[Sage] Welcome to our podcast. Today we're talking
about using automated acoustic
recording units to support
undergraduate learning by engaging students with original research
questions related to Bird Communities
and conservation. The Oregon
Community College Distance Learning
Association has research and
development funds available for faculty
interested in exploring emergent technologies that support
distance learning. Lane Community
College Faculty, Colin Phifer is here
with me today. Recently, he received
funding to explore bird monitoring
and automated acoustic recorders to establish a long
term dataset of bird species at Lane
Community College and the broader community. Colin is a biology
instructor and teaches both in person
and online courses. Welcome Colin! Glad you're
here with us today. [Colin] Oh, thanks for
having me Sage. I'm excited for it. [Sage] That's great. I am so interested about
your project. Sometimes we get
really unique projects that come through
our work group. And I think this
is definitely one of those very interesting and unique projects. So thanks for being
with us to share some information about it and talk about
your experience. I have a few questions for
you if you don't mind. [Colin] Absolutely. [Sage] Great. Well, first of all, call and tell
us what led you to become interested
in this project. [Colin] So I teach a variety
of classes at LCC, which is probably
no surprise to any community
college instructor. But one of the
classes I teach is an introductory general
education class called Birds of Oregon, the Boo class,
so to speak. And this is of course for anyone in
any discipline, any major if a certificate or if
they're coming for their transfer degree, And so this class
has a broad Everyone's coming in with no prerequisite,
I should say. And so need to I wanted to make
sure we had tools available that
were accessible to help students
learn about birds. Of all the biology
classes I teach, people get really
excited about birds, even people who think that maybe they don't know
anything about birds, or they took a class because that's one to
fit in their schedule. We've all had
that student. And then they
all come out, I think, learning
more about birds, but also about birds
in our culture. Birds are a part
of their culture and keep learning how
to identify them. And learning to
bird by ear, like learning to
identify birds solely on their call or their song is a real challenge
for most students. I think everyone
can learn it, but it's a lot to have that expectation that they can learn to do it in
a ten week course. And so these new automated recorders
are just with that. It's not just recording and listening and
playing it back. It's actually
doing the species identification at
the same time. So it's taking
advantage of AI, and now it's applying it to bird song
and bird call. And if you think about
what AI does a lot, you know, we call them like large
language models. Well, birds
have a language of repeated elements. So it's pretty easy from the I would say easy. I'm not a computer
scientist, but You can imagine taking that tool that helps
you autocomplete, your e mail response, which we have, or other tools that
helps you complete. And now apply
that to helping you predict what the next part of a
bird song would be and understanding that
what species that is. So this tool, they're from the manufacturers
called Bird Weather. Allow for species
recognition and identification
automatically. And that helps you
create a dataset that students
can now start learning the birds that
are in their area. So I think they're a
really powerful tool and something that's at I'm just really
excited for. I feel like I'm just
on the we've only been used them for one
term, so to speak. And so we're just
still learning how to use them even more effectively in
the classroom. [Sage] That is so interesting.
