re done
0:28:37 but that’s
0:28:38 that’s kind of
0:28:39 a thing that
0:28:40 we didn’t realize
0:28:41 you know,
0:28:41 this kind of
0:28:42 autonomous technology
0:28:43 would really open up
0:28:45 is much more
0:28:47 offerings for growers
0:28:48 so that they don’t
0:28:49 have to use
0:28:50 very expensive
0:28:51 ag inputs
0:28:52 or herbicides
0:28:54 that they’ve been
0:28:55 doing in the past.
0:28:56 All this work
0:28:57 that you’re describing
0:28:57 is it happening
0:28:58 in a like
0:29:00 commercial way now?
0:29:01 Like are you
0:29:01 out in the world
0:29:02 getting paid for this?
0:29:03 Yes, we are.
0:29:03 And what’s the business
0:29:05 model on that note?
0:29:06 Yeah, so
0:29:07 we’ve been operating
0:29:09 in paid pilots
0:29:10 in 2023
0:29:11 and 2024.
0:29:13 We never intended
0:29:14 to be paid
0:29:15 for pilots.
0:29:17 What happened was
0:29:18 we’d have these
0:29:19 demo days
0:29:19 and we’d invite
0:29:20 growers
0:29:21 and they would say
0:29:22 I want to buy one.
0:29:23 I said well
0:29:24 they’re not for sale.
0:29:25 Well, I want to rent one.
0:29:26 I’m like well
0:29:27 maybe we can do that.
0:29:28 Would you be interested
0:29:29 in doing a pilot?
0:29:31 Yes, and I’ll pay you.
0:29:32 And we’re like
0:29:33 okay.
0:29:34 So in those original
0:29:35 prototypes that we built
0:29:36 we actually
0:29:38 rented them
0:29:39 and it was kind of cool
0:29:40 because it was a very
0:29:41 small company
0:29:42 we didn’t actually think
0:29:42 we were going to
0:29:43 recognize revenue
0:29:44 until many years in.
0:29:46 And I think our first year
0:29:47 we did
0:29:48 you know
0:29:48 it was actually
0:29:50 close to a million dollars
0:29:51 but then the second year
0:29:52 we did several million dollars
0:29:53 and that was in 2024.
0:29:56 and this year
0:29:57 we went big time
0:29:59 and we have units
0:30:00 operating from
0:30:01 Washington, Oregon,
0:30:01 California
0:30:03 and now Australia.
0:30:04 And so
0:30:05 our business model
0:30:06 right now
0:30:07 is we work with
0:30:08 the manufacturers
0:30:10 they manufacture
0:30:10 everything
0:30:11 they procure all the parts
0:30:12 they bring it into
0:30:13 their manufacturing facilities
0:30:15 and they create
0:30:16 these very high quality
0:30:16 machines
0:30:18 that are factory fit.
0:30:19 Right?
0:30:19 A lot of people think
0:30:21 oh you do retrofit kits
0:30:21 on tractors.
0:30:23 We do not do that.
0:30:24 we believe it’s important
0:30:25 to work with the engineers
0:30:26 of these OEMs
0:30:27 to make the equipment
0:30:29 much more reliable
0:30:30 much safer
0:30:32 and lower cost
0:30:33 much lower cost
0:30:34 when you integrate it in.
0:30:36 And then when the farmer
0:30:37 buys the machine
0:30:39 with your software in it
0:30:40 with your autonomy package
0:30:40 in it
0:30:41 like
0:30:42 is it
0:30:43 is it
0:30:44 are they paying you
0:30:45 a subscription?
0:30:46 Is the price just
0:30:47 embedded in the machine
0:30:47 and you get some
0:30:48 of the money?
0:30:49 so
0:30:50 the answer is
0:30:51 yes to all of those
0:30:52 because in some cases
0:30:55 the software fees
0:30:55 will be embedded
0:30:56 in the machine
0:30:58 in some cases
0:30:58 you will pay
0:30:59 a monthly fee
0:31:00 or an annual fee
0:31:01 or a one time fee
0:31:02 right?
