July 8, 2025 • 23 mins
Autonomous Agriculture and the Farm
John Reid, Center for Digital Agriculture
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Todd Gleason (00:00):
This is a special edition of the Closing Market
Report. I'm University ofIllinois Extension's Todd
Gleason out of the office forthe afternoon, so no update of
the commodity markets. Thefollowing presentation was made
by John Reed, the director forthe Center of Digital
(00:22):
Agriculture at the University ofIllinois during the March 4 All
Day I Got Look at the Beef Housein Covington, Indiana. Reed
spent nineteen years at JohnDeere and has recently returned
for his second stint at the U ofI.
John Reid (00:38):
So thank you so much and thanks for the kind intro.
And I I just want to kind ofstart out is that how many
people use automatic guidancetoday on your farming systems?
Just raise your hand. So not oh,keep your hands up. How many
people were using it fifteenyears ago?
(00:58):
Okay. Quite a bit fewer. Howmany people use a farm
management information system tomanage your machine operations
on your farm today? A few. Okay.
So think about that for thispresentation because a lot of
this is an evolution oftechnology and guidance. I used
(01:19):
to when first was working inautomatic guidance, I was
driving up and down the road,and you don't see very many of
the GPS receivers on vehicles.You kind of wonder if the
technology that's beingdeveloped is actually maturing
very quickly. But then you startseeing over time that there is
more broader adoption as there'svalue for this. I think the
(01:40):
thing I want to kind of leaveyou with on automatic with
autonomous vehicles is many ofthe examples we're seeing today
are extensions of all the thingsin guidance and all the other
types of technologies, includingthe farm management information
system, and then using thosetogether with some opportunities
(02:01):
to, in some situations, toremove the operator from the
machine.
So just to get started withthis, what are we seeing in
terms of autonomous vehicles?There's of press release, and we
live in an age of social media,so you see CES, for example,
John Deere announces their newextension. For several years
(02:22):
they've been talking about this.And last year they talked about
doing one production step,tillage, and showing how it
could be done autonomously,whereas this year they're much
broader vision. They're talkingabout how these production steps
could be more of them.
Okay? So that's heading in theright direction. If we're going
to have autonomous agricultureon a farm, can't just be used
(02:44):
for one niche thing and then allthe rest of the time, you don't
get utility out of it. So that'sgood. They actually, really
interesting with with JohnDeere, they're also seeing that
this farm management informationsystem, which they call op
center, links to other spaceseven.
So it's used for agriculturetoday, but if they're going to
do autonomy in construction orin golf courses, these other
(03:06):
places where there's labor andlots of machinery challenges,
then you need to have the samekind of thing. So what is it in
an autonomous tractor? First ofall, it's a highly automated
tractor to begin with. Itdoesn't have to be. Actually,
you'll see a lot of companiesthat will retrofit a vehicle by
putting autonomy kit on anexisting tractor, but if it's
(03:27):
highly automated, it's got moreelectronic surfaces that can be
controlled through software,makes it easier to integrate
with the autonomous vehicle.
Actually, we're talking aboutthe example in Champaign. That
was the first Magnum tractorfrom Case IH that had a CAN bus
on it, And for a researcher, itwas really a nice way of doing
(03:52):
autonomy because usually I hadto give the tractors back every
year to the company. I'd spendsix months making it into a
robot tractor and in threemonths doing work, well, with
the changes of havingelectronics backbone and
networks, it became a one weekjob of converting a tractor into
an autonomous vehicle. So you'vegot this highly automated
tractor. It's got an ability tosense what's happening in this
(04:16):
environment that's calledsituational awareness.
Most commonly, we think aboutGPS telling where the vehicle is
and how you could use that toplant straight or to do
operations more efficientlythrough automatic guidance, as
an example. But it has othersituational awareness that's
needed too, because the mostimportant part of a tractor
(04:37):
operating in the field is theoperator sitting in the seat and
sensing what's going on aroundthe vehicle and the implement
system, as well as understandingwhat's happening in the field,
where do you want to go. So it'sjust not an issue that nobody
does automatic tractoring orjust driving the tractor around
the field. You have animplement, you're doing a work
(04:59):
task, you're trying to get a jobdone, so there's extra sensors
needed. Today, we're starting tosee cameras and other types of
sensing like that that arelooking out ahead of the vehicle
to look for people, to sense theroads, to and then also those
cameras are looking around thevehicle and seeing the implement
and trying to understand is theimplement doing its job.
(05:21):
And then as I mentioned, you seethis with some OEMs coming out.
Nearly every OEM is talkingabout autonomy today and how
they're exploring it, gettinginvolved in it. You don't see
very many price tags for it yetbecause a lot of this is
innovation that is stillemerging and is still evolving,
(05:42):
and I think it's almost likeworking with individual
customers to learn is part ofwhat many companies are doing.
However, there are companiesthat are selling retrofit
autonomy. Sabanto is one out ofthe Chicago area that will take
any color vehicle and basicallyadd an autonomy kit to it, and
then they have a managementsystem that helps run those run
(06:03):
that vehicle autonomously and itcould be remotely monitored and
managed.
What's really interesting inthis space, though, is really
not autonomy of the driving, butautonomy of the implement
itself. If you've heard ofprecision spraying technology
like ExactApply from John Deereor other see and spray types of
(06:25):
technologies, these technologiesare putting the AI and the
intelligence on the implementand allowing the implement to
work at a level that's beyondwhat has been possible in the
past, being able to veryprecisely spray individual weeds
and save chemical and get thosebenefits, but then still have an
operator in his seat. And Ithink this is really kind of an
(06:47):
important point about this isthat if we are going to be doing
these operations, we needimplements that can be able to
sense things that are workingand not working that today
depend on the operator or thefarmer in the seat to detect. So
think about a debris on a plowshank or some types of effects,
(07:11):
something breaking. Right now,lot of the implements are not
instrumented to be able to tellthat something's wrong.
