S8 E8 - Laser Weeding

podcast. I think the first thing we should do, though, is just kind of introduce you two to the audience. So what what do you guys do professionally? What are your roles? What are your institutions?

Thierry Besançon: So I'm Dr Thierry Besançon, and I'm an associate professor on extension weed science specialists for specialty crops in New Jersey. So I'm working at Rutgers University, and yes, I have similar responsibilities to lynn I'm working on every kind of crops which is not soybean or corn. So that includes cranberry, blueberry, grape, sweet corn, veggies, everything

old challenges as well. Clearly, the first one is that we have not getting any new mode of action, herbicide mode of action for quite a number of years now on, if it's a problem for angronomic crop, it's even more a problem for specialty crop, because in specialty crop we have even less herbicides. And you know what, what you can have for Corn or soybean. So less herbicide, lots of consolidation among chemical companies, herbicide resistance, big elephant in the room. It's also

affecting us in specialty crops. So that's, you know, that are some of the reasons. I mean, clearly, I mean, whoever's our reason as well. I mean, Lynn, you may want to discuss other stuff as well. Well, Well, I think for specialty crop world, we're also very heavily dependent on labor for weed management, and labor is becoming more difficult to source. The labor pool traditionally used in agriculture is getting older.

It's getting more expensive, and let's face it, hand, weeding is is a brutal job, and there are more efficient ways to use the labor that's available to us than than this tasks. So we're really Thierry and I and so many other weed scientists working kind of in a similar space you know, are very interested in novel technology for. Weed management. And laser weeding happens to be one of those tools out there.

Thierry Besançon: And I think I will just complete that is that if you're looking at where we are selling our specialty crops, we have a lot of export markets on. If you're looking at the continent, I will not just say countries, but continents on as a former European, I know a little bit what I'm discussing here, but it's a little bit more challenging these days to use pesticide in general, not just herbicide, but pesticide in

general for export market. I mean, in Europe, they're getting very sensitive to the fact that you may be using something just like glyphosate, you know. I mean, it may not be such a problem for corn or soybean, for example, but for fruit and veggies that you're consuming directly for the for the consumer, people are getting way more aware of that on. They want to know what has been used on. Yes, clearly some market, we know it from, you know, from experience here. Work on

specialty crops. Some markets are clearly not warm. They don't want us to use some specific chemical, because either it's not labeled in other countries or, I mean, the threshold is different from what we can accept in the US. So it's, it's, it's another big game, you know, it's a bigger difference in the room for us as well.

capacity to harness the power of artificial intelligence. It's it's growing on a daily basis, but up to maybe three, four years ago, we didn't have the capacity that we have right now, especially in terms of weed identification, right identifying the plants, separating the weeds from the crop. That's something which is really new. I mean, the laser,

it's, yeah, it's not a new technology. But clearly, the association between detection with artificial intelligence, on actuation with a laser weeder is really where to point where it's feasible right now, which was not the case for five years ago, because it was still under development.

eliminate the weeds. But the the ability to identify, discriminate and target is, is coming from the artificial intelligence and and and laser Weeding is, is one tool. And, you know, there's, you know, the the companies will suggest as to why you want to use this technology over, you know, a targeted herbicide or a targeted cultivation application,

Thierry Besançon: I will let Lynn discuss, because she's the one who started, I mean, getting a good relationship with carbon robotics, the company which is making the laser weeder, but she's the one who initiated the discussion with him, so she's probably better than me to discuss that. Yeah.

most units, commercial units in operation. And you know, they're, they're very interested in understanding how their tools and technology will work in different crops and in different environments. And we were able to establish a relationship with them. And so they brought up, they have kind. A demo research unit that's called the bud. They were able to bring that unit to both New Jersey and New York State for us to conduct research

trials with. They have a commercial unit called the laser weeder that you can see on their website, as well as some information about their demo unit.

Thierry Besançon: I think one, one thing both Lin and I were really interested is that if you're looking at these technologies, many of these technologies have been developed for environmental condition, weed condition on the West Coast, which are quite different from what we have actually on the East Coast, even further on the North East, because we don't have the same type of soil, we don't have the same time of precipitation, we don't have the same wheed species, we don't

have the same crops. So at some point, we managed to reach a kind of common understanding with carbon robotics that, yes, I mean, we are excited as weed scientists to be able to test it first, but they were also interested to eventually look at other production environment to see if this technology will be a good fit under other condition, if they don't really need to adapt. You know, their models, their artificial intelligence, on board to make it more efficient under a different

production on the amount. So, yes, last year was, well, this season was the best time when both our interests, Lin and I on carbon robotics interest came together on that's how we managed to get it for demonstration and for experimentation, not just demonstration.