0:31:04 Think of Sirius XM
0:31:06 Sirius XM
0:31:08 creates technology
0:31:09 it gets embedded
0:31:10 by the OEMs
0:31:11 in their manufacturing facilities
0:31:13 satellite radio
0:31:13 just
0:31:13 just to
0:31:14 satellite radio
0:31:17 and goes to the dealers
0:31:19 and then the dealers
0:31:20 basically sell
0:31:21 to the end customer
0:31:23 and there’s a free trial
0:31:23 that comes
0:31:25 and after the free trial
0:31:27 Sirius XM
0:31:28 retains those customers
0:31:29 and charges
0:31:30 a monthly fee
0:31:32 that’s a good metaphor
0:31:34 satellite radio
0:31:35 but for autonomous tractors
0:31:37 for high value crops
0:31:38 exactly
0:31:39 there’s another twist
0:31:40 I’ll tell you
0:31:40 by the way
0:31:41 the chairman
0:31:42 the chairman of our company
0:31:43 the chairman of the board
0:31:44 is Jim Meyer
0:31:45 he’s the former CEO
0:31:46 of Sirius XM
0:31:47 aha
0:31:47 okay
0:31:47 yeah
0:31:48 and by the way
0:31:50 he’s brilliant at business
0:31:51 and he really helps shape
0:31:53 this kind of business model
0:31:54 what’s a technical thing
0:31:55 you haven’t figured out yet?
0:31:58 I’ll give you a small example
0:32:00 but it’s kind of important
0:32:01 so
0:32:02 three point turns
0:32:03 if you think about
0:32:05 permanent crops
0:32:06 you’re going down
0:32:06 these rows
0:32:08 and there’s
0:32:08 you know
0:32:09 a person has a ranch
0:32:10 let’s call it a thousand acres
0:32:12 and there might be a fence
0:32:13 around the entire thousand acres
0:32:14 right
0:32:15 to keep deer out
0:32:15 or whatever it might be
0:32:17 and they try to
0:32:19 maximize the land they have
0:32:20 so they plant the crop
0:32:21 pretty close to the edge
0:32:21 of the fence
0:32:22 let’s say
0:32:24 and so when you turn
0:32:24 you know
0:32:25 in the headlands
0:32:26 right
0:32:27 the headland turn
0:32:28 you might not have
0:32:29 a lot of space
0:32:30 so you can’t just like
0:32:31 make a kind of u-turn
0:32:32 at the end of the row
0:32:33 and go down the next row
0:32:34 correct
0:32:34 correct
0:32:35 correct
0:32:36 so you know
0:32:37 right now
0:32:38 we can do about 80%
0:32:39 of farmlands
0:32:41 and permanent crops
0:32:42 but imagine
0:32:42 if we could do
0:32:43 three point turns
0:32:45 we can achieve 100%
0:32:46 but
0:32:47 but there’s a twist
0:32:48 right
0:32:48 because sometimes
0:32:49 you not
0:32:50 you don’t have an implement
0:32:51 that’s the device
0:32:53 you attach to the tractor
0:32:53 that does the work
0:32:54 so you have
0:32:55 implements
0:32:55 you can attach
0:32:56 for mowing
0:32:57 for spraying
0:32:57 for weeding
0:32:58 for disking
0:32:59 for all kinds of things
0:33:01 some of those implements
0:33:03 are actually trailers
0:33:04 uh-huh
0:33:05 okay
0:33:06 so imagine
0:33:06 you know
0:33:08 being on a slight slope
0:33:09 with rocky soil
0:33:09 and everything
0:33:11 and you have a trailer
0:33:12 and you have to do
0:33:13 a three point turn
0:33:15 to get into the next row
0:33:16 it’s a high skill
0:33:18 that’s a high skill moment
0:33:18 right
0:33:19 the operator has to know
0:33:20 what they’re doing
0:33:20 yeah
0:33:20 exactly
0:33:21 exactly
0:33:22 and you know
0:33:23 autonomy enabled vehicles
0:33:24 have lots of cameras
0:33:25 they have you know
0:33:26 minimum of eight cameras
0:33:27 including rear-facing cameras
0:33:29 and side-facing cameras
0:33:29 and so
0:33:30 you have to really
0:33:31 now start to