So if you really are going tohave effective autonomy, you
don't want to kind of come backto the field where a tractor
just finished operating andseeing that it hadn't completed
its operations. So anyhow, thisintelligent implement side is
(07:32):
not really autonomy, but it isan element that also has to
emerge if autonomous agricultureis going to take place. So
again, we have the tractor, wehave guidance systems, have
situational awareness orperception on the vehicle. And
then usually, these are tiedinto some kind of farm
(07:52):
management information systemfor autonomy. That system is
used to essentially understandthe fields where you're going to
operate, understand what kind ofpath plans and AI is used to
generate those plans, to cover afield.
For some operations liketillage, that's all you need.
You basically can designatethat, the operator, the farmer,
(08:16):
moves the vehicle to the field,and then you can get out and
execute the task. I want to comeback to this because we're
seeing a lot of the visualvideos of demonstrating the task
working, but when you reallythink about this, a lot of
operations, like your plantingor anything that has spring,
anything that has materials andinputs that you have to manage,
(08:40):
there's a lot of work besidesjust the execution of the
planting or spraying task. Haveto consider how do you resupply.
And actually, some cases,there's a robotics company
that's looking at automating thematerial logistics and moving
(09:01):
materials to and fro, whichseems like also a good idea and
be really important for a fullyautonomous system.
So you have these operationstaking place, and I want to kind
talk about the jobs to be done.People are showing autonomy in
terms of, okay, I'm going to getout of the tractor, hit go, and
this vehicle is going to performits tillage operation over a
(09:24):
field. But some of the bigchallenges are getting to that
point of being able to hit go inrunning that application. So,
for example, if you start atyour farm site and your field is
some distance away, you can'tautonomously go on the highways
today. It's not possible.
And in fact, I think it's areally hard challenge. I don't
(09:45):
expect to see that. Even inautomotive, we're seeing that
autonomy isn't where it needs tobe. And actually, in
agriculture, I think it's gotmore potential. But you have to
move the equipment to the field.
And maybe you have materials andthings to set up the job that
are done. Then actually, whenyou're running the job, you're
(10:06):
monitoring the task, and theseare things that the farm
management information systemscan do. They can essentially
load data that understands whatoperation took place, how
accurate was it done. And as aslong as you don't have any
materials to supply to thevehicle, like seed to a planter
or or fertilizer, then then, youknow, it's probably okay for
(10:29):
those types of operations. Butif you need to logistically do
those things, then somebodyneeds to be involved or
additional autonomy is needed tomake those things happen.
And then after the job is done,tractor finishes, stops, shuts
down, well, somebody has to gofetch it and bring it back and
clean up the operation. So allthe things that you have to do
when you're doing the operation,really, we're only talking about
(10:52):
one piece of it where thevehicle is driving autonomously
and doing these operations. Soit's really very, very possible
to do it today. The benefits ofit are still somewhat emerging.
And there are places where thereare benefits, but I just want to
kind of point out that this isstill kind of the first inning
(11:12):
of a long game, and we're at thevery beginning of seeing what's
possible.
Like guidance at the verybeginning, only a few people,
lead adopters, were adopting it.I'd expect this is even more
challenging because of theaccumulation of technologies
that are needed to make autonomyhappen, and then some of these
(11:33):
other operations to get the jobdone, which is part of the
complexity of that task. So somegood news, though, are there are
probably there are many caseswhich could be compelling
examples of autonomy. One ofthem, for example, you see with
combine harvesting operationswhere you have a tractor and a
(11:55):
green cart that's moving betweenthe combine and the edge of the
field to transport materials.Can that be made autonomous?
Because you're kind of in anopen field, there's kind of a
fixed location, except for thecombine, which is always moving,
and autonomy could be a way ofhelping a single operator be
able to harvest, and thelogistics management, at least
(12:16):
to the edge of the field, can behandled. And that's going to be
an example where theseaccumulation of technologies can
come together to provide thattype of opportunity. In spring,
chemicals in orchards, and Iknow that probably most of you
aren't dealing with specialtycrops, but in those kinds of
operations, you may not want tobe on the vehicle when you're
spraying because these chemicalshave a long period of time
(12:39):
before you can reenter theorchard, so why not get the
operator off the machineentirely? So where do these
examples pay off? I'll give acouple of examples for you.
It tends to be on, for many ofthe applications, where there's
(12:59):
a lot of already labor involvedand autonomy is a way of getting
better performance out outputout of labor that could be low
skilled and displace some ofthat. So one example, there's a
study by the Western growersabout a system called carbon
robotics. Autonomous It's notsystem, but it's a very
(13:23):
intelligent laser weed controlsystem. Very expensive. It's
about a million dollars just forthe implement.
But in those operations whereyou have three crews of 25
people weeding in the field,this kind of system has a
payback and can provide value inthat particular scenario. So,
(13:45):
again, that's one scenario.Another example that I worked on
in the past was working withcitrus operations in Florida,
where they have lots of acreage,they're spraying chemicals or
they're mowing, and they havefleets of machines that are
doing this all the time. And wewere able to show that we could
(14:07):
take the drivers out of three tofive machines and train a new
kind of skill level, somebodythat was a mission manager that
could sit almost like a securityguard, either in a truck or in a
remote air conditioned office.And they could monitor these
systems working, from our dataworking with these operators for
(14:31):
over eighteen months, theyachieved 30% more productivity
than people that were drivingthe machines.
And part of that was because,through planning, the orchards
are somewhat confusing andthey're not all just perfectly
straight rows, so they couldavoid passing through places
twice. That was one improvementin productivity. The other thing
(14:52):
was that a human operator inthese machines, with trees being
close, would only drive acertain speed, but it was safer
to drive faster with theautonomy system and it wasn't
more like in a cab. You would benervous, perhaps, driving that
speed, but with autonomy it wasvisibly and actually performed
safely in that way. So what'scoming with these technologies
(15:18):
is that, as I said, I think it'sa long journey.
