You're listening to the WeedSmart Podcast, where each fortnight we chat about dealing with those pesky weeds. Welcome to the WeedSmart podcast. I'm Jessica Strauss and on this edition of the podcast we're catching up with West wimmera farmer Sam Eastwood. Sam is particularly keen on managing weeds in non cropped and unproductive areas of the farm. And so we're gonna find out from him how he keeps on top of them and why it's such a priority for
him. And we'll also be chatting with University of Sydney precision weed control PhD student Guy Coleman. Earlier this year Guy, along with Dr. Michael Walsh hosted a workshop on weed recognition in Wagga Wagga New South Wales and Guy is going to walk us through the key learnings from that workshop and also reveal some details about a new open source weed locator his team is releasing, but my co host and Western extension agronomist for WeedSmart, Peter
Newman does join me. How are you going Pete?
I'm really well Jess. Rain keeps falling in WA. It keeps on coming just every three days. It's so funny. Just when you get over wet season like this, you just completely stopped looking at the weather forecast because and I'm sure the farmers are doing that too, because it just keeps on coming.
Yeah, it is very wet. And I stupidly did a hike on Sunday, which was our second wettest day in July in 20 years or something like that Pete and yeah, I was completely drenched. We had people falling over in the mud and everything.
So there's no such thing as bad weather. Just bad clothing. Were you adequately dressed?
Well, I thought my pants would be suitable but they weren't. So they were drenched. But I did have a nice warm core because I had a good GoreTex jacket on. So I was okay.
Yeah. And you've got some exciting news, Jess.
Yeah, so I'm officially taking over the role from Lisa Mayer as the project manager for weed smart. So that's really exciting and a great opportunity for me. And yeah, it's big news. It's scary. But it's exciting. And I'm really happy to take on the opportunity.
But yeah, fantastic. Yeah. Thank you. project has been going from strength is strength. I'm sure it'll keep going under your watchful eye.
Fingers crossed. But yeah, I'd like to really say a big congratulations to Lisa Mayer as well for doing such an amazing job for WeedSmart and managing the project. She's really enabled the project to flourish. So it is big shoes to fill Pete, but I am excited.
Yeah, absolutely. I remember high fiving Lisa, at the AHRI offices - that must have been in 2013 when we won the project. Yes. And Lisa was the one who really led that bid and won the project. And has led it ever since, so yeah, you know kudos to Lisa - huge job over a long period of time. And it was a good project to start with, but it's just grown and grown. And now we've got this national team so we can do things like this, Jess.
Definitely. And just so everyone is aware Lisa is not leaving completely. She'll still be project support for WeedSmart. So she's still in the office. Sometimes when you have changes in positions like this. You know, you think the the person has departed but Lisa is still working with us, but yeah, but just on a part time basis. So yeah, really great, exciting opportunity. But Pete, let's get into the podcast
interviews for today. I just wanted to firstly, because this is kind of related to our first interview Pete we do have a new article up on the WeedSmart website, all about double paraquat. And this one is basically looking at what the benefits are of using a double paraquat knockdown. James Jess, who is a research and Technical Services Manager at Western Ag in Ballarat provides answers on this one. So I'll put the link for that article into the
podcast show notes. But Pete, can you just run us through this whole concept of double paraquat and why this is, you know, potentially a good option for farmers?
Pretty amazing change, Jess, it's incredible how you know, we think that we might have a practice that we're going to be stuck with, well, not stuck we've been doing for a long time, then all of a sudden it changes and was talking to Phil Hawker from Western Ag recently, he will be James's boss, and he told me all about farmers getting into double paraquat really because they're having problems with glyphosate
resistant ryegrass. All of a sudden they were doing paraquat followed by paraquat, including group G's in there, and other pre-ems and massive adoption, he reckoned 80% of his clients in some areas are using this practice. And so a really quick adoption of a different knock
down practice. And so we've got a bit of work to do there, Jess, good article to start with here, written by Cindy Benjamin, but we're gonna have a bit of work to do here to work out how we look after paraquat with this double paraquat practice.