0:33:32 you know
0:33:33 do very sophisticated
0:33:34 modeling
0:33:35 of the
0:33:36 you know
0:33:37 physics of that
0:33:38 tractor
0:33:39 as well as that
0:33:40 implement or trailer
0:33:41 um
0:33:42 which we
0:33:43 which we do today
0:33:44 as we go forward
0:33:45 but now we have to do it
0:33:45 going backwards
0:33:46 so
0:33:48 we haven’t mastered that yet
0:33:49 but we’re getting close
0:33:50 and I’m super excited
0:33:51 about that
0:33:52 both from a
0:33:53 technological achievement
0:33:53 perspective
0:33:55 as well as
0:33:55 a business
0:33:56 need
0:33:58 um
0:33:59 it’s nice
0:34:00 when they go together
0:34:00 yeah
0:34:01 um
0:34:02 so
0:34:03 let’s zoom out
0:34:04 for a minute
0:34:04 like I’m
0:34:05 I’m
0:34:06 you know
0:34:07 there are other people
0:34:08 working on other
0:34:09 um
0:34:11 agricultural
0:34:13 technology projects
0:34:14 other ag tech projects
0:34:15 even other autonomous projects
0:34:16 like when you zoom out
0:34:17 and think about
0:34:18 farming and technology
0:34:19 more generally
0:34:20 in the
0:34:21 medium term
0:34:22 you know
0:34:23 whatever that means
0:34:23 five years
0:34:24 10 years
0:34:24 like
0:34:26 how’s the world
0:34:27 going to change
0:34:29 well
0:34:30 so
0:34:31 first of all
0:34:32 there are a lot of others
0:34:32 working on
0:34:33 similar problems
0:34:34 um
0:34:37 we all have a slight twist
0:34:38 either in business model
0:34:39 or in technology approach
0:34:41 I think
0:34:42 I think that
0:34:43 equipment in the future
0:34:45 is all going to be
0:34:46 autonomous capable
0:34:48 and
0:34:49 we see the writing
0:34:50 on the wall
0:34:50 right
0:34:51 when it comes to
0:34:52 labor challenges
0:34:53 that we’re having
0:34:55 we have a severe
0:34:56 labor gap
0:34:56 um
0:34:58 if I were to ask you
0:34:58 how many people
0:34:59 do you know
0:34:59 who
0:34:59 you know
0:35:00 whose kids
0:35:01 go into agriculture
0:35:02 the answer is probably
0:35:02 going to be
0:35:03 not that many
0:35:05 I have three children
0:35:06 Jacob
0:35:06 and
0:35:07 none of them
0:35:08 are taking over the farm
0:35:09 and then
0:35:10 you know
0:35:10 of course
0:35:11 we have things
0:35:11 like
0:35:11 you know
0:35:12 some
0:35:13 immigration policy
0:35:13 changes
0:35:14 that have occurred
0:35:14 that are only
0:35:16 exacerbating the problem
0:35:17 even in the absence
0:35:18 of a labor shortage
0:35:19 at some margin
0:35:20 automation wins
0:35:20 right
0:35:21 like that’s what
0:35:22 happened with the tractor
0:35:22 and like
0:35:23 whatever
0:35:24 that’s right
0:35:25 so yes
0:35:26 I’m fully prepared
0:35:26 to believe
0:35:26 and yes
0:35:27 I can see
0:35:28 how the current
0:35:28 politics
0:35:29 might be accelerating
0:35:30 the shift to automation
0:35:31 but yes
0:35:31 so okay
0:35:33 automation is going to win
0:35:35 I stipulated
0:35:35 like
0:35:37 what’s that going to mean
0:35:38 what’s it going to look like
0:35:38 you know
0:35:39 tell me something
0:35:39 about the future
0:35:41 based on that fact
0:35:42 think about
0:35:43 row crops
0:35:43 right
0:35:44 I believe
0:35:45 row crops
0:35:45 are going to
0:35:45 radically
0:35:46 change
0:35:46 so here is
0:35:47 one big
0:35:48 inflection point
0:35:48 that I think
0:35:49 is going to happen
0:35:50 right now
0:35:51 we have these
0:35:51 large
0:35:52 large tractors
0:35:53 in row crops
0:35:55 that cost a million dollars
0:35:56 fully autonomous ones
