Jessie Oliver designs citizen science research projects so that participants feel engaged, make meaningful contributions, and have fun. In this interview, she describes some of the considerations for a citizen science project, and about a challenge she's working on in Australia: to collect data on the endangered Eastern bristlebird. If you're curious about getting involved in citizen science, or you want to know more the tools, organizations, and resources involved in these projects, don't miss this episode!
Jessie Oliver designs citizen science research projects so that participants feel engaged, make meaningful contributions, and have fun. In this interview, she describes some of the considerations for a citizen science project, and about a challenge she's working on in Australia: to collect data on the endangered Eastern bristlebird. If you're curious about getting involved in citizen science, or you want to know more the tools, organizations, and resources involved in these projects, don't miss this episode!
Cary (00:04):
This is episode 21 of Soundproofist. My name is Cary.
Phill (00:09):
And this is Phill.
Cary (00:11):
And today we're talking with Jessie Oliver, a tech design researcher, and international liaison with the Australian Citizen Science Association. She's been working on a citizen science research project for the bristlebird, which is an endangered bird that's hard to find in its natural habitat. And a lot of people aren't familiar with this species. So it's more challenging for them to identify it in nature or from audio recordings or in a spectrogram. And these are some of the challenges she's trying to address by designing the methodologies, to get people engaged and collect useful data.
Jessie (00:47):
And you might not think creativity is part of science, but I would argue it's at every stage of science in different ways, and we can engage citizen science much more broadly with creativity, because if you can put meaning to that spectrogram, then you're going to remember it better.
Jessie (01:11):
I'm from the US originally. So grew up in California, worked in Hawaii and New York and California. I've moved around. And then in Brisbane. So I've been in Brisbane about almost 15 years now, which is unreal, but it's off and on 'cause I've moved different jobs. And it's been interesting moving back and forth to cause I've really been lucky. I've had some really great roles and I used to practice as an ecologist and an environmental educator before I started into my PhD as a tech design researcher for citizen science. And I first learned about citizen science when I was actually at Cornell's Lab of Ornithology working as an educator, which was really fun. And then I spent some of my time talking with guests and teach them how to use eBird and things like that. So it was very fun.
Phill (02:02):
Awesome. We're big fans of that. I was wondering if it comes with the territory, I know that there's some people that, you know, they do their PhD and they study a specific species. So then they kind of get tied to a certain place, but that doesn't necessarily sound like the reason why you landed in Brisbane.
Jessie (02:18):
No, I was in Brisbane already working as any colleges actually, but I got really heavily involved in the formation of the Australian Citizen Science Association simultaneously in my spare time. And it was from Cornell that I learned of citizen science. But then as soon as I arrived back here was actually when citizen science started being discussed on some kind of a national level. And so it was pretty neat 'cause I arrived just in time to jump into helping to develop the association. And then currently I'm international liaison of ACSA, which basically means I'm on a lot of different phone calls and emails all days, all hours of day and night, and keeping pretty busy to try and just understand what opportunities exist around the world and representing ACSA in those different initiatives where it can.
Phill (03:13):
That's really awesome. I should also say full disclosure a few weeks back, I participated in some citizen science and did the climate watch survey for we're looking for spotted towhees here in the bay area and among other things, lesser goldfinches, American goldfinches, but I'm a total rookie. So they were just like, "just look for spotted towhees." I didn't see any, I saw a California towhee, but it doesn't count.
Jessie (03:40):
Hey, fair enough. Hey, they count -- they're a bird. You know, we appreciate them all. Now my question to you, do you know the call of the American Goldfinch when it's flying?
Phill (03:53):
I don't know its flying call. I don't think.
Jessie (03:56):
So. A good way to remember it. It often flies around the same region as your chickadees, some of your local chickadees in that area and the American Goldfinch says "potato chip" and the chickadee says "cheeseburger." So you'll hear those two words together, which is quite funny. And once you hear it, you'll never unhear it. It's pretty fun. So when you're looking for those guys, that's what you listen for. "Cheeseburger, potato chip."
Phill (04:29):
Potato chip and cheeseburger. Wow. That's gold. That's....
Cary (04:36):
So let's pause for a minute and listen to a couple of examples of bird calls that I've found on the Xeno Canto website. First here's the American Goldfinch.... And now here's the chickadee. And now back to our conversation with Jessie,
Jessie (05:09):
Just diving right into the acoustics. So I was always interested in, you know, how to engage people with learning birds. And that is what I was lucky enough to do running around sapsucker woods with kids, teaching them the calls of the chickadee for example. And they go "chick a D D D D, D," as well. And yeah, so I was always interested, but I hadn't really thought about sound in the context of engaging with it through a computer until I dove into the work that I'm currently doing.