I have to say, once I realized what the sound of a humming
bird sounds like, I started being able
to recognize them, and they're so small,
right? So small. Yeah. [Colin] I almost sounds
like a large bee. [Sage] Yeah, and they have
a very unique sound. I'm certainly
not an expert on listening to birds, but I can relate
to this especially it's very interesting how students use their
auditory skills for identification because
a lot of time we use our vision, our touch, different
sensory skills. The incorporation of technology is
really interesting. Tell us were there
any specific skills, concepts or techniques
that you or your students learned
exploring this project? [Colin] So a great question. Let's start off
the technology, which is, I think it
kind of reinforce bird, bird song, bird call. Course, also helping
students remind them that not all birds actually have
calls and song. So that's an
important thing that we kind
of talk about. So one of the focus
of the class is on diversity of biodiversity, biodiversity, right? So what different
species they are in different landscapes
or habitats. And then one of the other parts
of the classes, we also talk about
conservation and talk about the
value of these tools. And these tools,
which are broadly we call this bio
acoustics, right? So this is actually
a really active area of research right now. And you can imagine being able to leave
recorder out in a particular forest or
a particular habitat and having it collect data long term and
learn a lot from that, versus if you were to
send a person out there maybe one time or two
times or three times, but they're only out
there for that 1 hour, those 2 hours that
they're there, and they're trying to
identify the birds. So it helps us think and develop a better understanding what's
happening long term, so we can collect
these long datasets. That's the idea
of this so that students are getting how can species
change over time. So for example,
my goal is to have my goal is I have 60 of these units with the
support of the grant, which was so
awesome of you. And so my goal is
that always have one on campus that's functioning and recording. And then have others
available for students to do small research
projects of their own. So maybe they
want to take that to a grassland
or to a wetland, and that's something
that we could leave to record
and collect dataset that
otherwise we wouldn't be able to do, right? So helping students kind of scale and think about questions they
can ask with bio acoustics is a really important
part about that. And these are kind of like an easy way and
introductory way that I can bring this really powerful
tool to students. I should say, and
I contrast it, which is In my class, I've moved away from
having students purchase a guide
book, right? Bird ID
identification book, just from an equity
perspective. And we usually use an
app called Merlin, which is works for Android and iPhone, which
is a great tool. In fact, that tool has
built in because it's free and open source maybe it's an open source,
but it's free. It's from Cornell
University. And so it has a
tool actually has an acoustic
identification tool built into that app, but you have to
use it in that moment when you are there. And so that's one
of the other things is we talk about kind of a threshold idea of
sampling going out to your community and learning what the
birds are there. Again, if the difference, if you're there one time or two times or three
times for that 1 hour, versus being
able to sample. Maybe I can set
it to record, for example, 2
hours, morning, 2 hours at dusk
and dawn for 2 hours and leave it
there for a week. How many more species
are we going to identify in a
week's sampling, versus just like three times
visiting the site. And the Merlin
app, I should say, that's on their
phone is great, but again, you have to be there in that moment to pull it out
and accessible. So this helps us expand beyond and let students, I think ask
different questions. Again, bio acoustics is a really hot
topic in biology, but it's also one
that I couldn't do in a 100 level general
education class without this tool
because everything else would be
too complicated. You have to learn software associated
with it. This gives me species
identification, probability,
and frequency, how many times did
we hear that bird? That's a really
cool, powerful set for students to
start thinking about questions they
can ask with that. I would imagine that it's highly engaging
for students, too, because reading
through your report, I don't know if I
understood right, but I saw a screenshot
of the app and I could see how it
would geo locate where on a map where the different birds were being
identified. For sure. That's
the cool part is it's a network
that you're joining. You have the option
to publicize or share your data.
But many people do. And so now you're
seeing this company, this network, which
is continental. So for example, you
click on and like, Oh, here I heard a Cedar Wax wing in my
neighborhood, and you can click on that and then you can see a map opens up a geospatial
map in real time. Who else is hearing
that same bird at the same moment, right? So this is a
really, I think, a cool part of it is that it continues to reinforce
students learning. So one of the goal
sill is like, Hey, help you learn that
bird caller yourself. So when you're on
the interface, You can pull up
and then you can pull up information
about the bird. And now you can actually see what we call sonogram, which is a visualization
of the bird song. That's one of the
skills. I do a bird bird by ear lab activity
with students. One way we do it is
help you learn to draw out and visualize what that sound might be. It's cool that we're potentially also
contributing to dataset that someone else might be using and
referencing as well. Yeah, I can imagine, and I would imagine
there's a lot of enthusiasts that are active with it as well
outside of education. Yeah. So these
tools have just kind of like I said, become I think
this last spring was the first
time I became available or
aware of them. And I should say they did offer an academic
discount. So I think they
are like $250, but they offered
an academic discount when I
requested one, so I think they were
$200 each unit. And then they
run on two or three D A batteries that you can use as well. I'm still learning.