0:35:57 cost well over that
0:35:58 and
0:35:59 why are there
0:36:00 big tractors
0:36:01 it’s because
0:36:02 these
0:36:03 these farms
0:36:03 you know
0:36:04 want as many acres
0:36:05 done per hour
0:36:06 by one person
0:36:08 but now
0:36:09 if automation
0:36:09 comes in
0:36:10 you don’t necessarily
0:36:11 need these
0:36:12 large vehicles
0:36:13 that have huge
0:36:14 ground compaction
0:36:15 this is your
0:36:16 trillion dollar
0:36:16 play now
0:36:17 you’re going to
0:36:17 disrupt
0:36:18 John Deere
0:36:18 is that
0:36:18 where this
0:36:19 is going
0:36:20 yes
0:36:21 so
0:36:22 I mean
0:36:22 but think
0:36:22 about it
0:36:23 right
0:36:23 if
0:36:23 if
0:36:24 you’re a
0:36:24 the only
0:36:24 reason
0:36:25 you need
0:36:25 to
0:36:26 make the
0:36:26 vehicle
0:36:27 as big
0:36:27 as
0:36:27 possible
0:36:27 is if
0:36:28 you have
0:36:28 to have a
0:36:28 dude
0:36:29 on every
0:36:29 one of them
0:36:29 that’s what
0:36:30 you’re saying
0:36:30 it’s not
0:36:30 necessarily
0:36:31 it could be
0:36:31 like
0:36:32 like the way
0:36:32 it’s like
0:36:33 a drone
0:36:33 swarm
0:36:33 you could do
0:36:34 a drone
0:36:34 swarm
0:36:35 but for
0:36:35 tractors
0:36:36 broad acre
0:36:36 swarming
0:36:37 exactly
0:36:37 right
0:36:38 and
0:36:39 that’s
0:36:40 there’s
0:36:41 some
0:36:41 additional
0:36:41 benefits
0:36:42 right
0:36:42 so
0:36:42 instead
0:36:42 of
0:36:42 spending
0:36:43 a million
0:36:43 dollars
0:36:44 you can do
0:36:45 more work
0:36:45 with five
0:36:46 tractors
0:36:46 that cost
0:36:47 less than
0:36:47 that
0:36:47 with their
0:36:48 implements
0:36:48 as well
0:36:50 and now
0:36:50 you have
0:36:51 some
0:36:52 additional
0:36:52 benefits
0:36:53 so you
0:36:54 actually
0:36:54 get more
0:36:55 work done
0:36:55 faster
0:36:56 and if
0:36:57 one of the
0:36:57 tractors
0:36:57 breaks down
0:36:58 you have
0:36:59 redundancy
0:37:00 right
0:37:01 if your
0:37:01 big
0:37:02 million
0:37:02 dollar
0:37:02 tractor
0:37:02 breaks
0:37:03 down
0:37:04 you got
0:37:04 issues
0:37:05 it’s
0:37:05 like
0:37:05 thousands
0:37:06 of
0:37:06 dollars
0:37:06 a minute
0:37:07 that
0:37:07 you know
0:37:08 that it’s
0:37:08 not running
0:37:09 so I think
0:37:10 there’s a lot
0:37:11 of opportunity
0:37:12 in agriculture
0:37:13 that we hadn’t
0:37:14 thought about
0:37:15 once automation
0:37:16 is achieved
0:37:16 I mean I
0:37:17 just I
0:37:17 described the
0:37:18 herbicide
0:37:19 challenge in
0:37:19 permanent
0:37:19 crops
0:37:20 I believe
0:37:21 that is
0:37:21 completely
0:37:22 can be
0:37:23 completely
0:37:23 eliminated
0:37:24 with
0:37:25 automation
0:37:26 and you
0:37:27 know computer
0:37:27 controls
0:37:29 computer vision
0:37:29 like you just
0:37:30 pick the weeds
0:37:30 instead if
0:37:31 you have
0:37:31 essentially
0:37:32 very cheap
0:37:33 robots that
0:37:33 are very
0:37:34 dexterous
0:37:34 you just
0:37:35 pick the
0:37:35 weeds instead
0:37:35 of spraying
0:37:36 them
0:37:36 absolutely
0:37:37 it’s already
0:37:38 happening in
0:37:39 row crops
0:37:39 you know you
0:37:39 look at
0:37:40 carbon robotics
0:37:41 they’re using
0:37:41 lasers to
0:37:42 kill weeds
0:37:43 you know in
0:37:44 lettuce fields
0:37:45 and other
0:37:45 kinds of
0:37:46 crops that
0:37:47 are called
0:37:47 specialty
0:37:48 crops
0:37:49 that’s huge
0:37:50 right because
0:37:51 before humans
0:37:51 would go in
0:37:52 there or
0:37:52 you’d spray