Phill (05:43):
So speaking of that, this is the bristle bird. Is that right?
Jessie (05:46):
Yeah. Eastern Bristlebirds. So there are critically endangered species for Queensland, but nationally they're endangered. And the challenge with these guys is they're super, super sneaky and they're rare. So, and they live in really remote areas, oftentimes in this region where people don't tend to hang out a lot. So not many people around know their calls. So if we want to look at something like using conservation technologies, what do we have as our options? We have camera traps. Well, they hang out in really tall, thick... Grasses, which are pretty bunchy. And they make these trails that they run around in and things like that. So not so great for the camera traps tried. It didn't really work, but acoustics are great because you can hear the birds calling from a fairly far distance, much further than you would if they happen to walk in front of a camera trap that snaps a picture with any movement, you know, and using those camera traps, you end up getting a lot of grass that's blowing in the wind and it makes it really annoying to go through a whole lot of pictures of grass.
Jessie (06:54):
And even at nighttime, I would get caterpillars at the ends of the grass that would trigger photos being taken. So, yeah, so sound is really great because it travels and it can be particularly useful for certain species. But then the thing we need to think about similarly to camera traps is you get a whole lot of data really, really quickly. As soon as you put a sensor out into the environment and how do we get through that data? And a lot of people say to me, well, why don't you just have a Shazam for birds? And it's a great question. And it's a fair question. And the lab of Cornell is very much working in that space as are a number of different groups in north America and in Europe. However, the thing to think about in the case of Australia is it's a numbers game.
Jessie (07:42):
You need to have a lot of data to be able to train AI, to be able to do that. And wildlife is a lot more difficult because we don't have that huge database of information that you might have from Google or Otter AI and things that do voice recognitions for humans. So we don't have the data to train often times on the scale that is needed for Australian species just yet. And particularly when we're talking about sneaky species that people just tend to not know about their acoustics. So that's a huge first step. And whether you're going to plan on doing machine learning or not, you still need a lot of really accurately annotated data to train your AI. So regardless, there's always this challenge of how do we get those initial data sets, right? And with Bristlebirds, I didn't know, starting out whether people did happen to know their call.
Jessie (08:41):
And I started working with the conservation team and found that a few people did, but they had very, very specific situated knowledge that's very hard to transfer to other people. So looking at acoustics potentially offers a way that we could actually share what we know, send each other recordings, like, "Hey, do you know this call or that call?" But the technology doesn't really exist to either share information like that easily in a way that is fairly seamless and even fun because looking at sound, if you're just looking at visualizations called spectograms can get pretty boring pretty quick. So how do you make it engaging, but also how do you make it useful and all these other variables of what do people want to know? And often we don't know what we don't know. So we can't really answer that question until we start diving in.
Jessie (09:36):
And so those are the types of questions I ask as I'm working with both the bristlebird team, the people that are dedicated to trying to save the species through a whole huge variety of amazing conservation efforts. Their expertise are diverse, not necessarily in acoustics, but we try all the acoustic technologies -- listening to audio. And I explored, you know, what did they know? What didn't they know what worked for them, what didn't. And then I thought, okay, we're definitely going to need a whole lot more people on this because the team is very small and I don't want to detract from the amazing other suite of things that we're already doing by adding yet another thing they've got to keep track of. And looking through acoustics is time-consuming. So then I started thinking about what other groups might be excited to listen to sound. And I thought birders, you know, that birders would be a great way to go because they've already got acoustic acuity.
Jessie (10:34):
They've already tuned their ears to really listen to the sounds of birds. So they might be able to pick them up more quickly and that might be well true. However, they might not be as drawn to sitting in front of a screen to do it, given their tenacity for questing, to look for different birds and spend time outside. And they rely on a lot of contextual information. And I say "they," but I'm totally a bird nerd. So I shouldn't realize it, but I guess I didn't realize the importance of that environmental context in how we identify the birds and that the lack of it might actually be kind of a frustrating thing. So then I started to think, okay, well, what other groups might be interested? And I started recruiting people from the broader citizen science community to explore how they engage with information. And looking at spectograms looking at photographs of birds or distribution maps of different species showing where, what areas of Australia they reside in. And how do people relate to calls that have a species that may be, it's a really vocal species, like the whip bird, the Eastern whip bird? Most people know what it's called, but they may not know what it looks like because it's a fairly furtive, sneaky species compared to like a cockatoo that's out on our banisters waking us up every morning.