We're seeing a lot of theinitial things like we asked in
the guidance questions. Thereare a few early adopters that
are trying these technologies,think we're going to see moving
through the hype as industrygets more experienced. We're
going to see more and moreexamples of this. It is not
(15:40):
going to be for everyone. Infact, I would say if you are
interested in autonomy, reallythe journey is still start with
automation and get the value ofthe productivity of automatic
guidance and some of these othertype of things.
And integrating into farmmanagement information systems
is kind of a next step thatleads to this. As I said, twenty
(16:03):
five years ago when I waslooking at guidance, actually
similar to what your answerswere here. Fairly small adoption
rate that grew over time tohigher numbers as these systems
matured and the costs came down.Autonomy will be more complex
because it's like a suite oftechnologies that integrate
(16:26):
together, and yes, thetechnology readiness is there.
We're still working on theviability and understanding the
business model, and there issome customer value on certain
types of applications in termsof especially around labor
productivity and performance.
So I'll just stop there and usethe time for questions, if
(16:50):
anyone has any. Yeah. It's it'scertainly putting sensors in the
ground, in the soil would besomething that many many have
looked at. And so I I don't knowthat it's commercially available
yet, but, you know, I know thereis things like a John Deere has
something called exact shotthat's spraying a little bit of
(17:11):
chemicals on the seed atplanting time with their with
their high precision exactemerge. But, yeah, I think
that's types of devices that weuse today for connectivity,
sensing and things like that.
Can they be integrated into theground engaging parts? It's
certainly a possibility.
Todd Gleason (17:31):
So the question was whether or not you could
soil test as you were plantingor doing some other operation
simultaneously. Other questionsabout autonomous agriculture
that might be on your farm. SoI've got a series of questions
for you. The university, to thispoint, and we started with
autonomous ag, with largetractors, have moved to, smaller
(17:56):
vehicles, usually robots. What'sthe in between or what what do
you see?
John Reid (18:03):
Yeah, actually, I'm you know, if you work for a
large OEM, and I did for twentyyears, you kind of see the
productivity we get out of thesemachines that last hundred years
wider, bigger, faster, highperformance. Adding autonomy to
those machines is an element oftrying to achieve increased
(18:24):
productivity with do more withless, like what John Deere says.
On the other end of the spectrumare small robots that are going
down between the rows. And, youknow, at my center you know, I
came into the center a coupleyears ago, but they've talked
about this a lot. And they'reshowing things like cover
cropping.
But, you know, think about theenergy and and power
(18:45):
requirements of agriculture. Ithink we all go hungry if we
just depended on small smallrobots to to do some of those
types of things. So they dounique things, like if you
wanted to plant seed in Augustbetween corn so that you start a
cover crop just before harvest,Definitely doable. Can it
connect collect data? Yes.
(19:06):
Drones also can collect collectdata.
Todd Gleason (19:08):
My next question. Yeah. What what what are your
thoughts about drone technology?
John Reid (19:11):
Some some of it is, one of my my greatest
experiences is for a period oftime, I had to work with
advanced marketing people andput technology and marketing
together. And, everytechnologist has a hammer. So if
you're doing the small robot oryou're doing the big robot, you
know, you wanna do that. Andthat's what you know. And, But I
think the reality is you have tounderstand what's the
(19:33):
effectiveness of the solution,what does it cost, how much work
does it take to get it done.
And think there's just a varietyof options that we have today,
including remote sensing. I'veused remote sensing services
that are available today thatgive me a prediction of yield on
a field that are are better thanthe yield monitor on a combine.
(19:55):
So so, you know, that's that'sactually easy. You can pay a
subscription fee to give themyour farm. I mean, you just buy
it for your own area of land,and and you have a layer of
information that can help manageto how the combine operates more
efficiently.
Again, we have to be a littleagnostic to the specific version
(20:15):
of technology. Now back to thispoint, I think there are going
to be some operations. Lymphtechis a company that shows a solar
powered sprayer. And they'veeven I talked about this idea of
tendering. They even show thatthey're building a unit that it
can park and tender.
And this this just runs reallyslow because it's solar powered,
(20:37):
but it runs all the time. So theidea of being able to do perhaps
some levels of weed managementin certain operations and not
have to touch it, if you reallydon't have to touch it, then
that's potentially valuable forthe farmer to consider. On the
other hand, if you have to spenda lot of time and resources to
get everything set up and useit, and then you're going to
(20:59):
stand there in the field andwatch it while it's operating,
well, that doesn't feel like itimproved your productivity very
much. It's kind of a noveltythat's pretty cool to see. So we
have to kind of get past thoseearly learning phases to where
you can trust, have trust inautomation, and you don't have
to sit there and babysit it.
Todd Gleason (21:18):
One of your colleagues, when you were still
at Deere, was here a couple ofyears ago talking about, and he
headed up the combine section,was talking about machine
learning versus AI. That's inyour wheelhouse.
John Reid (21:30):
Yep.
Todd Gleason (21:30):
What do you think about the difference between the
two, if there is one?
John Reid (21:34):
Well, learning is a form of AI, but one of the
reasons I'm really enjoyingbeing back in university is the
power of AI is getting cheaperand it's more capable today and
can do some really interestingthings. My team is working on
things like embodied AI that cango inside the machine and
(21:54):
collaborate with the farmer interms of advice and ask
questions, document information,so you don't have to hit the
display, but you can just haveinformation automatically
ingested. We actually have an AItool that's a large language
model called Crop Wizard, andthere'll be a big announcement
on this this week. Working witha number of companies where it's
(22:15):
kind of an expert agronomist.It's been trained on USDA data
and agronomy.
You can all go find it atuiuc.chat, and it's one of the
programs that's listed there.It's also in the early phases of
development, but it's based byfactual information that
scientists have collected, andit ends up being a quick way of
(22:37):
answering questions and gettinga perspective. Now, when you get
the answer, you still have touse your brain. You have to
understand, like everythingelse, is it telling me something
that factually true? Look at thereferences, but it's kind of a
nice aid that could beintegrated in the machinery as
well as used in the farm site.