Thierry Besançon: There's eight, eight laser, if I will remember, each laser is 500 watts, if you want to get a specification, but it's basically making a lot of light, so you're flashing the soil constantly, because you don't want to get any kind of shadow. If you want to get artificial intelligence to work efficiently at recognizing the weeds, seed on things, the weeds on the crop, you need to get an environment without any kind of

shadow. So it's making a lot of light on the soil, and then you can start to see you don't see the laser, you just see a.of light on a little bit of smoke. Actually, you smell it more than you see it. But that's that's the way it's working.

not just a plant. It's even more precise than that, it's targeting the meristem, the growing point of each weeds on, well, I mean, well, we're laughing, but I was, I was very impressed to see small rosette of, you know, let's say carpet weed. It's not touching the it's not touching the leaves. It's actually each shoot of the laser weeder was really targeting the meristem of each each weed on it was so precise. It was amazingly precise.

is targeting. And like Thierry said, they are hyper precise where you can, you can see it just, you know, hit the growing point of the plant, or, you know, the base of the stem. It's, you know, with, without causing, you know, kind of incidental damage to the crop itself.

the field. Um, everything else will be considered as a weed. But actually, when we're in the field on Lynn and I experimented on different crops, and we can discuss it later. But even at the very earliest, early, early stage of crop development, we never got any, I didn't see, personally, any kind of injury where the crop would have been burned by the laser weeders. The

weeds. Sometimes the laser weeder missed a few weeds depending on the size, but we never got any crop damage of the with the laser weeder. Just a personal fascination of mine. There's this split wave that these companies and others seem to be looking at is, do we identify the weed or the crop? And I'm I'm team, know the crop and kill everything else.

there's an additional layer for the weeds. But, you know, there's, there's a lot of details we're not necessarily privy to because of, you know, proprietary, yeah. Thierry Besançon: And I think the last thing I will say to regarding why to do it in the Northeast as well is that, yes, we may have some crops that are not growing on the west coast that you know they may have been interested. I mean, we, we tried it on on peas. So I know that they had that on peas. But we

tried it on spinach in New Jersey. We tried it on beats with Lynn in New York, we reinvestigated some crops that they may not be familiar with. So, you know, it was kind of good cooperation, because it was good research project for us to collect data, potentially publish the data, but also for carbon robotics to improve their database of knowledge regarding some some specific crop that they are not familiar with, for which you don't have any previous data.

itself, it's pretty close to $1.5 million per unit. So I think it's, it's the market is, you know, not for everyone. So that's one first limitation on clearly, especially for us. I would say New Jersey, where you have, you see, I mean, you know, people don't imagine New Jersey as a, as a veggie producing state, but we are the Garden State. We have a lot of small operation farm for Philadelphia, New York. We are one of the few states in the United States where the number of farm is

growing actually. So that's, you know, small farms, small, small operation units. But that's that would be a disadvantage, because clearly they don't have the financial structure to afford this kind of equipment, I will, I will let Lynn said other limitation, but we have some other limitation as well. Yeah,

takes time, so it can slow down the operation. One thing that I, I have read about carbon robotics is their, their newer units can integrate, you know, with, with the, you know, the the tractor operation itself, so that it can make decisions based on the number of weeds it's detecting and the number of weeds that it is acting against, so that it can speed up the tractor or slow down the tractor, so that you may maintain an efficient speed for, you Know, weed control. And so I

do think, you know, speed is certainly an issue. But it sounds like the companies that are involved in this, you know, this sector have that's something that they know and that they're absolutely working on. How can we, how can we make the processing, you know, portion of of this process

faster. You know, how, how can we, how can we optimize? And you know, the the technology, Thierry Besançon: when, when Lynn talk about optimization, something that we may want to, I will not set the bank, but I want to make it clear that it's not because you have a laser weeder that we will not need herbicide, and especially when you're looking at our condition in the northeast, where we have, well, not this year, at least in New Jersey, but when you have

frequent rainfall, you have a you have quite a few flush of weed energy during the season on if you have a long, long season crop, I mean, you may, you may need to get multiple passes of the laser weeder. I think what I learned with this technology, like with other new technologies, that you may still have the need for residual herbicide, for pre emergent herbicide, to just reduce your level of weed population, so

that this new technology will be efficient. Because sometimes we never really experienced it with the laser weeder, because we went we were not in this situation, but we know that if you have too many weeds, you can just overwhelm the technology. So you may still need to get this up from pre emergent herbicide application to reduce your weed population on to have a more efficient use of the new technology, laser weeder, for example, later in the season, because it's tricky. You know,