For example, there is a tool. You can buy an
extension cores, so you can just
plug it in. And that's where I
want to go now is plug it and leave
it on campus. I'm trying to find
the right place. I think we're going
to host it at our Childcare Center on LCC partially
because that area is fenced and secured. So there's less likely
someone might take it. But also because I
do an activity with a childcare center
in spring where my students come over
and they share about birds during international
migratory bird day. So I think kind of a
cool thing about it. One thing too,
I would say, if anybody is worried that there's a recorder
out there and it's recording people,
it doesn't do that. For those who are
concerned about privacy, it's not like just
leaving a voice memo running and you're picking up all this human chatter. It doesn't record
that at all. The system is smart
enough to only focus on the species of
interest, so not human. So that's That was one of the first
question people had, like, Is it always me running and
listening to me? I was like, No, it
doesn't record you. I'm not going to
hear you talk bad about me in class or something
like that, right? So. [Sage] That's great. Yeah, I mean, that privacy piece, I'm sure, would
be important. And it's great. Hear about the cross disciplinary
interaction you have with early
childhood education, too. [Colin] Yeah, we're
fortunate here at LCC to have a childcare
center on campus. That's for both students and staff and community. And this last term,
I decided to bring in to my course in person version
of this class. So Birds of Oregon,
sometimes I teach it as an online class, and then sometimes I teach it as an in person course. For my in person
version of the class, students did a service learning project where
they had to focus on bird adaptations and then how they were
prepare how to teach that to our two, three and
four year olds. We have a child
learning center. [Sage] Well, let's see another
question for you, kind of a general
question, but were there successes or challenges that
you encountered? I know you talked
a little bit about the power sources
for these. [Colin] Yeah, so to our surprise, maybe just a challenge, would just be
recognizing that or kind of the continued costs beyond the unit,
right? Is batteries. So they are battery run, and they do work on
rechargeable batteries. So that's something
we adopted. But we did find them to be more hungry or use more power than
we've anticipated. So something to be
thoughtful about, and maybe that's just a software
issue as well. One thing about this company bird
weather is nice, they do encourage
feedback, so we've been
able to do that. Using batteries as an important
piece about that. The one way around that, like I say, for a
long term dataset is to have an external
power source, right? So you can plug
it into a wall, and they do have
that tool or another power core
that you can use. So I think that's where we're hoping
to move for when we have the one on campus. That's just one challenge. The software itself uses a smartphone to visualize the data, to set it up. And that was
another one thing to be thoughtful
about is at least currently the current
version of the software only allowed one person
one user name for it. And so that's something to think be
thoughtful about if could have a certain
circumstance if you have six students using
six different units. They're going
to be the ones responsible for the data. And that always makes
me a little bit uncomfortable because what happens
if that student, drops a class or something
like that, right? They're taking that
data with them. So that's just where I think
I need to develop maybe a generic LCC user name student
project name or something like that, would be the way to
problem that so that, you know, I would always
have access to it. So things like that are just kind of
like needing to work through something
that was clearly designed for a
backyard enthusiast. And now I'm
trying to scale it multiple units
in a classroom. And they seem to
understand that. So I say the bird
weather developers, They've been really
generous in either answering questions and
asking for feedback. And in fact, I know
they sent out a 100 of these units to
K 12 classrooms across the country
this last year. So they clearly understand
they're interested in education and the power of these
for education. So that's something
to think about. I think, just the
back end of it. In the moment, of
course, when you're using it and it'll
give you a list, a running list
of the birds that is identifying. But the value of it is the dataset over time
and just getting the data out of
the unit out of the unit in a CSV file, for example, that's just one thing that
some students, that's a little
more challenging. And so I know I need
to be able to offer more support for
spreadsheet tools and working
with that data, and how to extract
it, and how to do some data cleaning so that
we can look at those trends
across time. So That's a really
great opportunity. I always like to have
quantitative piece in all my biology classes. So I don't think that's a challenge
or a problem or a limitation is just
something that I need to think through