Jessie (11:51):
Right. So how do relate to those different things? And then how do we inform the way we design future tech, taking in all those different insights that I gained from the conservationists, the birders and the broader public that I worked with, and that's what I'm trying to work out now. And so I came up with some really interesting insights from all the amazing people that I worked with that I'm not writing up in my thesis.
Cary (12:18):
So you're really focused mostly on the sound, but you're also dealing with these birds that you probably can't actually see very easily because of they blend into their environment, I guess. And they're also very rare. So can you tell me when you talked about camera traps, like exactly how do those work? You said they don't work in this case, but how do you set one up?
Jessie (12:40):
Yeah. And it is a case by case basis. So I need to be clear and that's true for acoustics as well. So any there's lots of different types of what we call now, conservation technology technologies that help us support conservation activities, learning different things. Camera traps is just one of those options. It can work beautifully for some things, especially bigger things that don't live within a recruited area, but all a camera trap is, is basically a camera that you set into a location and it's set with different specifications. So it might take a picture every hour of a habitat. Or it might be triggered to take a picture when something moves. And so I tried the motion-sensitive ones because I thought, well, I want to get a bristlebird if they walk right in front of the sensor. Right? But then I got a lot of pictures of grass, and not a single bristlebird.
Jessie (13:36):
So you have to really know where the animals are going to be. And some people try to attract them to the camera to try and see if they're in an area, but that's not really going to work for a bristlebird 'cause what would attract them? Not really anything we have access to. So, you know, thinking around the behavior of the animals and how it can actually support what it is we're trying to learn is kind of how you pick your tech. But you try things and see what works as well. And so that's what I've done. And acoustics were really becoming quite popular because the, the hardware itself has become quite inexpensive in recent years. But I do get contacted very regularly by people in Australia saying, "Hey, Hey, I've got this audio recorder and I've collected all this data. Can you analyze it for me?" And I'm like, well, no. Right? Because I need to know, well, what do you want to know? And machine learning is difficult. You've got to analyze all that information to understand what it is you want to learn. And what it usually takes is like for example, some PhD students who sit there and hand annotate everything and get hundreds, sometimes thousands of calls, depending on the characteristics of the call and the broader environment.
Cary (14:59):
So with the bristlebird project. First of all, you need to recruit some people to be citizen scientists. What are their assignments? What do they do?
Jessie (15:10):
Yeah. So to be clear, I don't have a deployed project, if you will. Because my role as a tech design researcher is actually to understand what do we need to design the tech for the future? What I have published so far is four papers that are exploring the different aspects of what is needed and why these certain things are needed. So for example, I worked first with the bristlebird team and I really had done ethnigraphic work to understand how do they actually work as a team and what technological advances could we create to augment what they're already doing. But then I also trialed new interventions with this sensor and then looking at that data through a website to see what are the challenges and what are the potential opportunities. There were four main lessons from that. They need ways to exchange information. They need ways to learn calls through, you know, being able to actually gain the insights that they need both to translate what they know from outside to a visual spectogram.
Jessie (16:22):
But also if they've never heard the bird, how do they gain that knowledge? How did they exchange information between each other, but also to be able to recruit broader interests given how busy they are. Right? So then I started working with those birders, cause I thought, well, I need to understand what their baseline knowledge is about sound and more broadly and how that relates to use of tech, right? How do their practices relate to digital citizen science, for example. Because in this instance, we're asking ...if you're looking at camera trap images or you're looking at audio, you're going to be analyzing data. That's already been collected in this instance. So we need to think about what do people already do outside naturally. And then how can we kind of digitize that to be interesting or compelling, or do we need to go towards a whole different group of people?
Jessie (17:18):
Because we don't know, people have had a couple of projects that have engaged folks to look at bats and whales and their calls online. But relative to camera-trap projects, they've had kind of low participation and that's okay. You don't need to have everybody loving and engaging in everything necessarily. But when you're looking at critically endangered species that we know little about and camera traps don't work that well, time is of the essence. So if we can design technology to be more compelling, more enticing and more informative, given how people interact with stuff, then we have a lot better potential to actually inform, you know, how to support bristlebird conservation. How to develop these algorithms, but also just to expose people just at large to birds they might not otherwise ever be able to hear or see.
