Using low cost microphones to monitor biodiversity

Dr Casey Woodward, Founder of AgriSound, speaks with Thomas Slattery about the company’s use of low-cost listening technology to better understand biodiversity in the field and help farmers and growers to both target interventions and protect pollinators.

AgriSound’s technology utilises bioacoustic algorithms to produce a user interface that shows deficits and offers high-level recommendations. With advice from growers about challenges they face relying on manual observations and what they would want from an automated solution, AgriSound has developed strategies to enhance insect levels, map impacts of agrochemicals and support national governments with data-led policymaking.

By combining acoustic technology with environmental sensors, AgriSound has developed a range of innovative products that can aid in maintaining an optimal pollination environment, necessary for not just the survival of future pollinators, but for farms, gardens, and the planet.

Transcript

0:04

Hello and welcome to another episode of Seedling Sessions. Today we're speaking to Casey Woodward of AgriSound. Hi, Casey, how are you doing? Hey, Tom. How are you? I'm very well, thank you. So really excited to have you on the podcast. I've been following AgriSound for some time now and I've been really looking forward to getting the opportunity to have a chat to you about what you're doing and also the future plans for the business. But just for listeners of the podcast, would you mind doing a little introduction about yourself, AgriSound, and a little bit maybe about your mission? Yeah, definitely. So those of you who don't know me, my name is Casey. I've been in the Agri tech space now for the last five years or so, working in livestock innovations, and then over the last couple of years set up a company called AgriSound, which aims to use low cost listening technology to understand biodiversity in the field and help farmers and growers to really be able to target interventions to protect pollinators where needed. So when you say obviously the listening device themselves are quite low cost, two things that'd be interesting to me is one kind of what's the recommendation for a producer or whoever the customers are using? What are the number of devices per hectare could they expect needed to use? And then do you expect that to be kind of the whole farm? It's a really, really good question. It's a question we get asked the most probably, and I think it really depends on what you're trying to achieve. So if you're an apple grower and you've got hundreds of hectares of orchards, you might want to optimize your pollination in those orchards to make sure you get the best yields and the best fruit quality. And we know that suboptimal pollination costs around £5000 per hectare in the UK, so it's pretty big challenge that growers are facing. So we would recommend kind of two to four devices per hectare. In that case, it would be around about a couple of hundred pounds. But conversely. If you were an arable grower or a dairy farmer and you wanted to make sure that you were protecting pollinators. You might want to monitor your hedge rows. Or you might want to look at any forms of the habitat that you have on your farm was operating in the best possible way. Or you getting the most return for the investment that you've made. In which case it might be a much smaller number of devices placed in those areas and you're just tracking very small areas rather than monitoring a whole field. In which case it might only be a handful of devices in that case. So it really depends on the question you're trying to answer. Yeah, and I'm assuming that kind of the meat of the innovation. Obviously, even with low cost listing devices, as we all know, once you get them out into an agricultural setting, I'm sure there was some work required to make them robust enough. But my assumption is that the real meat innovation is in the algorithmic and data processing. Absolutely. So we use some standard components. We built our own hardware, but ultimately it's effectively a low cost Alexa that you might have in your house, applied in your garden or field. So it's a series of microphones with some environmental sensors. So the idea being that we can use bioacoustic algorithms, particular algorithms, which interpret that sound information to tell us what type of insect is present and the abundance of that insect, alongside some of the factors that might be contributing to why your pollinator levels are high or low. If the temperature is very high or humidity is very high, that might change. So it's a case of trying to listen out to the sounds of those insects, but also being able to interpret that sound file in a very scalable way, which means that we don't need very high powered computing resource or we don't need to get the device confounded by lots of different background noises and try to strip all of that out. So you're right. The algorithms that we developed really are the core IP, as it were, of the devices. That's interesting. Obviously, you mentioned earlier, not only working across a number of different kind of agricultural sectors, but also what people are trying to achieve. And I think one of the things that we see a lot in agtech, particularly in sort of the more cutting edge stuff, is a lot of different ways of capturing data. But then sometimes there's a bit of a gap between them giving the end users actionable insights. And it sounds like that's something you've already spent a lot of time thinking about, how to take all of this data. And actually then, depending on the particular use, case, deliver easy and actionable insights, it'd be interesting to hear a bit about that. Yeah, absolutely. So we spent a lot of time talking to growers to kind of look at what they were doing around biodiversity monitoring as is, and then what would they want from an automated solution? So we know that most growers tend to rely on still on manual observations. They're bringing in volunteers or consultants or ecologists to come out and stand on the farm for a few days to really understand kind of what's going on. And clearly that's fraught with challenges if the day is a bit overcast, or whether it's poor or too windy, that can really skew the results. And then it means you've got to bring someone back out and incur more costs. So it's a