and how to support students and how. And I wasn't yet
very familiar with how the data was
going to come to me, so to speak, in terms of when
you download that. So I just need
to build a lab. Is more structure around that to help students, more scaffolding,
so they're more successful with
using data. That's coming down
of these units. [Sage] That's great. Yeah, I truly sounds like
an exploration, and you're learning as you go with this project. But it's really
exciting to hear. I'm just listening to you. I'm already
getting the sense in terms of how you might answer this next question. But do you feel
like you achieved your desired project
goals or outcomes with I think
we started to. [Colin] I think we're just on
the cusp of it, right? And that's where
these tools are a permanent part of biology stock room. And that's the important
piece about that, which is And when I say we achieved
our goals, yes, we got student
engagement, we collected some data. We learned what we need to learn in terms of to make it more effective and how I think to scale
it across. So meaning how to help
lower the threshold, for example, students who might be needing
more support and how do you
spreadsheets, right? That's something I'm
going to have to give more time
and attention to. Developing a universal
login for the system, so I can have
multiple units where we would always have access to the data, right? Things like that. I can work towards
an excited way. Yeah, I think
met our goals. I think it's going
to be now like moving them into more
uses and more classes. So at LCC, we
really try to stress research opportunities in
all our classes. So with that might look like in
different places, different scales. For example, if
you were doing the science life
science courses, the biology principles
of biology, that's a 200 level
class intended for students who are transferring a
four year school, they get to design
and develop their own
research project. And this is now
going to be a powerful tool that
lets them do that. It compliments
other research projects we've
done in the class. So for example, we have a long term
running project on Bird Window collisions, LCC, which is sad. But in a factual
statement, actually, Bird Window
Collisions is one of the second largest
contributor to bird species decline. And so we've had a
long term project now of students doing
the monitoring project. So you can imagine
creating dataset, here's the birds
we're hearing and sensing in the area. Here's the ones
are finding with our bird window
collision project. What's that
relationship? Are there certain species which
are more prone to that? Are there certain
ones that we only hear
and never see, we never see evidence that they've collided
with the window? Right? So that would be an example of a
research question or an extension
that we can do, and we could do
only because we're going to have access
to these tools. I'm going to
use them in my biology forest ecology class
this coming fall. Students do a variety of things in that course, including learning how
to identify trees. But we always ask
a question how are species depended
upon a forest. I'm going to imagine, for example, Using
these tools, creating the
data set where students will
have a chance to see here's an
older forest, here's a younger forest. Is there a difference in the species that we
detect between that? That's a really
nice contrast. Even more basic. I'm still learning,
for example, how sensitive
are these tools, meaning, how close does a bird have to be
to be detected. One of my first activities I think with this
class will be because it'll be
an in person class is Let's put one of
the units and install it on a tree and get a
tape measure and play back songs at
known distances and see where
it's detecting. Let's get an idea
what we call a sensitivity analysis. How sensitive
are these tools. Because that tells
you a lot about how far you can
effectively sample. There's a lot of
questions that are really scalable
and accessible, even to a 100
level student who is taking a course
for general education, and that they can
really contribute to other students
using it as well. That's what we
want to have this idea that
we're building data set that they
are more than just a one off lab. I don't like to
do one off labs. I like to do
things which are consequential and
develop and push students to develop
their skills and that they're
creating something and creating a user
for something. That's what I'm hoping to keep moving
these tools for. I feel like we're just
on the edge or just on the cusp of being all using this
next full year. So That's in
the classroom. I will say,
anyone there who is considering applying
for this grant, whether it's for these
tools or another tool, one place I did
stumble and say, is this one of your follow up questions or his? [Sage] Oh, no, feel free
to contribute. [Colin] Okay. Just to recognize
if you are thinking of purchasing any
new technology that you want to
use on your campus, I encourage you to engage with your IT department. Because that was
actually slowed us down in terms
of being able to deploy these tools because our IT department
got a little uncomfortable about something that's