laser weeder, it's tricky post emergence solution. So, and that brings up another point with it. You know, this is post emergence weed control. You know that there, there is no residual activity unto itself, and just like every other post kind of control tool or technology, whether that be chemical, physical, thermal in this case, is that weed size is definitely going to affect performance, And we need to be timely with whatever operation we have in

place. So it just having, you know, new, you know, exciting technology doesn't, doesn't absolve you from being a very good you know, grower or weed scientist. You know, you know crop manager is you. Have to get in there, you have to be timely, you have to be effective, you have to be efficient, and you have to be thinking about the system as a whole, and how this, you know, how this technology works in a system. Because we

don't even need to talk about, you know, laser weeding. We can talk about, you know, almost anything we do, it can't be standalone. Thierry Besançon: Lynn brought a really good point too. Is that, compared to condition in California, we are we have much more rainfall on this kind of equipment. I know that the laser weeder, one time we were getting rain on the person from carbon robotics was operating the laser weeder. The bud, was really he

was hurrying up. He wanted to get it done before the rain. So same thing. I mean, some of these new technology, because you have artificial intelligence on board, because you have very high technology equipment, this kind of equipment may be very sensitive to environmental condition, especially rain, things like that. So that's that's another limitation is that sometimes, you know, depending on the kind of condition you have, you may not be able to use it because when it's sensitive to that.

New York are a very short seasons. You know, they the spinach and the peas, you know, they're very, they're, they're very fast as as far as crops go. These aren't 110 20, you know, plus day in, in the field. So we were really using them, using the the technology, and investigating it, you know, kind of in that, that very early period where weed control is most critical. So in that, in that respect, our crops were smaller, you know, and, and, you know, our weeds were also newly

emerging. So I can't imagine that there becomes a point where, you know, the crop is, is too big to really effectively go in with the with these kinds of tools for weed control. Thierry Besançon: Well what, what we saw in in the Pea trial, I mean, we're looking at that, we were able to, we had a Peatrial, second Pea trial, and we're able to push it up to

harvest on. We just needed to get two passes of the laser weeder, because later, what is going on is that you're getting so much Pea biomass that you're covering the row, you know, the spacing between the rows on at this point. I mean, you don't even need to get any kind of weed management, because you're reaching canopy closure. If you want on, you don't have new weeds emerging, or if you have it was very interesting. We saw

a few plants emerging, but they never produced. I mean, we saw a few weeds emerging later, but they never produced enough biomass to be really competing with a crop. So that was interesting. You know, regarding biomass production of the crop as well.

with the lasers on the unit. So we actually have running counts of of how much, how many they detected and eliminated.

cell on your getting a nice burn. Actually, it's so powerful. The level of energy is so powerful that in in my sound design in New Jersey, I managed to get some glass made of laser eating the ground on creating. Summary, there was some high energy, and you could see a little bit of melting of of the sun at the surface of the ground because of the laser weeder.

Thierry Besançon: Well, I have to sign carbon robotic, because it made me aware that it was really time for me to get new glasses. Because actually, sometime I was not able, I was not able to see the weeds emerging. There was a laser shot at the surface of the soil on I was, Wow, that's amazing. But yeah, they were detecting something, and there was a laser shot on I was I didn't see anything. I didn't see any weed. Yeah, I think it did highlight how old Thierry and I

are. Is because there were seeming the laser beam and like we're, like, down on the ground, like noses practically to the ground, like I don't like it, don't see it, Thierry Besançon: and it's there. I'm still nine months earlier on Earth, so I'm still fine.

California in June last year, it's still extremely slow. You know, I think for many of these agronomic crops, it would be a disadvantage. Can you define slow?

probably, of what it can do. But I do think that this technology is, you know, obviously, has, has, you know, gained access into the markets where we are talking extremely high value crops that are grown on, you know, certainly compared to a corn and soybean, uh, small, you know, small acreage that have been so heavily reliant on human labor. Uh, you know, they, they do have a fit there, but that doesn't mean that's where

they're staying. You know, these, these companies are, if nothing, but inventive, you know, and you know, investigating their options. But for, you know, even for as much as we talk about the cost, Western Growers did a case study of of what, what, what, you know, the use of this

technology. And in some of the companies you know, the farms that we have talked to that are are using the technology you know, they have found return on investment in their systems again, out in California of like, 300 some days. Thierry Besançon: So in organic cropping system. It was an

organic cropping system as well. But, you know, the other thing to consider too, that was part of the work that Lin and I did on we had really good data on that is, if you're looking at herbicide in specialty crop, they are much more limited than