Phill (18:21):
That's super interesting. And well, a few things. One is I think I did participate in that bat ID thing or some online system like that, where it shows you a spectrogram and it gives you a small tutorial. Like this is a generic kind of bat call for these species. You'll see bugs in this frequency range and these ones. And you kind of go through and help identify it. So I guess in the broader picture, based on what you've said — and then here's the very naive take — because I think you've pointed out all of the difficulties in this process. But like if we wanted to have a magical app, the Shazam of birds, you're saying the first we know how to do data acquisition, right? You put up field recorders, and we record them. But then we need to annotate this data. So this is where naive me could be like, well, "can't we just throw that audio up?"
Phill (19:06):
And then you have your citizen scientists come here and they say, "Boom! Cheeseburger, cheeseburger, potato chip. And, Shazam.com. We're ready to go." Right? But obviously there's a lot, you're identifying all these difficulties in that process that we can't just do that overnight. And also I find it very interesting how you're talking about different cohorts of citizen scientists. Perhaps that have different skill sets where it makes sense. Once you say it, that the birders are maybe more... They don't necessarily want to sit in front of a computer all day. That's not what draws them to this activity, but then there's gamification aspects where maybe you draw in these more indoor people to outside things or potentially other sides of the world. Like you're saying bristlebird. So this is a very interesting and complex problem as you've laid it out.
Jessie (19:54):
I think just to reiterate with the birders, it's not to say that they wouldn't. But we need a lot more research to understand what their motives would be. Right? So I know from my discussions with the amazing people that I interviewed, that they're like, well, I mean, it does take a lot of practice to remember the calls. You've got to do it regularly. And, and it's often a very collaborative process. So then I started thinking about collaboration and I know some birders love events., Like the Twitchathon, where you run around for 24 hours and you compete in teams for what team can find the most species. But then some people are more adverse to competition. So even if you're thinking about gainful or playful designs, you need to really understand your communities and thinking around, I mean, I could have done a whole PhD on just birders, right?
Jessie (20:46):
But I wanted to set a foundation kind of for very exploratory work to say, Hey, look, birders are great, but here are the boundaries of what they might consider within. And here's what we need to think about more. And let's think about what other groups might also engage with this process and why and how, and so just to hear with you, for as an example, when I was interviewing one gentleman who was a birder, I showed him a spectrogram. He'd never seen one before. So that wasn't obviously knowledge he was already had, but I said, "you know what species that is?" Without even playing the sound. And sometimes people could work it out if they were familiar with sound. Even if it was through music or something like that, they could interpret and be like, "oh, that's an Eastern whip bird." And I was kind of blown away because there's no way I could have done that, but I'm not musically trained or inclined.
Jessie (21:40):
So just understanding how people process information and integrating that into design can be really important, but even more broadly than that, how do we have fun doing it? And one fella said to me, "oh, that looks like car tracks. I have no idea what the bird is -- but doesn't it look like car tracks?" And that was a real moment of serendipity for me. I remember driving home thinking that was an awesome interview, but that one moment made me think of Rorschach tests, where people look at ink block paintings. They look at ink blots, and they're kind of asked to conjure up what they see in these images. And I thought, well, wouldn't that be cool if we did that for spectograms and I just really started thinking about it. So then I integrated that into my next iteration and my prototype, which I showed the video to you of.
Jessie (22:28):
And that's where I got the idea of interpreting spectograms creatively by sight and sound. And you might not think creativity is part of science, but I would argue it's at every stage of science in different ways. And we can engage citizen science much more broadly with creativity, because if you can put meaning to that spectrogram, you know, it makes you think of the Eiffel Tower or something, then you're going to remember it better. Which then actually might feed into you, annotating it more accurately because it's more memorable. And a lot of times people would very much attach their own personal experiences to these things, to make them meaningful. And I think that's a really cool insight and you very rarely see discussions around tech design and citizen science with things like creativity. It's mostly around how do we get accurate data? How do we get as many people engaging as possible?
Jessie (23:28):
And I would argue that you don't need as many people as possible because a lot of the times certain small numbers of people do huge amounts, but also we can think much more broadly about, well, what does it mean to have accurate data? If it's more memorable data that could make it more accurate, but how do we make it more compelling and enticing and fun to explore? But that's what I mean when I say I'm a tech design researcher, I'm thinking very broadly about the implications of this stuff. But not really going to say, oh, this is all generalizable to every species on earth. No, not at all. I very much studied in relation to a particular species with specific groups of people.