0:37:53 and use
0:37:54 ag inputs
0:37:55 or herbicides
0:37:56 so I think
0:37:56 there’s a
0:37:57 huge
0:37:59 there is a
0:37:59 huge inflection
0:38:00 point but
0:38:00 I just I
0:38:01 want to make
0:38:01 it clear to
0:38:02 your listeners
0:38:03 that just the
0:38:04 core base of
0:38:05 you know being
0:38:06 able to produce
0:38:06 enough food for
0:38:07 the population
0:38:08 we need
0:38:09 automation for
0:38:10 there are going
0:38:10 to be these
0:38:11 other benefits
0:38:11 some of which
0:38:12 we don’t even
0:38:12 know yet
0:38:15 so I’m such a
0:38:16 big believer in
0:38:17 automation it’s
0:38:17 why I brought my
0:38:18 two worlds
0:38:18 together and it’s
0:38:19 why I’ve really
0:38:20 dedicated my life
0:38:21 to this mission
0:38:23 and I just I’ve
0:38:23 never been as
0:38:24 excited as I
0:38:25 as I am now
0:38:26 in terms of
0:38:27 the impact
0:38:28 that this
0:38:29 company can
0:38:30 have our
0:38:31 share of
0:38:32 impact but
0:38:33 all of us
0:38:33 together there’s
0:38:34 so much
0:38:35 opportunity here
0:38:36 and you know
0:38:37 I think the
0:38:38 incumbents are
0:38:38 the ones that
0:38:39 the incumbent
0:38:40 equipment manufacturers
0:38:41 are the ones
0:38:41 that are really
0:38:42 going to make
0:38:43 it happen in
0:38:43 terms of
0:38:44 distribution and
0:38:45 support and
0:38:46 you know getting
0:38:47 it out there to
0:38:48 the world so
0:38:49 couldn’t be more
0:38:50 excited
0:38:54 we’ll be back
0:38:54 in a minute
0:38:55 with the
0:38:55 lightning round
0:39:06 run a business
0:39:07 and not thinking
0:39:08 about podcasting
0:39:08 think again
0:39:09 more Americans
0:39:10 listen to podcasts
0:39:11 than ad supported
0:39:12 streaming music
0:39:13 from Spotify and
0:39:14 Pandora and as
0:39:14 the number one
0:39:15 podcaster I
0:39:16 heart’s twice as
0:39:17 large as the next
0:39:18 two combined so
0:39:18 whatever your
0:39:19 customers listen to
0:39:20 they’ll hear
0:39:21 your message
0:39:21 plus only
0:39:22 I heart can
0:39:23 extend your
0:39:23 message to
0:39:24 audiences across
0:39:25 broadcast radio
0:39:26 think podcasting
0:39:26 can help your
0:39:27 business think
0:39:29 I heart streaming
0:39:30 radio and
0:39:31 podcasting let us
0:39:31 show you at
0:39:32 I heart advertising
0:39:33 dot com that’s
0:39:34 I heart advertising
0:39:35 dot com
0:39:37 got a business
0:39:38 problem there’s a
0:39:39 TED talk for that
0:39:40 stay updated on
0:39:41 everything business
0:39:42 on TED business a
0:39:43 podcast hosted by
0:39:44 Columbia business
0:39:45 school professor
0:39:46 Modupe Akinola
0:39:47 every week she’ll
0:39:48 introduce you to
0:39:49 leaders with unique
0:39:50 insights on work
0:39:51 answering questions
0:39:52 like how do
0:39:53 four-day work
0:39:54 weeks work to
0:39:55 will a machine
0:39:56 ever take my
0:39:57 job get some
0:39:58 surprising answers
0:39:59 on TED business
0:40:00 wherever you listen
0:40:00 to podcasts
0:40:04 let’s finish with
0:40:04 the lightning round
0:40:06 I heard you say
0:40:07 on another interview
0:40:08 that you are not a
0:40:09 big fan of IPOs
0:40:10 and that you
0:40:11 generally prefer
0:40:12 acquisitions why
0:40:13 what do you got
0:40:14 against IPOs
0:40:16 well that’s from the
0:40:17 old days I went
0:40:18 through a couple
0:40:19 of IPOs and
0:40:20 just about killed
0:40:22 me so a lot of
0:40:23 people think you
0:40:23 know IPOs are