problem. And what ultimately growers want to know is kind of where are the black spots on the farm? Do I have sufficient pollinator coverage? Yes or no? Very binary. And if no, where are those black spots and then what can we do to plug those gaps? So what we've done is we've developed a very simple user interface which ultimately revolves around a heat map which shows kind of where the devices are, what's the activity happening in those areas. And therefore, if you have got an area which has some form of deficit, then we give some very high level recommendations around short term being to move manage parameters there. So if Honey Bee Hive or Bumblebee Box but longer term, look at kind of how can we use more sustainable interventions like sowing wildflower seeds, create a new bee habitat to create local native pollinator communities that can ultimately deliver all of the good stuff in terms of precision pollination that's required, but also in a way that's kind of environmentally and economically viable and sustainable. Great. In essence, essentially you've got kind of 24/7 monitoring, heat mapping and recommendations which you can then work with the grower, their Agronomist, another company you can provide pollinators and advice on. That was some of the feedback we had quite clearly was from the growers perspective, they were already engaging in agronomist and they didn't see it as our job to be telling them what to do on the farm. And we take that to heart and we tell growers that we're engaging with now that we're a technology company, we're not agronomists. We'll show you where those deficit zones are, but ultimately it's your decision on how you want to deal with that. And if you want to make any recommendations at all, if you want to make any interventions at all, then that's completely up to them and they could do with that as they see fit. I don't want to put you on the spot at all, but have you got any kind of particular success stories that you want to talk about publicly? Yeah, absolutely. We're really proud to say that our first big project is kicking off next month with Innocent Drinks. So Innocent Drinks are obviously one of the smoothie brands and part of the Coca Cola group. And they came to us recognizing that there was a problem around getting enough pollinators across all of their fruit production estates in southern Europe. And so we're working them on a pilot project now to deploy some of our devices onto some of their apple orchards to help them evaluate different types of interventions that they are going to apply to remediate low pollinator levels, but also then use this as a case study to them, really get this technology used across their wider supply chain, which we're really excited by. Actually, I think for a company that's only been going just over two years now. I think the fact we work with such impressive brands like Innocent Drinks is something that we're really proud of, actually. Yeah, it's very impressive. But I think it obviously speaks to the value of the mission of the product and the growing trend towards sustainable and regenerative. What is interesting to me is that I've spoken to a number of ag tech companies over the last year who interestingly, has started to work more and more with processes and food brands. And what you're seeing is personally, I think what you're seeing is a consumer push towards more regenerative and sustainable products, which is thereby getting some of the more progressive food brands obviously always been one of those. I think they just got their B Corp status looking at ways that they can work further down the food chain and pre-farm gate. And it's really encouraging for me that we're starting to see this whole food value chain working together towards these kind of admirable goals. Absolutely. What's great for us is that consumers are really vocal about protecting pollinators. Everyone knows that bees are dying out and that presents a really big problem. And I think what's great is that the consumer is now making their voice really clearly heard and they're supporting the brands which are doing their part to protect bees and protect biodiversity. And from our perspective, obviously we're really lucky that we're supporting that agenda and we've got brands that want to work with us to help them to achieve that and tell consumers a really positive story about what they're doing to help protect the bees and other pollinators. Yeah, no, it's wonderful. The nice place to move on is like I know you've got some pretty ambitious goals. It would be great to hear about some of that. Absolutely. Yeah. We're really going for the stars, Tom. We want to create the world's largest insect biodiversity database and the way we want to do that is to deploy a network of our listening devices across the globe. We're aiming at our first million devices over the next few years and then onwards and upwards from there, really. And what that will enable us to do is really take this big problem of the loss of biodiversity and the loss of pollinators in particular, but address it at a hyper local level. That's always the problem. We know that climate change is a big issue, same as as an example, but when we talk about how we solve it, we often get sort of scared by the big actions that are required. What we're trying to do is help growers and farmers, and even broader than that, into gardeners and construction companies and so on. So we help them take action at a really local level that can just start to create more areas where bees can be protected and native bees can thrive. And from that we expect to see the outputs being greater than the sum of the parts, as it were. Yeah. So obviously you're a UK company. What does the mapping of your kind of global geographic rollout look like? I'm assuming you've already started a bit and if you got any particular listings, for example, who might be able to help out with that. Yeah, definitely. It's still early days for us. Vast majority of our devices are forecast to go into the UK over the next twelve months, but we're certainly not closed off to discussions with people in other countries. We've been talking to plenty of people in New Zealand and in Australia around some of their almond plantations and fresh produce in those markets. We've been talking to people in the States as well, particularly around kind of big greenhouses and making