Yeah, certainly. Yeah. We'll do some more interviews and yeah, potentially webinar details and that kind of thing down the track. Yeah, we'd love to hear from you. If you've got any specific questions around it and areas you think we should cover? Let us know by sending as a direct message on Twitter. But Pete, let's get into our first interview for the podcast up first, we're going to be hearing from Sam Eastwood from Kaniva.
He's a grower there and he really has got a good priority around managing his uncropped areas and controlling those weeds on fence lines and uncropped areas of his farm. Pete, why is this so important?
Well, it's a bit of an old chestnut Jess, we've
talked about it before. But we don't want to destroy one of our best herbicides if not our best ever herbicide glyphosate on the least productive part of the farm, and so that's your fence lines or drainage lines or just some of those unproductive areas and so I've spoken to a lot of growers where they have developed glyphosate resistance on their farm, they developed it on the fence line, and then it spread into the paddock, and now they're dealing with it over
large areas. And so yeah, Sam really talks about how he prioritised this early on, and just made sure that it used a good double knock went early and, and has avoided those big blowouts on the fence lines.
Yeah. All right. Well, let's take a listen and hear from Sam. In today's interview, we're catching up with farmer Sam Eastwood. Sam is based in Kaniva in the West Wimmera. And he grows canola cereals, legumes and pulses for grain and hay. And Sam is particularly keen on managing weeds in non cropped and unproductive areas of the farm. And so we're going to find out how he keeps on top of them and why it's a priority for him. Sam does Join us now. Hey, how are you going, Sam?
Yeah, good. Thanks, Jess.
You've been very busy. So I really appreciate you taking the time to do this interview. What have you been up to?
So we've been busy the last yeah, the last week or so getting on top of some grass sprays. And yeah, getting ready to for our first round of urea spreading. So we've had some really good rain here in the last few weeks, which has brought the crop along with a really late start. So now we're just trying to fit everything in in between these rain events, which is a challenge, but we're getting there.
That's great to hear Sam and before we jump into the topic today, can you just give us a bit of an overview of your farming system just in a bit more detail?
Yeah, Jess so we're a no till continuous cropping operation in the West Wimmera. We only have livestock on the property over summer for grazing stables, which we do on adjustment so basically no stock in the system as such. And yeah, we're roughly 30% canola 30% cereal 30% legume give or take a few percent obviously for for the cropping rotation.
Great. And what kind of weed burdens do you have at your place?
So yeah, typical of a Wimmera kind of clay farm. We've got obviously ryegrass is probably has been in the last wild forever probably probably public enemy number one that also tares, which is vetch straw. They're probably the main problem weeds, radish as well. But we tend to keep on top of that pretty easily. They're the main probably the main ones, marshmallow and deadnettle.
They're all in there. Yeah, but ryegrass and tarrs and bedstaw seem to be the probably the main one that gives us issues.
Yeah, right. And Sam. So we are focusing on controlling weeds in non cropped areas in this interview, it is something that you do keep on top of at your place. Can you first tell us what made this a priority for you?
Yeah Jess, as we moved into a continuous cropping
Yeah. Okay, that makes sense. And so how did system, and so we got completely out of livestock bed decade ago, or a bit over. And we're focusing heavily on on growing crops for grain. And we've also introduced hay, but it was when we started to really focus heavily on continuous cropping.
And the fact that we were tryin to get on top of rye grass at the time probably was the mai one and rye grass resistance And yeah, really trying to ge our numbers, it was all bout getting our numbers down as a numbers, you know, the nu bers game. And yeah, controllin the fence lines just became pa t of that integrated appr ach. Because we were getting rea
onable control. We now in padd cks in crop if you like, bu we'll bring them back in fr m from fence lines and around tr es and those non crop kind of areas you go about controlling these weeds in non cropped areas?
The main approach has been to spray them out. So we use a double knock, spraying scenario, just like your double knock paddock for you profit mix of non selective herbicide with glyphosate thing during the main leg work there for the grass and the double knocking with paraquat that couple of days later, which is yeah, with the double knock probably was introduced maybe four or five years in, because we found that we'll try to pull down these really large populations of ryegrass which were quite
advanced and that became harder to do. So yeah. We moved to double knocking, which we've had quite a success with.