Phill (24:14):
That's super interesting. And I'm also kind of blown away by that insight of creatively interpreting a spectrogram, you know, especially in a citizen science, 'cause that occurs to me as being a great opportunity for what they call the wisdom of crowds. Right? Where kind of like memes on the internet, where like the crowd gets to the main idea of the gestalt of what this is. And maybe it's not obvious to everyone, but once you tell it to them, it clicks. That seems like there's potential for that in that as you pointed out, it's engaging them creatively, not in this necessarily like technical way or like some other like niche kind of, you know, you've gotta be a nerd about this or that. And much more like casual. That's a very interesting insight.
Jessie (24:57):
Yeah. And I think most citizen science is very visually driven, and we are visual creatures for the most part. But I did find there are a number of people that I interviewed that were like, "now when I'm birding, it's all about sound." And they may not know. I think there is a misnomer too, that you must know the calls of the birds out in nature. You know, if you're a real birder... And that kind of thing. But a lot of times actually when I'm teaching people to bird, the first thing I have them do is to just close their eyes and listen. Because if you start to tune your ears to even recognize where the birds are by sound — no matter what species they are — we don't care about that yet. Right? You're not at that level or whatever. So engaging people acoustically is not done very often in citizen science either.
Jessie (25:47):
And I think there's lots of opportunity to do that. And I think it could also allow communities to drive their own projects. If we had a platform that allowed for them to easily engage with calls. And I interviewed some folks that have been involved kind of behind the scenes and helping review audio as volunteers for threatened species. And I asked them, "well, what kind of barriers did you encounter?" And sharing information is huge one and what you were talking about too. It's also known as collective intelligence and social media can play a large role in this too. And I actually tested my Rorschach idea on Facebook. First, I just chucked this spectogram out there and I said, "Hey, friends of mine." And it was just who I was connected to, but I said, "Hey, how would you call this?"
Jessie (26:40):
If you think of it like a Rorschach test of what it like creatively and just the way that people engaged with that. Or if I chucked out a call I didn't know, they would tag other people that they knew that were experts in X, Y, or Z. And next thing I knew I had this huge wealth of ideas and information. And in fact, I did publish some of that info in my last paper that I think I sent to you. But, about the posting on social media and people drew over the spectrogram to demonstrate what it was they imagined and then posted it back to share with the group. And it was really cool. And that was just trying it before I put it formally in a prototype that I was going to test with a number of people in a more formal controlled setting. So it was quite good. A lot of fun.
Cary (27:31):
The thing I got the impression of is that you had -- I think you even mentioned it -- you had no knowledge at all of bristle birds before you got involved in this project.
Jessie (27:39):
Well, I knew of them in a broad sense. Yeah. I knew of their plight. I mean, I knew of them as a species because I was involved with our local bird life committee. And, you know, they often brought updates to us to say, Hey, this is what's happening. And they're looking for funding and that kind of thing. And I knew that they had hired a detector dog to help locate them in different regions and that they were going to bring it to Queensland briefly and have a go at trying to find them here. And I was thinking at the time as an ecologist, you know, I was thinking, wouldn't that be awesome if we could actually validate what the dog is telling us, because you don't necessarily see the bird. And it's not like when you're looking for koala with the dog and they find a pile of poo or something like that. You often don't see evidence of bird in particular, unless you happen to come upon a nest, which you kind of don't want to do. Particularly if it's near breeding season, because you don't want to spook them.
Jessie (28:36):
And they're quite a sensitive bird. And so, you know, they only, and they're very strict about when they do these things to make sure they don't disturb them to be clear, but I just thought, wouldn't it be neat if we could put sensors out and then find the calls, but I was completely naive to how difficult going through the audio really would be. And I went and recorded captive birds to understand their calls first, but then not as an ecologist, I was thinking, "I wonder if the captive birds differ from the wild birds in their calls?" We don't know. These are all open questions. And I never had time to go through those huge amounts of audio that we have, and the systems that we have to do, it proved to be a bit difficult for people that I tried to recruit to help out with that.
Jessie (29:25):
But it was also just because there was no easy way to help people learn, okay, this is what a bristle bird sounds and looks like. And you know, it just was not an easy task to do so, but it revealed a lot in terms of tech and thinking about what we need in the future. So all of those exercises were not in vain. They're just now informing, okay, this is actually a much bigger issue that could actually be useful to, you know, a variety of species, but I am sensitive and cautious to not over-generalize, it would need to be studied in those different contexts to know for sure. But also it's important to recognize. And another thing I was totally naive to is that it all depends on both the environment and the species that you're looking at together. Because it's very much a holistic system.