0:40:24 glamorous and
0:40:25 everything but you
0:40:25 know the road
0:40:26 shows are pretty
0:40:27 intense and the
0:40:29 pressures are really
0:40:31 intense and you
0:40:32 know I’m not just a
0:40:33 fan of acquisitions I
0:40:34 mean I think you’re
0:40:35 seeing a lot of
0:40:36 differences in the
0:40:37 private market than
0:40:38 you have in the
0:40:39 past companies like
0:40:40 open AI who are
0:40:41 raising billions of
0:40:42 dollars in the
0:40:44 private market and
0:40:45 not having to go
0:40:46 public you just
0:40:47 stay private forever
0:40:48 you could just be a
0:40:49 company that makes
0:40:50 more money than it
0:40:50 spends you could do
0:40:51 that for as long as
0:40:53 you want exactly
0:40:55 exactly so I think
0:40:57 the IPO you know
0:40:59 I mean just look at
0:41:00 the last few years
0:41:01 of IPOs there just
0:41:01 haven’t been that
0:41:03 many and so there
0:41:04 are other ways to
0:41:05 build incredibly
0:41:06 valuable companies I
0:41:07 get asked all the
0:41:08 time what’s your
0:41:10 exit strategy and I
0:41:10 always have the
0:41:12 same answer I
0:41:13 only focus on
0:41:13 building value
0:41:15 opportunities come
0:41:16 and I think
0:41:17 entrepreneurs who
0:41:19 build companies for
0:41:21 an exit aren’t
0:41:21 building the company
0:41:22 for the right reasons
0:41:23 and aren’t building
0:41:24 really valuable
0:41:25 companies either
0:41:27 what’s one thing
0:41:28 that Steve Jobs
0:41:29 told you that stuck
0:41:29 with you
0:41:33 I had the pleasure
0:41:34 of working for
0:41:35 some incredible
0:41:36 entrepreneurs Steve
0:41:36 Jobs being one of
0:41:38 them and what I
0:41:40 did learn is that
0:41:41 these incredible
0:41:42 entrepreneurs they
0:41:43 each had like a
0:41:45 singular mantra and
0:41:46 for Steve it was all
0:41:47 about design it’s all
0:41:49 about design and
0:41:50 there were things he
0:41:51 would push for that
0:41:52 we’d sit there and
0:41:53 scratch our heads and
0:41:54 go wait that’s going
0:41:55 to add like a hundred
0:41:56 dollars cost to this
0:41:59 and and we would do
0:42:01 it and the product
0:42:02 would be successful and
0:42:03 we’re like okay he
0:42:05 was right you know I
0:42:07 I learned from from Bill
0:42:09 Gates that it’s all
0:42:11 about software and you
0:42:12 know what he was
0:42:12 right I learned from
0:42:13 Michael Dell that it’s
0:42:14 all about cost and
0:42:15 guess what he was
0:42:17 right so I think you
0:42:18 know the every
0:42:20 entrepreneur that makes
0:42:22 serious impact you know
0:42:24 has these kind of core
0:42:26 mantras so what’s
0:42:28 yours I view it as a
0:42:29 combination but there
0:42:30 is one thing that I
0:42:31 will tell you very very
0:42:33 clearly and your
0:42:35 listeners and I tell
0:42:37 my teammates probably
0:42:38 every day if not every
0:42:41 week you know it’s all
0:42:43 about show me don’t
0:42:45 tell me and that
0:42:48 particularly applies to
0:42:49 these industrial markets
0:42:52 the best way to sell
0:42:54 industrial equipment and
0:42:55 this is what dealers do
0:42:58 is you know here Susie
0:42:59 Farmer take this piece
0:43:00 of equipment and use it
0:43:02 for a day and she’ll
0:43:03 take that piece of
0:43:04 equipment use it on her
0:43:05 ranch on her farm
0:43:07 whatever and inevitably
0:43:08 she will buy it because
0:43:10 she gets to try you know
0:43:12 before so I think I think
0:43:13 the show me mentality is