Yeah, okay. And was that due to glyphosate resistance? That's obviously an issue in lots of parts of the country. Was that part of the reason why you moved over to do Double Knock?
Yeah, it was, yep, definitely, because we were really concerned that we were actually going to breed resistance on our fence lines by applying or under applying really, glyphosate to large ryegrass plants and having them survive. So then they set seed and we drag those into the paddock we're actually going to make resistance issues or rye grass issues worse. Yeah, so, you know, bringing a susceptible seed into the into the paddock is a pain because it's there. And you've got to then deal with
that. There's always a bit of dormancy in some of those seeds. And so it lasts for a number of years. But if you've if you've brought a resistant seed across, or seeds that are partially resistant, because you haven't given them a fatal dose on the on the fenceline, you actually escalated the problem. So double knocking became a really important so that we had no survivors, basically.
Yeah, perfect. No, that makes total sense. And so what impact do you think this has made on controlling weeds on your farm? Sam?
Yeah, look, it has, it has made a large impact, we could, we could definitely see where we were, where we had rye grass. I'm gonna focus mainly on rye grasses that's mainly the weed that we've been trying to control on our fenclines. That, that was definitely noticeable that once we started getting on top of the weeds in the crop, and we were doing the fence lines as well, we weren't bringing them in off the fence line. And we did see
big reductions in that. Because obviously, the harvesters, when they pick up a rye grass seed or any weed seed in in a crop or harvest situation they don't just deposit that seed straight back out the back of the header, right there, they carry it and distribute it across the paddock quite effectively. Which is no good. Yeah. So yeah, by controlling those fence lines, we definitely saw a reduction in weeds coming in off those non cropped areas for sure.
That's great to hear. And so for other farmers who might be listening to this, what would your advice be to them in terms of going about controlling weeds in non cropped areas? Is there any tips you'd like to share?
Yeah, my kind of thoughts, our observations over doing it over time would be go probably earlier rather than later, we can get caught up doing other spray operations and fertiliser spreading in the fence on tend to be the last one obviously, that you've sometimes that you can go and tackle. And by that time, your populations are large, the plants are big. And they can take a huge amount to to get under control or to
kill. So really robust rates of chemicals, whatever you're using high rates of sulphate of ammonia in the mix, as well has proven for us with our glyphosate application to really help the glyphosate and all those other general spraying tactics like don't be spraying after or around frosty weather, just totally avoid frosty weather. If you've had a couple of frosts, you've just got to wait and get your double knock
spot on. So only spray as many fence lines as you can get back to within that, you know two day period to get your double knock on and make sure you get that that control. Because if you miss it and they survive, you've got a really big problem.
That's great advice, Sam. Well, thank you so much for joining us on the WeedSmart Podcast today. I know you're so busy and yeah, very valuable for you to share your story with us on managing non cropped areas for weeds. So we really appreciate you taking the time.
No problem at all.
Thank you so much to Sam Eastwood there and Pete, I really like how Sam has got his priorities right and he really knows how important it is to manage those unproductive areas and he goes out and really looks at them as a priority and kills weeds along fence lines and unproductive areas, first. Pete is this common, or is Sam getting a golden star here?
No, common practice Jess, is to finish everything else and then get an old man, normally your father or someone who's hanging around and they go spraying fence lines to keep them busy. I'm being a little bit unfair here but no, often you know farmers get very busy looking after the main gig which is putting in crops and spraying them and fertilising them in the fence lines get left
til last. Whereas Sam has really focused here on though they are a priority because we've got to make sure we don't blow out big glyphosate resistance problems or resistance problems in general in those areas. Yeah, so he talks about going early going, double knocking and just taking no prisoners and not accepting any survivors. So that's just like everything. If you prioritise something, yes, you can make a go of it.