Jessie (30:21):
And if it's a noisy environment, it doesn't matter how crisp and long these animals call. It's going to be trickier if it's really windy, if it's rainy all the time, that kind of thing. If there's lots of cicadas calling and other things that share that same bandwidth as the call that you're looking for, that's going to be difficult. Whereas if you happen to pick a species that calls late at night in a quiet desert, it's going to be a lot easier to build recognition software for something like that. So, I didn't understand the complexities of any of that. And I probably would have asked about a Shazam for birds before starting my PhD as well. So to be fair, like I totally thought about it. Like most people do when I started. It was only when I started diving in and learning about these bristlebird calls.
Jessie (31:11):
They have a huge amount of versatility in their calls, which also makes it very difficult. Their calls are super short, which makes it difficult as well. What you want is you want long consistent calls and habitats without lots of background noise, and that doesn't fit any aspect of bristle birds at all. So I picked a really challenging species, 100% naively. But it turned out to be great for a tech design research project because had I picked an easier species. I don't think I would have been nearly as pushed to be creative as I was to come out with useful outcomes.
Cary (31:53):
Interesting, also -- it sounds almost like in this case, it's almost easier to train the detector dogs than it is to get humans trained up for this particular project. You mentioned there were detector dogs that were in use already.
Jessie (32:06):
Yeah. Well, detector dogs are great for certain things. And yeah, definitely, but I mean, it takes years to train them and they only work for a few years. It's very intensive work. So I'm no specialist in that side of things. I didn't actually work with the dogs directly at all, but I think they compliment each other. If we had the systems to go through that data would have been amazing to be able to say, Hey, look what this dog is telling us is accurate. I mean, we're pretty sure that that what she is telling us is, 'cause in a few instances there was a little bit of evidence and things like that. But you know, it's just still always good. And if I'm thinking in my former training of an ecologist, it's always good to have different ways of validating what you find. So it's always a good thing to have multiple tools in your toolbox.
Cary (33:01):
Can you describe what are your ideal tools for the future? Like if you could build ideal tools, what might they look like?
Jessie (33:09):
Well, I am so glad you asked me that. I think there actually does need to be a lot more research before I would know for sure, to be honest with you. If I'm fully honest, I mean, I'm tip of the iceberg here. But I would love if we could create a system that was kind of based on the bristle whistle prototype that I created and you can create a prototype in PowerPoint. Amazing. I had no idea PowerPoint could be used like that, all interactive and things. But I explored it with different people and it had lots of different playfully and gainfully designed things. Playful is kind of open-ended exploratory that Rohrshach-focused stuff, whereas gainful is more structured and goal oriented tasks and had a number of those different activities to see what resonated with people. And now that I have a pretty good sense, at least with the few people I worked with because it all is very much exploratory.
Jessie (34:08):
It's not about numbers necessarily. It's more about gaining a rich understanding of people's struggles and what they loved about it and what they didn't. So now that I understand that acoustics are really cognitively demanding. And that was the biggest hurdle I found. So what we need is systems that really give a variety of different tasks. And they mix them regularly so that that cognitive demand is reduced. But then allow people to share their creative ideas and see, Hey, if we share these ideas and we get that collective intelligence idea going, they call that folksonomy. When you have people naming things like tagging them with their own creative names. If they do it over a long period of time, they might develop kind of a system. If we think of something like a Flickr account for photographs, you know, there's kind of a nomenclature that naturally developed.
Jessie (35:07):
And what if we see that there's trends in the way that people see objects in these calls, then we might actually be able to use that information as a way to tag the calls later in a more systematic way. And then you could know a whole heck of a lot more about not just our bristle birds here or not, but do their calls vary between habitat a or B, do they sing different at different times of the... I mean, there's just so many questions we could ask if we even knew the basics. So I think for me having a system like that, that really allowed for creative thinking to be integrated into the scientific part of it to an end also to have plans to directly inform conservation action would be ideal.
Cary (35:59):
So is there, right now, any sort of a web-based -- as you mentioned -- sort of like a Flickr for bird calls, that's crowdsourced and tagged? Is there something that exists like this now?
Jessie (36:10):
I'm glad you asked that too. Great questions. So Xeno Canto is an example of that and it's a static kind of website where anybody can go and make a recording and then they can upload it to Xeno Canto. And then if they don't know what the call is, they can tag it as a mystery call. And then if somebody wants to, they could come and actually identify it. And those calls are used quite a lot. And there's also other apps now that have augmented to be able to include calls. So even if there's the global eBird platform for birders, they go outside, they observe their birds, they enter their checklist. They might add some photographs. They can now also upload audio recordings, that's Cornell Lab of Ornithology. And then it goes into what's called the Macaulay Library and Cornell's Macaulay Library and Xeno Canto libraries have been used in a number of different ways for different studies of various species around the world.