what we can have in agronomic crops. And in addition, what we really saw this summer, that was amazing, especially in spinach and beet, was that these are besides that are labeled on some of these crops still have a huge judgmental impact on the crop by limiting emergence of some of these crop, especially for small seeded crops like spinach on beets, yes, the herbicide will

still negatively impact the emergence of the crop. So one advantage for us to use a laser weeder is that you are getting the same level of weed control, but in addition yet less crop injury, less crop damage, because you are not using pre emergent herbicide on the crop. Yeah, so that that's just to bring up the point that Thierry did about the potential that that some of our crops can be so sensitive to the herbicides that we rely on, you

know, to suppress the weeds, to be able to grow the crop. I had a I had conditions in New York, where we had some kind of we had torrential downpour, and then we had such hot, dry conditions that we had significant crusting in the plots where we had just used the laser weeder alone, those the beets and the spinach were out of the ground. You know, they were out of the ground already, and they were actively growing where we had used our traditional herbicide program, those those plants were

stunted. We had less emergence at the time that that crusting event occurred, and so we delayed the crop significantly because of that. Thierry Besançon: And we had. So Lynn was mentioning what she was talking about, the use of pre emergent herbicide. But in my experimentation in New Jersey with peas, we used post emergent herbicide. We use a standard herbicide, okay, on we had the same level of weed control with the carbon with a laser weeder

than what we are getting with the post emergent herbicide. But when you're looking at crop biomass accumulation, we had less crop biomass accumulation where we use the post emergent service side as compared to the laser Wheeler. So you may still have some detrimental effect, maybe not visible in terms of chlorosis, chrotic response or necrosis, but you may have some stunting, like, what Lynn was mentioning, even caused by post emergent herbicide that we did not see with a laser weeder.

Yeah. So I think this is this for for us, from an academic standing, is that this is, is making us think about really, like, how best to use some of these novel tools and technology, you know. And my thoughts are, well, gosh, you know, maybe we can use that laser weeder early, and then we can have, maybe products that have residual control later, when our, you know, some of our specialty crops are more established, maybe less likely to be impacted. You know, at

that early, you know, emergence phase. I think there, there's just so many things that like again, we're just, we're just hoping to be able to do this again next year, because there, there's so many more questions to answer, of of Integra, integrating the technology into different production systems, uh, effectively. So, so

are true. So this isn't intended to be like, to feel like a gotcha question, but I think is the next line of research then really to figure out how to integrate a non chemical strategy like laser weeding with herbicide tools we have,

alternatives that can be used. But I think what this research really does highlight is that a single tools may not be standalone, you know, and we need to be thinking about, okay, if we're going to invest in this one type of technology, you know, what other types of technology are going to work best with it? You know, for for some for some tools, maybe, you know, using a cover crop is going to be effective. Maybe

it's not going to be effective with others. Maybe this is going to, you know, integrate well, with a chemical control program. Maybe others won't. And so I think we really need to be. We need to be, I think sometimes, as weed scientists, we think about within a season, right? We think start of crop to finish of crop. And I think this is where we really need to be thinking, Okay, how do we, you know, before the crop goes in the ground, to after the crop is harvested, to the next season's

crop? Because, you know, maybe, by using laser weaning technology, maybe we pull out a tool that has not allowed us to rotate to a different crop, and now we can move into that cropping system, but maybe to use this technology, we have to have something, a different crop in advance that provides us with

a level. And I'm being I'm being vague, because I just think there are so many permutations based on who you are, what you're growing, what your weed problems are, I think it's just again, highlighting we need to be thinking about the system and how to integrate it into a system, rather than how to use it with just one specific other tool. Thierry Besançon: To be honest, some of this solution will not be working together. You know, when you're looking at the laser

weeder. How do you make the laser weeder working with, you know, a permanent cover crop. So if you use clover, for example, how do you integrate the laser weed? If, if you're using clover, can we use a laser weeder just on the bond of crop that you want to keep weed free without? You know, we don't know. I mean, there's just one additional tool on we need to

learn. How can we integrate it with other technologies, other solutions that we are all that we are investigating at the same time, because we have been so focusing so much on every side for quite a number of years now, we are starting to look at new, new approach, especially in specialty crop. It was maybe the focus was, you know, for example, on cover crop. It was mostly grown we crop, but now we are starting to look at cover cropping, specifically for weed control in specialty crops. Yes,

we are looking at all these tools at the same time on yet. I mean, just the first year. We don't know the way they will be integrating with each other. It's just the beginning. So and I think this is what makes it everything so exciting. I mean, just this sector is in this novel technology sector is just so dynamic, and there is so much innovation going on, and all of these companies like just, just what they're doing, what they intend to do, what they hope to