Todd Gleason (22:54):
Yeah. So essentially, you can go to
uiuc.chat.
John Reid (22:57):
Chat. C h a t.
Todd Gleason (22:59):
Yeah. Do you have to put the e d u on the end? No.
John Reid (23:01):
Just uiuc.chat.
Todd Gleason (23:05):
Yeah. And then it's like ChatGPT or any of the
others. So you can ask itquestions. It's Crop Wizard, so
you can ask it questions aboutwhatever as it's related to
crops, and it will go out andfind within a defined set of
good facts an answer for you.Yes.
(23:25):
So try that out. That would beinteresting. Yes. I have a
couple of other questions. So onautonomous agriculture, I don't
think that your colleague saidupfront, but I know he thinks
that the platform, the combinesitself, can do a better job of
setting the combine.
John Reid (23:47):
Did you have Don Pfeiffer here by any chance? No.
I'm trying to figure
Todd Gleason (23:52):
does anybody remember the guy's name? I don't
remember his name at the moment,but no, that one doesn't ring a
bell.
John Reid (23:57):
Actually, well, this is this is actually my one of my
last projects at Deere. Weworked on combine automation.
Actually, the X nine, has a apackage on it with forward
looking perception that sees thecrop and gives information about
what's coming at the combine.Actually, you know, the thing is
this, is that we we documentedover 500,000 acres around the
(24:19):
world of people harvestingdifferent crops, and we found if
you gave them a zero to a 100%score, the best operators were
only performing in the 80%range, and that was only for a
few hours a day. And actually,it was more common to see that,
especially large operations thathad lots of unskilled labor,
(24:40):
that they're operating in the50% range.
In some cases, they were settingup the combine for one crop,
using it for a different one,not changing it, and the
performance was because youweren't getting maximum out of
the machine itself. So that's anarea where you don't even have
to automatically control it.Combine harvesting changes with
the temperatures during the day.It changes with moisture
(25:01):
content, weather. It changeswith topography, uphill,
downhill.
How many people here, whenyou're busy harvesting, take
somebody in your family and putthem in a cab and say, set the
settings here and don't touchanything? How many people do
that?
Todd Gleason (25:17):
Raise your hands.
John Reid (25:20):
I mean and so, actually, we had a lot of fleet
customers that told us theydidn't care about fuel
consumption, but with thedigital information that's on
the on the machines today, oncewe showed them the significant
dollar that they were spendingon harvesting crop and losing
some of it out the back of themachine, or burning too much
(25:42):
fuel to get that same type ofperformance, then it is a real
interesting area ofoptimization. And frankly, I
think it's probably a precursorto autonomy. If I can get that
kind of performance and reallyoptimize a machine, then I'm
ready to talk about stepping outof the cab because it doesn't
need me so much.
Todd Gleason (26:01):
Any other questions? Yes, sir? Oh, I
didn't catch that. Can yourepeat
John Reid (26:08):
Is it actually you're asking about having a service
model? Yeah. I think it's reallyinteresting because when you
think about the machinesthemselves, significant
investment. The autonomy,especially in the early phases,
is going to need a lot of handson or observations by the
company, the perception systems,are they working, is it the same
(26:32):
as it was last year? So having amodel that's similar to
satellite TV, you buy your TVbut you buy the satellite
service, that kind ofsubscription model could work.
I don't see very many of theOEMs or even the small players.
Some of them may be going thatway. I think there's value in
(26:55):
them being able to see how thefleet of machines work as long
compromising the farmer's dataso that they can make those
systems better by the learningthey get between systems. So in
the early phases particularly, Iactually believe in the long
run, services model might bebetter than thinking about a
farmer who's going to basicallyset this up and stand in the
(27:17):
field. I'd rather have a serviceprovider that train community
college students to understandhow to run these systems, and
they can afford the cost of theassets and they deploy it by
serving lots of farmers inparticular areas.
It's not a bad model.
Todd Gleason (27:34):
Give John Reed a nice round of applause. John, if
you could take that to NeilDahlstrom. John Reed is the
director for the Center forDigital Agriculture here on the
Urbana Champaign campus of theUniversity of Illinois. You've
been listening to excerpts fromhis presentation made during the
March all day ag outlook at theBeef House in Covington, Indiana
(27:55):
on this agricultural programmingfrom Illinois Public Media. I'm
(28:20):
University of IllinoisExtension's Todd Gleason out of
the office this week.
Here's an update of what'shappening from Extension across
the state next week. Starting onthe fifteenth, there is webinar
hosted by the Farm Doc team.I'll be the emcee for that event
from noon to one that's onTuesday related to the one big
(28:41):
beautiful bill and the farmpolicy changes it has in acted
related to commodity programsand crop insurance farmers
landowners and bankers all needto be in attendance please
register now you can do that onour website at wilag.org
willag.org Look in the calendarof events on Tuesday, July.
(29:06):
There'll be some other thingsyou can register for as well. On
the sixteenth of next week, theOre Agricultural Research and
Demonstration Center field daywill take place.
Agronomist will be on hand inBayless. That's in Western
Illinois to the West ofJacksonville. The Monmouth field
day is the following week onJuly. They'll follow that up
(29:29):
with the July 24 EwingDemonstration Center field day
and then wrap things up here oncampus with the Hemp Research
Open House. That one is on Julyand will require registration.
You can find all thatinformation on the PharmDoc
Daily website or willag.org.
This is a special edition of the Closing Market
Report. I'm University ofIllinois Extension's Todd
Gleason out of the office forthe afternoon, so no update of
the commodity markets. Thefollowing presentation was made
by John Reed, the director forthe Center of Digital
(00:22):
Agriculture at the University ofIllinois during the March 4 All
Day I Got Look at the Beef Housein Covington, Indiana. Reed
spent nineteen years at JohnDeere and has recently returned
for his second stint at the U ofI.