So Pete, moving on to our next interview for the podcast today, we're going to be hearing from Guy Coleman. Guy is going to be talking to us about some of the key learnings that were presented at a recent workshop he helped organise with Dr. Michael Walsh. And that was all about open source data and weed recognition. Pete, you've done a bit of work with guy, can you just talk us through how some of this stuff works? Because it is complicated. And it's very interesting, isn't it?
Look, Guy is one of the smartest people I think I've ever met. He probably doesn't want to hear me talking him up all that, but he's a very clever fella, we're very lucky to have him in agriculture. And he just loves this stuff. He's built his own robot to drive over plots and measure things and transcribe things and pluck things out weeding, and he just absolutely is fully dedicated to the cause of, of machine
learning. And, you know, using cameras to detect weeds and he is trying to now apply that over a big area and allow people to work out whatever weed is a problem in what situation, take photos of it, upload it to this website, and then potentially have algorithms developed in the future so that we can make the most of this scene in spray technology. So we're lucky to have Guy. I've done a little bit of work with him firsthand on this stuff. And I've felt like an absolute din
Yeah, no, he's such a clever guy. And I have met Guy in person and seen some of his work. And yeah, very passionate, which we love. And it's great that he's in agriculture. Is there anything else you want to mention before we take a listen to this interview? Pete?
No, I think we should just go for it, Jess, let's take a listen.
In this interview, we're catching up with University of Sydney Precision Weed Control PhD student Guy Coleman. Earlier this year guy along with Dr. Michael Walsh, hosted a workshop on weed recognition in Wagga Wagga, New South Wales. And so Guy is going to talk us through the key learnings presented at this week workshop and also reveal some details about a new open source weed locator his team is releasing. Guy does join us now. Hi, Guy, thank you for
joining me on the podcast. So you're no longer actually based in Australia, but in Texas in the US, which is really exciting. Can you tell us a little bit firstly, about your journey over to the States?
Thanks much, Jess for having me on. Very happy to be here to talk about few of those things you mentioned. So currently based in Texas A&M in College Station in the US as part of a bit of research for my PhD. I came over here just about a month ago, I think now five
weeks ago. And what I'll be doing over here is looking at the detection of Palmer amaranth, which is a key weed in this part of us, I guess, the whole of the West in cotton and how different growth stages of Palmer amaranth basically affect its detection accuracy, it's as I'm sure everyone would be well aware, weeds change a lot in their appearance as they grow over the season. And as does the
crop. So if we can better understand how that detection accuracy changes over the season, then perhaps we can work out better how the performance of these green on green sprayers might change as well.
That's really interesting. And yet, Palmer amaranth, I know about that weed. That's an awful weed..so I think anything we can throw at it is a good thing. And so Guy, today we're going to talk a little bit about the open source way data that University of Sydney and the Precision Weed Control lab, which runs out of Sydney university and has made available. Can you just firstly give us a bit of an overview of what this involves, obviously,
there's a few moving parts. So if you can give a bit of a broad overview that will be great.
As you mentioned, there, it's the the open source nature is probably the first thing to start off with. And those people who haven't come across the term before it, it's all about people being able to contribute and access data software, even hardware designs or free, but also having this ethos around community based
development. So lots of tools these days in the machine learning world and machine learning community, people tend to call it open source, so anyone can use them, they can use them potentially based on their licence for commercial purposes. And in different licences, they can even privatised that that as well. And then sell a product based on that open source code or data. But it's all depending on the different type of licences that are out there that are well worth looking into. That's I
guess, the background to it. But what it really means is that anyone can access and anyone can really start to develop these sorts of algorithms based on those those open source architectures. So big companies like Google and Facebook, have released all their open source tools. So one of them is called
TensorFlow. And so what TensorFlow allows people to do, in their coding language called Python, is actually train machine learning models based on another set of open source tools, which is these open source architectures of algorithms. So it's like getting a house and you feel a little furnishings, right. So all the house is open source, and you get to put the furnishings inside it. And what we're providing through the open source database is effectively
those furnishings. And in this case, it's actually weighed data. So the images of weeds in Australian context and it could be global context. At the moment it's of images of weights in weight, as well. And previously all this work and research and development of architectures. So that like structure, Guess that you can develop into a specific algorithm. All that work has been done on things like coffee cups or chairs, or dog cats. And so all these algorithms are typically designed for that sort
of detection. So one of the benefits and what we hope to see down the track from this platform of imagery is more specific development that focuses on weed detection. So we have these larger architectures that are actually designed for weed detection that anyone can use. So really focusing on open source side of things. That's probably the ethos of it. And like the the reason behind it, and I'll probably talk a bit more about the the structure of it as well, if you feel like that would help.