Jessie (37:13):
So those are two bird examples. There's other ones in Australia, for example, where people are going on, recording frogs on their phone, and then they're sent into a frog ID and then a scientist will validate the calls and identify, confirm what they are. And then that information has informed research about frogs across Australia. So they're out there, but I think the analysis part of it isn't really in existence just yet. So Xeno Canto is probably the closest, but it doesn't really guide people in any way for how to actually learn bird calls or analyze the bird calls that are uploaded. So it's just expected that birders will come that know the calls and they'll identify them. But what I want to do is reduce the entry barrier that allows lots of different people to learn about their local environments or far away environments. Because why limit ourselves?
Phill (38:20):
So that leads me to another question. And are you aware of any tools that are in this kind of bird song learning or like teaching people to understand songs? I'm aware of maybe Lark Wire. I've done a few, like Audobon kind of things, but I will have to say -- I guess it's more my critique -- that there's kind of like, "here's three examples of this bird vocalizing" and that's kind of like, "here's three words in French. Is that person speaking French, Portuguese or Spanish?" And you're just like, "uh, if they happen to say those words, then it's good"... but right, anyways. The main question is, are there tools that you've seen that do help educate people on bird calls like this?
Jessie (39:01):
I think you tapped right into where my critique of the current moment, and it's not a critique so much as it's kind of a newer area that's being looked at. So the technology is likely to expand in rapid gears if people like me keep pushing that agenda forward. But I would agree with you in that there are loads of field guides out there, right? So you can download your local Sibley Guide for example, if you're in North America. And it's now a field eGuide for your phone, and that will often include calls of birds. If you're new to birds in North America, it's pretty awesome. 'Cause Cornell Lab recently came out with Merlin ID, which is much more geared towards beginning birders. And now they also include bird calls in there as well, which is great. They also have, like I was saying before, some regions of the world have more calls already included and they've been able to develop some AI and Merlin ID has actually done that. So if you record a call, it can give you a sense for what it might be in terms of species calling, which is pretty awesome.
Phill (40:15):
The Bird net application.
Jessie (40:16):
Yeah, so it's the Cornell Lab of Ornithology and it's Merlin ID.
Phill (40:22):
Okay. So those are separate. I assumed that was a Cornell thing. I attribute everything to them, but that's my limited knowledge.
Jessie (40:30):
Yeah. So eBird kind of is for people that already know their birds and their calls and they want to go out and they want to contribute their observations to citizen science. Whereas Merlin ID is more for like, you're a beginning birder, but ["your cane"??] is anything to learn and you want to go out and figure what things are. So that's kind of the difference I would say. And I imagine if they haven't already, they'll probably kind of augment more entry from one to the other in due time. But I don't know where they're at with that. When I was at the lab, they were at the very early stages of training, the, the visual AI part of it. And so I was helping to annotate images, but that was a long time ago. So they've come a long way since then. And it's really awesome. I just saw last week that...
Phill (41:20):
They got a huge grant, right?
Jessie (41:21):
Well, yeah. I don't know specifically about grants, but I just noticed last week in particular, I got the announcement from their newsletter that they now have the ability to actually identify sounds. So that's really, really cool to see. And I hope we get that someday in Australia, but I would still caveat that it only applies to the 400 most well-known species, if you will. And so you're still gonna, we want to be sure that as our technologies advance, we don't kind of cost out the species that are not the low-hanging fruit. Right? So we don't want to miss out on our Bristlebirds just because we don't know anything about them and they're hard to know about. Right? So we need to maybe think about ways of extending what we know for even species that are more difficult. So they don't kind of get left out of the discussion, really.
Phill (42:21):
Totally. And I would, maybe I would also argue that doing what you're doing with Bristlebirds potentially leads to way more knowledge than like let's identify a rock pigeon or something very common, right? Like you're solving more difficult problems that could have cascading implications for other researcher or work or technology.
Jessie (42:42):
Yeah. I wish I could have a whole team of people work on this, because it really does need a lot of research to really understand how to unpack all this stuff. My outcomes of my project certainly were very different than I envisioned. But I also should confess coming from an ecology background and transferring to design thinking was a very challenging transition for me. And I didn't even realize why for the longest time, but the training and the way of thinking that you're trained to do, you know. Science is about understanding generalizable trends. Whereas design is really about understanding the nuances of what people need to interact with information. Once I got it, I really see that they compliment each other quite well and that it really does take both ways of thinking and other ways of thinking and just not limiting ourselves to think too narrowly in one way or another.