John Reid (00:38):
So thank you so much and thanks for the kind intro.
And I I just want to kind ofstart out is that how many
people use automatic guidancetoday on your farming systems?
Just raise your hand. So not oh,keep your hands up. How many
people were using it fifteenyears ago?
(00:58):
Okay. Quite a bit fewer. Howmany people use a farm
management information system tomanage your machine operations
on your farm today? A few. Okay.
So think about that for thispresentation because a lot of
this is an evolution oftechnology and guidance. I used
(01:19):
to when first was working inautomatic guidance, I was
driving up and down the road,and you don't see very many of
the GPS receivers on vehicles.You kind of wonder if the
technology that's beingdeveloped is actually maturing
very quickly. But then you startseeing over time that there is
more broader adoption as there'svalue for this. I think the
(01:40):
thing I want to kind of leaveyou with on automatic with
autonomous vehicles is many ofthe examples we're seeing today
are extensions of all the thingsin guidance and all the other
types of technologies, includingthe farm management information
system, and then using thosetogether with some opportunities
(02:01):
to, in some situations, toremove the operator from the
machine.
So just to get started withthis, what are we seeing in
terms of autonomous vehicles?There's of press release, and we
live in an age of social media,so you see CES, for example,
John Deere announces their newextension. For several years
(02:22):
they've been talking about this.And last year they talked about
doing one production step,tillage, and showing how it
could be done autonomously,whereas this year they're much
broader vision. They're talkingabout how these production steps
could be more of them.
Okay? So that's heading in theright direction. If we're going
to have autonomous agricultureon a farm, can't just be used
(02:44):
for one niche thing and then allthe rest of the time, you don't
get utility out of it. So that'sgood. They actually, really
interesting with with JohnDeere, they're also seeing that
this farm management informationsystem, which they call op
center, links to other spaceseven.
So it's used for agriculturetoday, but if they're going to
do autonomy in construction orin golf courses, these other
(03:06):
places where there's labor andlots of machinery challenges,
then you need to have the samekind of thing. So what is it in
an autonomous tractor? First ofall, it's a highly automated
tractor to begin with. Itdoesn't have to be. Actually,
you'll see a lot of companiesthat will retrofit a vehicle by
putting autonomy kit on anexisting tractor, but if it's
(03:27):
highly automated, it's got moreelectronic surfaces that can be
controlled through software,makes it easier to integrate
with the autonomous vehicle.
Actually, we're talking aboutthe example in Champaign. That
was the first Magnum tractorfrom Case IH that had a CAN bus
on it, And for a researcher, itwas really a nice way of doing
(03:52):
autonomy because usually I hadto give the tractors back every
year to the company. I'd spendsix months making it into a
robot tractor and in threemonths doing work, well, with
the changes of havingelectronics backbone and
networks, it became a one weekjob of converting a tractor into
an autonomous vehicle. So you'vegot this highly automated
tractor. It's got an ability tosense what's happening in this
(04:16):
environment that's calledsituational awareness.
Most commonly, we think aboutGPS telling where the vehicle is
and how you could use that toplant straight or to do
operations more efficientlythrough automatic guidance, as
an example. But it has othersituational awareness that's
needed too, because the mostimportant part of a tractor
(04:37):
operating in the field is theoperator sitting in the seat and
sensing what's going on aroundthe vehicle and the implement
system, as well as understandingwhat's happening in the field,
where do you want to go. So it'sjust not an issue that nobody
does automatic tractoring orjust driving the tractor around
the field. You have animplement, you're doing a work
(04:59):
task, you're trying to get a jobdone, so there's extra sensors
needed. Today, we're starting tosee cameras and other types of
sensing like that that arelooking out ahead of the vehicle
to look for people, to sense theroads, to and then also those
cameras are looking around thevehicle and seeing the implement
and trying to understand is theimplement doing its job.
(05:21):
And then as I mentioned, you seethis with some OEMs coming out.
Nearly every OEM is talkingabout autonomy today and how
they're exploring it, gettinginvolved in it. You don't see
very many price tags for it yetbecause a lot of this is
innovation that is stillemerging and is still evolving,
(05:42):
and I think it's almost likeworking with individual
customers to learn is part ofwhat many companies are doing.
However, there are companiesthat are selling retrofit
autonomy. Sabanto is one out ofthe Chicago area that will take
any color vehicle and basicallyadd an autonomy kit to it, and
then they have a managementsystem that helps run those run
(06:03):
that vehicle autonomously and itcould be remotely monitored and
managed.
What's really interesting inthis space, though, is really
not autonomy of the driving, butautonomy of the implement
itself. If you've heard ofprecision spraying technology
like ExactApply from John Deereor other see and spray types of
(06:25):
technologies, these technologiesare putting the AI and the
intelligence on the implementand allowing the implement to
work at a level that's beyondwhat has been possible in the
past, being able to veryprecisely spray individual weeds
and save chemical and get thosebenefits, but then still have an
operator in his seat. And Ithink this is really kind of an
(06:47):
important point about this isthat if we are going to be doing
these operations, we needimplements that can be able to
sense things that are workingand not working that today
depend on the operator or thefarmer in the seat to detect. So
think about a debris on a plowshank or some types of effects,
(07:11):
something breaking. Right now,lot of the implements are not
instrumented to be able to tellthat something's wrong.
So if you really are going tohave effective autonomy, you
don't want to kind of come backto the field where a tractor
just finished operating andseeing that it hadn't completed
its operations. So anyhow, thisintelligent implement side is
(07:32):
not really autonomy, but it isan element that also has to
emerge if autonomous agricultureis going to take place. So
again, we have the tractor, wehave guidance systems, have
situational awareness orperception on the vehicle. And
then usually, these are tiedinto some kind of farm
(07:52):
management information systemfor autonomy. That system is
used to essentially understandthe fields where you're going to
operate, understand what kind ofpath plans and AI is used to
generate those plans, to cover afield.
For some operations liketillage, that's all you need.