Yeah, let's hear a little bit more about the structure as well. And then we'll get into some of the key learnings from the workshop in a moment did you want to just give us a bit of bit of structural overview of how it works?
Yeah, so anyone can access to the moment so it's all live and ready to go. So if you head over to weeds, dash ai.sydney.edu.au, or wait AI is a website. And what it is, is this at the moment, it's a series of images that people have uploaded, mostly at the moment that people can upload any images, they will like that have been annotated of weeds in any crop. And people can also download that data that would like to use, I guess, to train
some of the algorithms. So someone could upload a set of images to I have later recently, your blue lupins in an airlift lupins from the northern growing region in NWA. And that data can then be downloaded and used by a machine learning engineer anywhere in the world to then solve that part of the problem or, or run algorithms and develop these algorithms that solve that problem. And previously, that wouldn't have
been possible. So what has gone into the structure in the development of this database is a lot of development around data formats and standards. So one really big issue about all this is trying to make sure everything talks to each other. And I'm sure they will be well aware of the interoperability issues that would be on farms. And that extends equally into data sets and machine learning to to making sure everything talks has the same annotation format. So all the weeds are in
the right places. And that when they when you download it, you know that this is a weight and that's the crop potentially. And then things like agricultural contexts are really important as well, to what growth stage crop is it in what background conditions are there, things like surface soil coverage, soil colour, even types of weeds, you have all the lists of like taxonomy, the hierarchy of the waves, and that gives the ability then to search and filter by different weed
species. So if you want to just detect broadleaf weeds, then you can go up to the family level and select all your Rasta Casey, weeds that might be relevant for wild radish detection, perhaps all that is equally as relevant for grasses too. So what this this platform then does is opens up the world of weeds and we detection, the whole range of new people that might not necessarily have been able to access it before. And also, I guess, contributions to so farmers and researchers and
anyone can also contribute. So it's that the contribution side as well as the access side.
Definitely. And we'll give a bit more information later on, I think on how farmers and agronomists can contribute. But firstly, let's talk about this workshop that you hosted not too long ago. So earlier this year, you hosted a workshop on weed recognition, which is what we're talking about today and what you've explained there about the database. But what did participants actually learn at this particular workshop? What were some of the key takeaways?
The workshop was all about trying to understand the requirements of the end users of this data. So it was bringing people in the weed control industry. And also farmers, economists, end users of all this information and weed recognition technology, and bring them into the same room.
So you can work out if this platform in its current form is useful, or what sorts of developments would be required to make it more useful for uploads potentially, at the grower level, or even at a grower group level, or how that would work in effect. And so what people probably took away from it was more about how the platform would contribute to win
recognition development. And like the open source, ah, I've been talking about, and probably secondly was also the machine learning side of the project too. So that workshop was part of a larger GRDC machine learning projects. And while the database was one key feature of that, the second aspect of that project was also developing some machine learning models that recognised ryegrass and turn away in wheat. So that was, I guess, based on that and northern growing region in New
South Wales. So those turn away it is a bit of a allegory for wild radish as well. But what people I guess learn from that as well was that while Tennant weed is in wait detection is was
possible, as was ryegrass. What we saw was the turn away was easier to detect, then then ryegrass and also easy to annotate to so if you told someone to annotate a nice roset of wild radish alternate Wait, it's fairly straightforward because you can see where it's clearly delineated, whereas trying to annotate potentially ryegrass or other sorts of grass species in Grass crop is fairly
difficult. So we really benefited from hearing that feedback from growers and and the weed control industry about how they could contribute and how the platform could better match their expectations and how it could hopefully gain a bit of traction and really contribute to the development of these algorithms.