sure that we can optimize pollination in controlled settings as well as in outdoor settings. And the imperative area is very strong because it's a very large commercial cost buying in pollinators is expensive ultimately, and so everyone wants to make sure they get enough bang for their buck. For us it's a case of I think we're staying close to home for the time being to make sure we've got the evidence base that really shows how the technology works and how much impact you can ultimately have and then I think from there we'll then start to roll that out. But we're really excited. I think this is just the tip of the iceberg for AgriSound and what we're going to achieve over the next few years. Absolutely. And I think it makes real sense to focus on pollinators and focus on a single market and really be able to demonstrate impact case studies on that. I've got a real passion for regenerative systems, particularly in agriculture, obviously, and of that biodiversity is a major kind of pillar for that. And interestingly, obviously around the carbon side we've seen a huge growth within ecosystem services. As much as a strange phrasing, carbon has been the big one up until now, I don't think we've seen the peak of carbon ecosystem services. One of the next big ones, I mean water infiltration is of course one, but another big one is biodiversity, and that's not necessarily just going to be around pollinators, which I appreciate a big indicator of that. So is there in the longer term, the potential for your devices to also listen out for a wider range of biodiversity? Absolutely, yeah. So I think we know that we agree pollinators are just the start and they are really important indicators species around the health of an ecosystem. But actually we can use our technology to monitor things like pests as well and make sure we're helping growers to use integrated pest management and reduce pesticide application where necessary. And so we've already started that work. We had a proof of concept project recently looking at how we can listen out for the coddling moth, which is a really important economic pest of apple orchards. And so we definitely want to look at that and expand that library of algorithms present to be able to recognize more pests as well as more beneficial insects as well but I think there's no reason why the technology can't be used more broadly than that either. So if we wanted to use the device to monitor birds in the environment, we absolutely could do that. If we wanted to monitor small mammals, we absolutely could do that as well. Anything that makes a noise, we can pick up and distinguish that noise by isolating the kind of the key acoustic features against the background noise. So if people have got interest in ideas of other use cases, then we'd be absolutely delighted to hear that and be open to working with anyone to take the device to the next level. Yeah, absolutely. And obviously, you said working outside of agriculture and you mentioned construction, and I wonder if there's a pitch that needs to be made to the HS two? Because I know that I'm aware that for that build, working out the impact of biodiversity across that construction is going to be a big thing. Probably worth reaching out. Yeah, definitely. We are talking to quite a few big infrastructure project kind of teams at the moment to look at how the technology can be used to help provide an indication of what's happening to the environment. And as you'll notice, in the UK, under the new environmental bill, all of these products have to show biodiversity net gain of over 10%. So it's quite significant. And I think what we can do is to highlight the areas where a landscape might be struggling and then suggest some interventions that can help developers get over that threshold. But then 10% should be the minimum. We could hopefully exceed that as well. We know that if you've got bees on a site, you'll then start getting pollination of hedrows that provide berries for birds and then the whole ecosystem benefits. So we think there's loads of great stuff that we can do by protecting pollinators. Absolutely. To listen to us. I think what's so exciting about the tech development like this is one of the core principles around trying to make progressions and sustainability and regeneration is this idea of measure to manage. And if you can't benchmark where you are and where you're going under all of these different things, then it's impossible, one, for it to be regulated, one for growers or construct people to understand where they're going right, where they're going wrong, where they should be focusing. And so a tech solution like this, which core is a kind of measurement and management tool, is so exciting and so needed. Well, look, it's been really interesting in getting a catch up on the project. I mean, it's incredible to think you've only been going just under two years and what you've achieved so far, is there anything else that you'd like to kind of put out there to listeners or any kind of collaboration requests that we might be able to help you with? Yes, absolutely. I think for us, we're a small company and we recognize our limitations in terms of we can't achieve everything alone. And so collaboration is really important to us and I think we definitely see ourselves as being a partner who can bring biodiversity elements to projects, particularly bigger projects, looking at regen farming as a concept or more kind of environmentally sustainable forms of crop production or even beyond that. So I think my message would be if there were companies out there that were interested in looking at biodiversity and they wanted to work with a partner that was used to working in partnerships, very collaborative, then definitely reach out, get in touch and we'd be delighted to grab a virtual coffee and have a chat, if nothing else. Amazing. Well, we are looking at doing a regen focused event at the end of March and where we're looking to bring technologists producers, farmers, etc, etc. To talk about this. And I think it would be really great to potentially have you guys at. That sounds brilliant. Yeah, brilliant. Well, thank you so much for taking some time. I know you must be incredibly busy as a startup. There's been a really enjoyable conversation and I'm really looking forward to following up and seeing how things go over the next, over the coming months.