Cary (43:49):
I was wondering about also is there are a number of different citizen science organizations. I'm just starting to learn about some of them. And I'm wondering how much do they cooperate with each other or share information or share citizen scientists? You know, for example, SciStarter, Zooniverse, EU Citizen Science Organization. There's probably many more.
Jessie (44:14):
There's many, and I'm happy to say there's associations popping up all over the world. So now we've got one developing in Africa and there's definitely already one in Asia. And there's one that's looking at all the Spanish and Portuguese speaking communities. And it's really an exciting time to be heavily involved in the space. There is cooperation to a point, but there could certainly be a lot more, a lot of it's just capacity. Most of these organizations are volunteer. And so it's a matter of how much time do we all have to actually network and do different things. So those are the kind of high-level associations, if you will. And there's also a global partnership that's forming that I've been heavily involved in, and there are discussions around the sustainable development goals and how citizen science can plan to that. And I've been involved in those types of discussions and with platforms like SciStarter.
Jessie (45:13):
So SciStarter is a global project finder. So if you're interested, they are North American-based and they do have the largest proportion of their stuff is North American. But knowing of them and knowing that Australia's global or sorry, Australia's biodiversity repository group called the Atlas of Living Australia was interested in having a project finder. I was like, "Hey, we should start some conversations." And they have what's called an API. And it allows for transfers of information about different projects, in between projects. So there is sharing where possible, but you know, there's always room for more. I think the biggest thing though, is there needs to be support for that to happen as much as possible. Yeah. There's lots of collaborations. I mean, I decided to take on the international liaising role because I started to see like, Hey, there's all these amazing discussions happening and we need to make sure we're exchanging information and you can look at the publications to see there's a lot of global collaboration among the practitioner community.
Jessie (46:21):
It tends to be academic, but I'm also very interested in ensuring that our community groups, I do a lot of stuff locally to share information with our local groups that are interested in different things from gardening to you name it. And I'll give a talk to them if they ask me. And they have the support for it. But being able to give them the means to do what is of interest to them, but then maybe think about how do we scale it so that it can feed into bigger initiatives is really in need of a lot of thinking. And I'm really interested in ensuring that whatever we do is contextual on a local level. So it's important to the people and empowers them to learn what they want to about their own natural environments or otherwise. And citizen science is much broader than nature too. So understanding what is of interest to people.
Cary (47:16):
That's what I hope to do through this podcast. Increase people's awareness of how they can contribute to citizen science projects and learn about the different organizations and how to get involved.
Jessie (47:28):
It's a great goal. And I totally agree, but I think also your point to sharing, I mean, we don't need to reinvent the wheel in a lot of ways. But it's also important to realize we've only just tip of the iceberg moment, right? So just because one person is doing it one way, doesn't mean that there's not room to do things in a more creative or different way for a specific community. But then how can we design our technology so that the information we're gathering from those is shareable and useful to both groups, right?
Jessie (47:58):
So that's, that's a design question actually, because you need to understand the different communities and what their respective goals are. And then thinking about how to design the technology to harmonize that information is really, really important. And it's really a cool space to work. And it's got its challenges for sure, because every community is different, right?
Phill (48:28):
I think that's a really good point that ties back into your earlier comments, around, you know, communities deciding for themselves what kind of research they need to do. And maybe that's outside of nature or what kind of citizen science. And, you know, when you were talking about this collaborative effort of one community doing it their way and sharing the knowledge they gained with other groups, this occurs to me that has applications beyond science, even, or citizen science and all the social justice movements. We see organizations, organizers, workers like unions, a variety of different groups assembling around the world and how are they're communicating and sharing knowledge. And not overgeneralizing from some top-down approach, but like a kind of more grassroots, bottom-up method of what works in this community and a collaborative effort of sharing these ideas. And it seems to me that you're at the crux of how do we design this technology to allow this kind of communication to happen.
Cary (49:28):
I'd like to thank Jessie Oliver for talking to us today. The practice of citizen science is really interesting and we've covered some apps and some projects in previous episodes of Soundproofist. Today, we talked about some challenges of project design and about engaging different groups of participants about collecting data for an elusive and endangered species, and about collaboration. I'm going to put some links in the Soundproofist blog. So if you're interested in learning more about this topic or getting involved in a project, go to soundproofist.com and we'll give you more information there. Thanks for listening.