Certainly guy and as we know, weed recognition software, it has come a long way. And we're seeing real big breakthroughs with this technology, not only with green on brown weed detection, but green on green. And that's already being seen in optical spraying. But we're really only at the beginning, as you're talking about this open source data that has been made available, there's still a lot of progress that can still be
made. So I just wanted to get your perspective, since you're in the in the thick of it, what's your perspective on how this open source data will help in this process of innovating, weed control solutions going forward.
So what I think really excites me about open source data and open source software and the hardware is that in other industries, so in the machine learning industry, all this open source stuff really promoted and advance the industry at breakneck speed almost be and that was really only potentially possible because of the open sourcing of this technology. because that meant that anyone could use the technologies, all of a sudden, you opened your end user group
up considerably. And that meant that people could start finding potential issues with it, they can start finding new uses new adaptations, I guess that the existing technology. And that means that all of a sudden, you have this much wider group of people that can start contributing where they might not have been able to contribute
before. And if we do that agriculture, and particularly these data sets, and as I guess I mentioned shortly, this open weed locator and even more open source projects, we can start to involve with the weed control community farmers, really at the development stage, if they're interested in and start to improve that feedback cycle between the people, the end users, and they might have a whole range of different, I guess, contributions in terms of how these things might look in
the field or how I work and what's the features they might need, and how that technology translates to real build impact.
Yeah, that's really exciting, Guy. And like we've mentioned farmers and agronomists, they can help in this process. And they are invited to contribute to the wave library to help with the machine learning of different ways at different growth stages. Can you just talk to that a little bit more about how people can actually be involved in this process.
As you mentioned that it would be fantastic to have as many people contributing as possible. And the best way to contribute I guess, is to religious get out there in the fields and collects imagery of or weeds that you're interested in. But what's most important is that the collection must be consistent and consistent heights that are reportable and might represent a use case in
the future of app. So if your camera is held directly above the iPhone cameras held directly above the weed, then that might be a good detection, if the camera on a boom or robot is directly above the way in the future as well. So it's not perfect that translation between a phone camera and machine vision cameras or the deployment level, the expensive Taiwan's that they operate on these machines, there is a small benefit, I guess, or a benefit to having that initial data
people can work on as well. And so probably the first step is getting in contact with myself or other people on the web so that we can point you in the direction of the sorts of weeds you want, might want to collect images of and how you might want to collect that imagery as well. So maintaining specific bytes, and we can point in the direction of the right data to record as well.
Yeah, certainly, Guy. And we'll provide the details for that in the show notes. And I'll put your Twitter handle in the show notes as well so people can follow along. Now. Lastly, guy we teased in the intro there that you do have some exciting news to share with us on a new development you've been working on as well. And this is the open source weed locator. And you have also talked a bit about this on your Twitter, can you tell us a little bit about this new development.
As I've mentioned, with the open source software and the hardware side of it, the open weed locator is trying to bring those two things together. So what it is, is just really cheap off the shelf computer a Raspberry Pi minus E and not that not the PI version. It's a small embedded computer. And what you can do is run simple algorithms on that that just at the moment, pick up green in a
fellow paddock. So fallow weed control is very important for moisture management and nutrient management are preventing that weed seed set for subsequent seasons. So what we thought would be a good step is developing these open source camera based fallow weed detectors. So they have a camera they have this computer and then what the next part of that is actually mapping each of those detections where that green was strong. So wait, just texted that grain from the camera feed.
Then it activates a relay, which you can then connect to any sort of solenoid it could be a hydraulic solenoid in the case of target and tillage or even a solenoid on it. On a boom sprayer to activate a, a nozzle, then you can spray away. And so all this code and all this hardware and all the results will be published open source that have been mentioned a few times now. So people can really contribute to the development of these weed locators and all this different sort of technology.
And that will mean that people can improve it and give feedback and really, hopefully, speed up the rate of adoption and the rate of advancement of this technology so that more people can can start to use it.