0:43:14 really really important in
0:43:16 this era and in ag tech
0:43:17 in particular because
0:43:19 there have been some
0:43:20 companies that have been
0:43:22 developed by pure tech
0:43:24 people and pure tech
0:43:25 people you know telling
0:43:27 farmers that you know we
0:43:28 can farm better than you
0:43:30 is not a good recipe for
0:43:32 success right and so by
0:43:35 focusing on show me it’s
0:43:37 really what resonates with
0:43:39 our customers it’s really
0:43:42 what our customers need and
0:43:43 it’s just how business is
0:43:44 done in these industrial
0:43:48 worlds so that’s my mantra
0:43:49 show me don’t tell me
0:43:53 what’s one ridiculous word
0:43:54 somebody wants you is to
0:43:56 describe the wine that you
0:43:56 grow
0:44:05 I didn’t even know what
0:44:06 this I still don’t know what
0:44:07 this word means but
0:44:10 ethereal I’m like
0:44:10 ethereal
0:44:12 I’m like what does it what
0:44:13 does that mean it doesn’t
0:44:15 stick around it’s like I
0:44:16 don’t even know I drank it
0:44:17 like what yeah I don’t know
0:44:20 right so you get all kinds of
0:44:21 feedback is that a fancy way of
0:44:23 saying easy drinking sounds
0:44:24 like easy drinking to me
0:44:26 probably yeah yeah yeah
0:44:28 but uh you get all kinds of
0:44:29 great feedback and I mean
0:44:31 that’s the cool thing about
0:44:32 sort of the show me mentality
0:44:34 is when you develop products
0:44:37 that end users you know
0:44:40 consume or use and you see
0:44:41 their reaction and you know
0:44:44 they’re you you please them
0:44:45 you give them something of
0:44:47 value that’s that’s what gets
0:44:49 me motivated all the time and
0:44:51 and that’s why you know this
0:44:52 space is is so exciting for me
0:44:54 because that that impact is
0:44:55 felt like really really quickly
0:44:57 when when you do have a
0:44:58 solutions that work or when you
0:44:59 do make great wine or when you
0:45:00 do make great olive oil
0:45:03 um yeah I don’t think of
0:45:04 autonomous tractors as
0:45:05 ethereal
0:45:07 please don’t
0:45:17 tim booker is a farmer and the
0:45:18 founder and ceo of
0:45:21 agtonomy please email us at
0:45:23 problem at pushkin dot fm we
0:45:25 are always looking for new
0:45:27 guests for the show today’s
0:45:29 show was produced by trina
0:45:30 menino and gabriel hunter chang
0:45:32 it was edited by alexander
0:45:34 garrettin and engineered by
0:45:37 i’m jacob goldstein we’ll be
0:45:38 off for the next few weeks for
0:45:40 the holidays i want to thank
0:45:42 you very much for listening to
0:45:43 the show this year it really
0:45:45 means tremendous amount to all
0:45:46 of us i hope you have a great
0:45:48 holiday happy new year and we
0:45:50 will be back with more episodes
0:45:52 of what’s your problem in
0:45:53 2026
0:46:04 everybody has a hot take on the
0:46:05 economy and whether you’re
0:46:07 curious about inflation trade
0:46:09 wars or the markets what you need
0:46:11 is reporting you can trust hi
0:46:12 i’m kai rizdal the host of
0:46:14 marketplace our award-winning
0:46:15 reporters talk to everybody from
0:46:17 ceos to farmers to help you
0:46:18 understand how the economy takes
0:46:21 shape in the real world you’ll be
0:46:23 smarter every time you listen and
0:46:25 these days that’s priceless listen
0:46:26 to marketplace on your favorite
0:46:27 podcast app
0:46:32 this is an iheart podcast
0:46:34 guaranteed human