do. I mean, I probably, probably can't even fathom some, some of what they are are thinking about and who they're partnering with. I just think it is, it's an exciting time to be thinking

outside of the box for for weed management. And, I mean, I'm, personally, I am thrilled to to kind of be part of it and and to see where it's going to go. Because it's, it's, it's, it's wild, and it's it's thrilling, and Thierry Besançon: I see so, so good time to be a weed scientist, because many of these companies recognize the value of working with weed scientists as well people who know the system, who actually are interested in integrating this new technology

in other strategies on that may not have been exactly what it is right now, at the beginning, you know, they are technological people, they are engineers. They may not be plant biologists, they may not be weed scientists. So now they start really integrate the value of us. As weed scientists bringing the knowledge of, okay, you know it can work because we can kill specifically this weed or your system is great, but I'm sorry it's not really working well on yellow nutsedge. So what are we

doing? Because that's our number one problem for us in New Jersey. So they start to see the value of that on it's exciting time, like Lynn said, because, yes, I mean the this, they understand that, yes, it's not just them working. It's working as a team on, working with, with researchers as well.

something more. It's we, again, it's a system, it's a system, it's a system, we and and it's hard, because there are constraints on growers, whether you know it's there's financial constraints, there's environmental constraints, there's social constraints, there are so many constraints that affect what you can do. And, you know, I sometimes I feel like the bad guy saying we need to diversify. We need to be

thinking about, you know, adding this in and changing this. And it can be hard because it's, it's not necessarily feasible, but truly and honestly, nothing is going to be stand alone. We have to, we have to diversify Thierry Besançon: we are just I mean, diversity is critical clearly. I mean, that's in life, but in Weed Science as well. On, truthfully, with this technology, we are just starting

to investigate it. On we some of the limitation is that we don't even know how the weeds over time as a weed population, not you, not just the weed as a species, but the weed population in general. I mean, what kind of shift are we getting in our weed population by using laser weeder, we know that it's targeting the meristematic tissue of the weed, so the merostatic tissue has to be exposed many, many grasses, the

meristematic tissue is underground. So we have all these questions that we'll have to investigate when as we are using this technology to see, okay, I mean, yes, we have limitation, because we realize that by using laser weeding, maybe we'll be shifting our weed population, maybe to more weed, more grasses on which case the laser weeder may be less efficient. So we don't know yet, but I Well, Lynn and I mean, something that at least we learn in weed science, there is

nothing like a silver bullet. I mean, it has been debunked by herbicide resistance quite a few years ago already.

And I think this is a tool. This is a tool, and it's going to be for some people and not others. And I think we are just lucky that we, we live at a time when this tool does exist, and that there is the ability to to conduct research and to investigate it, and to work with growers and to work with the companies you know, to to just make food available for everyone, Thierry Besançon: and we have to stay, I mean, optimistic as well. You know, we are just at the beginning of the emergence

of this technology. We are just at very early stage on, like with any other technology, over time, you can see that there is a democratization of the technology, if you want things are getting smaller. Things are getting more affordable over time. So we are just starting right now on, maybe 10 years from now, we'll have smaller units that we can use in smaller

operations, that people can use in their garden. Who knows? I mean, you would have told me 20 years ago that I would be using a smartphone today on, you know, check my emails on my smartphone every five minutes, every day. I would have been laughing. Okay, yes. And I remember when I was 30 or 20, we didn't have access to this technology. So it's, it's moving so quickly that,

yes, we have to stay optimistic. Am I saying that this company, yes, I mean, you know, it's a start up date, but they understand that over time, if they want to access new market as well, they need to come up with units that are smaller, more affordable, for people, for farmers who were making the you know, our farmer population here In New Jersey, we have small operation units again on I think it will come with time. It will come with time.

all. I think they were very generous with their time and and generous with with bringing their units for it, you know it, this is, this is a cost on their part, and so I think we have to acknowledge them and express my gratitude for being given the opportunity. I clearly can't buy a laser weeder on my own. So you know, thank you to carbon robotics for giving us the opportunity to conduct research.

Thierry Besançon: I think on my own, I have a Rutgers weed science instagram account, but it's probably not as frequently updated as Lynn, because she's really the social media media Bee and I'm glad she's doing it in the team. But yes, I mean, we're really grateful that we're able to work with carbon robotics this year on, we built up a really good relationship

with them as well. But hey, I will thank Lynn as well. Because we are, we are making a good team as well, where we have a good research team for specialty club in the north seas between our two labs, and actually it's almost like one team

crop protection. Thanks again for listening, and we hope to see you next time you.