You basically can designatethat, the operator, the farmer,
(08:16):
moves the vehicle to the field,and then you can get out and
execute the task. I want to comeback to this because we're
seeing a lot of the visualvideos of demonstrating the task
working, but when you reallythink about this, a lot of
operations, like your plantingor anything that has spring,
anything that has materials andinputs that you have to manage,
(08:40):
there's a lot of work besidesjust the execution of the
planting or spraying task. Haveto consider how do you resupply.
And actually, some cases,there's a robotics company
that's looking at automating thematerial logistics and moving
(09:01):
materials to and fro, whichseems like also a good idea and
be really important for a fullyautonomous system.
So you have these operationstaking place, and I want to kind
talk about the jobs to be done.People are showing autonomy in
terms of, okay, I'm going to getout of the tractor, hit go, and
this vehicle is going to performits tillage operation over a
(09:24):
field. But some of the bigchallenges are getting to that
point of being able to hit go inrunning that application. So,
for example, if you start atyour farm site and your field is
some distance away, you can'tautonomously go on the highways
today. It's not possible.
And in fact, I think it's areally hard challenge. I don't
(09:45):
expect to see that. Even inautomotive, we're seeing that
autonomy isn't where it needs tobe. And actually, in
agriculture, I think it's gotmore potential. But you have to
move the equipment to the field.
And maybe you have materials andthings to set up the job that
are done. Then actually, whenyou're running the job, you're
(10:06):
monitoring the task, and theseare things that the farm
management information systemscan do. They can essentially
load data that understands whatoperation took place, how
accurate was it done. And as aslong as you don't have any
materials to supply to thevehicle, like seed to a planter
or or fertilizer, then then, youknow, it's probably okay for
(10:29):
those types of operations. Butif you need to logistically do
those things, then somebodyneeds to be involved or
additional autonomy is needed tomake those things happen.
And then after the job is done,tractor finishes, stops, shuts
down, well, somebody has to gofetch it and bring it back and
clean up the operation. So allthe things that you have to do
when you're doing the operation,really, we're only talking about
(10:52):
one piece of it where thevehicle is driving autonomously
and doing these operations. Soit's really very, very possible
to do it today. The benefits ofit are still somewhat emerging.
And there are places where thereare benefits, but I just want to
kind of point out that this isstill kind of the first inning
(11:12):
of a long game, and we're at thevery beginning of seeing what's
possible.
Like guidance at the verybeginning, only a few people,
lead adopters, were adopting it.I'd expect this is even more
challenging because of theaccumulation of technologies
that are needed to make autonomyhappen, and then some of these
(11:33):
other operations to get the jobdone, which is part of the
complexity of that task. So somegood news, though, are there are
probably there are many caseswhich could be compelling
examples of autonomy. One ofthem, for example, you see with
combine harvesting operationswhere you have a tractor and a
(11:55):
green cart that's moving betweenthe combine and the edge of the
field to transport materials.Can that be made autonomous?
Because you're kind of in anopen field, there's kind of a
fixed location, except for thecombine, which is always moving,
and autonomy could be a way ofhelping a single operator be
able to harvest, and thelogistics management, at least
(12:16):
to the edge of the field, can behandled. And that's going to be
an example where theseaccumulation of technologies can
come together to provide thattype of opportunity. In spring,
chemicals in orchards, and Iknow that probably most of you
aren't dealing with specialtycrops, but in those kinds of
operations, you may not want tobe on the vehicle when you're
spraying because these chemicalshave a long period of time
(12:39):
before you can reenter theorchard, so why not get the
operator off the machineentirely? So where do these
examples pay off? I'll give acouple of examples for you.
It tends to be on, for many ofthe applications, where there's
(12:59):
a lot of already labor involvedand autonomy is a way of getting
better performance out outputout of labor that could be low
skilled and displace some ofthat. So one example, there's a
study by the Western growersabout a system called carbon
robotics. Autonomous It's notsystem, but it's a very
(13:23):
intelligent laser weed controlsystem. Very expensive. It's
about a million dollars just forthe implement.
But in those operations whereyou have three crews of 25
people weeding in the field,this kind of system has a
payback and can provide value inthat particular scenario. So,
(13:45):
again, that's one scenario.Another example that I worked on
in the past was working withcitrus operations in Florida,
where they have lots of acreage,they're spraying chemicals or
they're mowing, and they havefleets of machines that are
doing this all the time. And wewere able to show that we could
(14:07):
take the drivers out of three tofive machines and train a new
kind of skill level, somebodythat was a mission manager that
could sit almost like a securityguard, either in a truck or in a
remote air conditioned office.And they could monitor these
systems working, from our dataworking with these operators for
(14:31):
over eighteen months, theyachieved 30% more productivity
than people that were drivingthe machines.
And part of that was because,through planning, the orchards
are somewhat confusing andthey're not all just perfectly
straight rows, so they couldavoid passing through places
twice. That was one improvementin productivity. The other thing
(14:52):
was that a human operator inthese machines, with trees being
close, would only drive acertain speed, but it was safer
to drive faster with theautonomy system and it wasn't
more like in a cab. You would benervous, perhaps, driving that
speed, but with autonomy it wasvisibly and actually performed
safely in that way. So what'scoming with these technologies
(15:18):
is that, as I said, I think it'sa long journey.
We're seeing a lot of theinitial things like we asked in
the guidance questions. Thereare a few early adopters that
are trying these technologies,think we're going to see moving
through the hype as industrygets more experienced. We're
going to see more and moreexamples of this. It is not
(15:40):
going to be for everyone. Infact, I would say if you are
interested in autonomy, reallythe journey is still start with
automation and get the value ofthe productivity of automatic
guidance and some of these othertype of things.
And integrating into farmmanagement information systems
is kind of a next step thatleads to this. As I said, twenty
(16:03):
five years ago when I waslooking at guidance, actually
similar to what your answerswere here. Fairly small adoption
rate that grew over time tohigher numbers as these systems
matured and the costs came down.Autonomy will be more complex
because it's like a suite oftechnologies that integrate
(16:26):
together, and yes, thetechnology readiness is there.