That's awesome Guy. And we have covered a lot today. But is there any other final thoughts that you wanted to leave with us before we wrap things up? Just that
I think this technology is really opening doors to our engineers exciting things, and having that feedback Avenue through open source and this community development around data and hardware for site specific value management, I think will really be an exciting avenue for improvement in weed control in Australia, and also potentially around the world.
Definitely. Well, thank you so much for joining us, all the way from Texas in the US via the magic of zoom, we really appreciate it.
Thanks so much having me Jess, and yeah, all the best down there in Australia.
Thank you so much to guy Coleman there for giving us such a great overview on the weed recognition work that he's doing. And pay as was mentioned in the interview, it is encouraged for farmers to get involved and take photos, which they can upload and provide to, to guy and the team so that they can include it in their machine learning open source data pack. But obviously, there is a little bit to it, and he needs you to take photos in a specific way
for them to be useful. So I'm gonna throw the links to all of you guys, Twitter and the website that you need, so that you can be across how to help with that project properly. But yeah, it's exciting, isn't it fate,
It is exciting. And look, there are guidelines, Jess and there's a specific way that he wants the photos take . But I can tell you that prett quick, we did a bit of this nd just had a selfie stick nd just holding the using your i hone or whatever smartphone you ve got, and just holding it certain height above the crop certain angle and click, clic , click click, you can capture lot of images really fast. And the theory is that it you now, it is actually very easy to
do. And if we do get farmers oing it and get 1000s of images ploaded then down the track, we' e going to have some algorithms hat can detect and see and spr y those weeds in those crops. So a great initiative that his has happening and just rea ly good that it is open source nd that it's going to create al orithms for all sorts of ways, ot just the probably the most pr fitable ones, if you like this o viously going to be the ocus of companies that get i
to this space. It just means that we could get algorithms for the weeds in all sorts of si uations at farmer
Yeah, so we definitely encourage you to get involved in some community science. It's really cool. And yeah, you can be part of the future of weed detection technology. So that's a cool thing to be a part of. But the that is our podcast for today. And I just wanted to give a big thank you to our guests once again to conniver farmer, Sam Eastwood and of course University of Sydney precision weed control, PhD guy Coleman, and we've got a few things that I'd like to keep you in the loop
on as well. So we've got our next regional update, which will be out next Monday, and we're heading to the southern region and hearing from our southern extension agronomist Chris Davies. So make sure you're downloading that one. And we've also got some new content on the website for you to check out as well. We've got our article explainer on what to expect the weight Smart Week in Esperance, and we've released a programme now so you can check the
programme out in detail. I've got the link to that article, which has everything you need. And we've also got that new article we mentioned at the top of the podcast on double paraquat knock downs. I'll put the link in the show notes for that too. And we also have a new webinar recording. So if you couldn't attend our webinar last week on ryegrass management in the high rainfall zone, that recordings on the website and I'll provide the link for that
one too. That's with Dr. Krista Preston and southern farming systems James Manson. They provided the info and it was hosted by our high rainfall zone extension agronomist Yana Dixon, and make sure you get your tickets to weed Smart Week they early bird does end at the end of this month on July 31. And it is quite a bit of a saving. So at the moment the tickets are $190. And after July 31. That will go up to $250. So yeah, Pete definitely worth getting your tickets earlier rather than
Absolutely. Alwa s a great event, Jess. And don't forget to follow us on Fac book and Twitter. You can also sign up for our monthly log, the weed smart whip aroun all of our content from the last month is featured so it's easy to get the latest news. And we'll provide the sign up link in the podcast notes. Also, it ill mean the world to us if y u could spread the weight map m ssage by sharing this pod ast with your friends. We're g tting great listeners, but w
always want more. So please eave us a review on Apple odcasts. And it's also a mass ve help to get the word out.
Yeah, exactly. Pete, thank you so much for joining me and being the co host today for the podcast. And yeah, like we said, make sure you subscribe so you don't miss s next time. Thanks, Pete.
Thanks, Jess.