We're still working on theviability and understanding the
business model, and there issome customer value on certain
types of applications in termsof especially around labor
productivity and performance.
So I'll just stop there and usethe time for questions, if
(16:50):
anyone has any. Yeah. It's it'scertainly putting sensors in the
ground, in the soil would besomething that many many have
looked at. And so I I don't knowthat it's commercially available
yet, but, you know, I know thereis things like a John Deere has
something called exact shotthat's spraying a little bit of
(17:11):
chemicals on the seed atplanting time with their with
their high precision exactemerge. But, yeah, I think
that's types of devices that weuse today for connectivity,
sensing and things like that.
Can they be integrated into theground engaging parts? It's
certainly a possibility.
Todd Gleason (17:31):
So the question was whether or not you could
soil test as you were plantingor doing some other operation
simultaneously. Other questionsabout autonomous agriculture
that might be on your farm. SoI've got a series of questions
for you. The university, to thispoint, and we started with
autonomous ag, with largetractors, have moved to, smaller
(17:56):
vehicles, usually robots. What'sthe in between or what what do
you see?
John Reid (18:03):
Yeah, actually, I'm you know, if you work for a
large OEM, and I did for twentyyears, you kind of see the
productivity we get out of thesemachines that last hundred years
wider, bigger, faster, highperformance. Adding autonomy to
those machines is an element oftrying to achieve increased
(18:24):
productivity with do more withless, like what John Deere says.
On the other end of the spectrumare small robots that are going
down between the rows. And, youknow, at my center you know, I
came into the center a coupleyears ago, but they've talked
about this a lot. And they'reshowing things like cover
cropping.
But, you know, think about theenergy and and power
(18:45):
requirements of agriculture. Ithink we all go hungry if we
just depended on small smallrobots to to do some of those
types of things. So they dounique things, like if you
wanted to plant seed in Augustbetween corn so that you start a
cover crop just before harvest,Definitely doable. Can it
connect collect data? Yes.
(19:06):
Drones also can collect collectdata.
Todd Gleason (19:08):
My next question. Yeah. What what what are your
thoughts about drone technology?
John Reid (19:11):
Some some of it is, one of my my greatest
experiences is for a period oftime, I had to work with
advanced marketing people andput technology and marketing
together. And, everytechnologist has a hammer. So if
you're doing the small robot oryou're doing the big robot, you
know, you wanna do that. Andthat's what you know. And, But I
think the reality is you have tounderstand what's the
(19:33):
effectiveness of the solution,what does it cost, how much work
does it take to get it done.
And think there's just a varietyof options that we have today,
including remote sensing. I'veused remote sensing services
that are available today thatgive me a prediction of yield on
a field that are are better thanthe yield monitor on a combine.
(19:55):
So so, you know, that's that'sactually easy. You can pay a
subscription fee to give themyour farm. I mean, you just buy
it for your own area of land,and and you have a layer of
information that can help manageto how the combine operates more
efficiently.
Again, we have to be a littleagnostic to the specific version
(20:15):
of technology. Now back to thispoint, I think there are going
to be some operations. Lymphtechis a company that shows a solar
powered sprayer. And they'veeven I talked about this idea of
tendering. They even show thatthey're building a unit that it
can park and tender.
And this this just runs reallyslow because it's solar powered,
(20:37):
but it runs all the time. So theidea of being able to do perhaps
some levels of weed managementin certain operations and not
have to touch it, if you reallydon't have to touch it, then
that's potentially valuable forthe farmer to consider. On the
other hand, if you have to spenda lot of time and resources to
get everything set up and useit, and then you're going to
(20:59):
stand there in the field andwatch it while it's operating,
well, that doesn't feel like itimproved your productivity very
much. It's kind of a noveltythat's pretty cool to see. So we
have to kind of get past thoseearly learning phases to where
you can trust, have trust inautomation, and you don't have
to sit there and babysit it.
Todd Gleason (21:18):
One of your colleagues, when you were still
at Deere, was here a couple ofyears ago talking about, and he
headed up the combine section,was talking about machine
learning versus AI. That's inyour wheelhouse.
John Reid (21:30):
Yep.
Todd Gleason (21:30):
What do you think about the difference between the
two, if there is one?
John Reid (21:34):
Well, learning is a form of AI, but one of the
reasons I'm really enjoyingbeing back in university is the
power of AI is getting cheaperand it's more capable today and
can do some really interestingthings. My team is working on
things like embodied AI that cango inside the machine and
(21:54):
collaborate with the farmer interms of advice and ask
questions, document information,so you don't have to hit the
display, but you can just haveinformation automatically
ingested. We actually have an AItool that's a large language
model called Crop Wizard, andthere'll be a big announcement
on this this week. Working witha number of companies where it's
(22:15):
kind of an expert agronomist.It's been trained on USDA data
and agronomy.
You can all go find it atuiuc.chat, and it's one of the
programs that's listed there.It's also in the early phases of
development, but it's based byfactual information that
scientists have collected, andit ends up being a quick way of
(22:37):
answering questions and gettinga perspective. Now, when you get
the answer, you still have touse your brain. You have to
understand, like everythingelse, is it telling me something
that factually true? Look at thereferences, but it's kind of a
nice aid that could beintegrated in the machinery as
well as used in the farm site.
Todd Gleason (22:54):
Yeah. So essentially, you can go to
uiuc.chat.
John Reid (22:57):
Chat. C h a t.
Todd Gleason (22:59):
Yeah. Do you have to put the e d u on the end? No.
John Reid (23:01):
Just uiuc.chat.
Todd Gleason (23:05):
Yeah. And then it's like ChatGPT or any of the
others. So you can ask itquestions. It's Crop Wizard, so
you can ask it questions aboutwhatever as it's related to
crops, and it will go out andfind within a defined set of
good facts an answer for you.Yes.
(23:25):
So try that out. That would beinteresting. Yes. I have a
couple of other questions. So onautonomous agriculture, I don't
think that your colleague saidupfront, but I know he thinks
that the platform, the combinesitself, can do a better job of
setting the combine.
John Reid (23:47):
Did you have Don Pfeiffer here by any chance? No.
I'm trying to figure
Todd Gleason (23:52):
does anybody remember the guy's name? I don't
remember his name at the moment,but no, that one doesn't ring a
bell.
John Reid (23:57):
Actually, well, this is this is actually my one of my
last projects at Deere. Weworked on combine automation.
Actually, the X nine, has a apackage on it with forward
looking perception that sees thecrop and gives information about
what's coming at the combine.Actually, you know, the thing is
this, is that we we documentedover 500,000 acres around the
(24:19):
world of people harvestingdifferent crops, and we found if
you gave them a zero to a 100%score, the best operators were
only performing in the 80%range, and that was only for a
few hours a day. And actually,it was more common to see that,
especially large operations thathad lots of unskilled labor,
(24:40):
that they're operating in the50% range.
In some cases, they were settingup the combine for one crop,
using it for a different one,not changing it, and the
performance was because youweren't getting maximum out of
the machine itself. So that's anarea where you don't even have
to automatically control it.Combine harvesting changes with
the temperatures during the day.It changes with moisture
(25:01):
content, weather. It changeswith topography, uphill,
downhill.
How many people here, whenyou're busy harvesting, take
somebody in your family and putthem in a cab and say, set the
settings here and don't touchanything? How many people do
that?
Todd Gleason (25:17):
Raise your hands.
John Reid (25:20):
I mean and so, actually, we had a lot of fleet
customers that told us theydidn't care about fuel
consumption, but with thedigital information that's on
the on the machines today, oncewe showed them the significant
dollar that they were spendingon harvesting crop and losing
some of it out the back of themachine, or burning too much
(25:42):
fuel to get that same type ofperformance, then it is a real
interesting area ofoptimization. And frankly, I
think it's probably a precursorto autonomy. If I can get that
kind of performance and reallyoptimize a machine, then I'm
ready to talk about stepping outof the cab because it doesn't
need me so much.
Todd Gleason (26:01):
Any other questions? Yes, sir? Oh, I
didn't catch that. Can yourepeat
John Reid (26:08):
Is it actually you're asking about having a service
model? Yeah. I think it's reallyinteresting because when you
think about the machinesthemselves, significant
investment. The autonomy,especially in the early phases,
is going to need a lot of handson or observations by the
company, the perception systems,are they working, is it the same
(26:32):
as it was last year? So having amodel that's similar to
satellite TV, you buy your TVbut you buy the satellite
service, that kind ofsubscription model could work.
I don't see very many of theOEMs or even the small players.
Some of them may be going thatway. I think there's value in
(26:55):
them being able to see how thefleet of machines work as long
compromising the farmer's dataso that they can make those
systems better by the learningthey get between systems. So in
the early phases particularly, Iactually believe in the long
run, services model might bebetter than thinking about a
farmer who's going to basicallyset this up and stand in the
(27:17):
field. I'd rather have a serviceprovider that train community
college students to understandhow to run these systems, and
they can afford the cost of theassets and they deploy it by
serving lots of farmers inparticular areas.
It's not a bad model.
Todd Gleason (27:34):
Give John Reed a nice round of applause. John, if
you could take that to NeilDahlstrom. John Reed is the
director for the Center forDigital Agriculture here on the
Urbana Champaign campus of theUniversity of Illinois. You've
been listening to excerpts fromhis presentation made during the
March all day ag outlook at theBeef House in Covington, Indiana
(27:55):
on this agricultural programmingfrom Illinois Public Media. I'm
(28:20):
University of IllinoisExtension's Todd Gleason out of
the office this week.
Here's an update of what'shappening from Extension across
the state next week. Starting onthe fifteenth, there is webinar
hosted by the Farm Doc team.I'll be the emcee for that event
from noon to one that's onTuesday related to the one big
(28:41):
beautiful bill and the farmpolicy changes it has in acted
related to commodity programsand crop insurance farmers
landowners and bankers all needto be in attendance please
register now you can do that onour website at wilag.org
willag.org Look in the calendarof events on Tuesday, July.
(29:06):
There'll be some other thingsyou can register for as well. On
the sixteenth of next week, theOre Agricultural Research and
Demonstration Center field daywill take place.
Agronomist will be on hand inBayless. That's in Western
Illinois to the West ofJacksonville. The Monmouth field
day is the following week onJuly. They'll follow that up
(29:29):
with the July 24 EwingDemonstration Center field day
and then wrap things up here oncampus with the Hemp Research
Open House. That one is on Julyand will require registration.
You can find all thatinformation on the PharmDoc
Daily website or willag.org.
Popular Podcasts
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.
Crime Junkie
Does hearing about a true crime case always leave you scouring the internet for the truth behind the story? Dive into your next mystery with Crime Junkie. Every Monday, join your host Ashley Flowers as she unravels all the details of infamous and underreported true crime cases with her best friend Brit Prawat. From cold cases to missing persons and heroes in our community who seek justice, Crime Junkie is your destination for theories and stories you won’t hear anywhere else. Whether you're a seasoned true crime enthusiast or new to the genre, you'll find yourself on the edge of your seat awaiting a new episode every Monday. If you can never get enough true crime... Congratulations, you’ve found your people. Follow to join a community of Crime Junkies! Crime Junkie is presented by audiochuck Media Company.
The Clay Travis and Buck Sexton Show
The Clay Travis and Buck Sexton Show. Clay Travis and Buck Sexton tackle the biggest stories in news, politics and current events with intelligence and humor. From the border crisis, to the madness of cancel culture and far-left missteps, Clay and Buck guide listeners through the latest headlines and hot topics with fun and entertaining conversations and opinions.