connecting to our Wi Fi network
potentially for long term, right? And they have
security concerns, right? In terms of that. So That's something that as a biology instructor, I was not even
on my radar, not even a
question to ask. So any new tool I think that you
want to bring into the classroom
that's going to take advantage or use
your network, which is probably a
lot for this group. Due to due diligence
of disengaging with your IT department
and understanding the any security concerns they may have, right? So that was we ultimately were able
to resolve that. We put them on
kind of like a secondary network is
how we had to do it. But, they became a little bit uncomfortable
with something that in their view is not as vetted
and may not have the same security features that you're going
to have from other from a
Microsoft product or your cell phone, right which is
regularly updated. So they don't like things that might join
the network. This is my understanding that might join
the network, but then not
regularly be updated like no virus
protections or no security protections
regularly, right? And you can imagine a tool designed for bird
enthusiasts isn't going to have the same
security rigor that something else that, you know, like your cell phone
might have, right? So just a point to really do yourself
that due diligence, support your
campus IT team, and just ask them what their needs are
going to be. No. That makes a lot
of sense. And it certainly sounded like
it was a challenge that you encountered
that was not anticipated as part
of the project. Yeah. The challenge
being like, Oh, what do you
got there, right? Like, they want to support you in
a classroom, but they legitimately have that's their
job, right? That's there to help us keep our campus
community safe. None of us want to deal
with a campus that doesn't have a functional
network or one of these sort of ransomware
circumstances, right? So that's where
they're coming from, and we have a newer
director of IT, and that was their job is to help keep the bad entities
out of our network, and that was their
concern, of course. [Sage] Yeah, that's a
great point. You know, It's
excellent to hear these research
opportunities that your students
have and some of the emerging research
opportunities you out. Just a question about
distance learning since we're the
distance learning Association for Oregon
Community College. Now you teach
online classes, do you see opportunities
or were there any unique challenges
in trying to or how you might approach
incorporating some of this into
your online teaching? [Colin] Yeah, I think originally, I had had the
expectation that my online students
would want to potentially host one
of these units in their home and use it to collect data in their home or
their environment. And that just became
a little a challenge in terms of handoff
and getting them back. So students, I ended up giving them out to faculty members. Is what I ended up doing to get collect
data sets in different places
during the first trial because I just found that most students
weren't did not want to come to campus
to pick up the unit and then
that challenge, but how do we get
it back, right? So that part is just something we're going to think about whether
or not in the future, if we're willing
to You know, students just
didn't want to take responsibility
for it, right? Which I
can understand. They're coming from
different places. So that was just
one thought. That was like my
original point, wouldn't it be
awesome all across? You know, my first
thought was maybe these would be some
student has it for two weeks at
their home and then they return and goes out to
another student. And that may still
work in the future. But the way originally, that didn't work
this last time. Let me just say
it that way. So that might just be me, still say understand where students are coming from, maybe identify
students who are an online class, but they're
also on campus, and so they're going
to be interacting with us into
physical space. That might be another
way to do it. But when I offered, nobody took me up on my offer this last
term in terms of hosting one of
their own units or one of these units
in their own homes. But in terms of helping create a dataset
that you can now develop lab activities
for students to use, whos?
Absolutely, right? That's still going
to be possible, even if even
if these units only stay local on
the LCC campus. So that's an important
thing to think about I'd like
to like you say, your question
for online in the distance learning
is the cool part about this is that you're
part of a network. So I want to develop more comprehensive
lab activity. Here's one of these bird Weather pucks records
on this location, find another place of interest to you any place, for example, Maybe maybe it's where your
grandparents live. Maybe it's where
you used to live. Maybe it's across
the country. Maybe it's a state
park that you know and recognize where
someone else is develop has one
of these units, and let's access those
datasets together. This is what we call a regional level
biodiversity or Species list, right? And see what
species are common, and which ones are
unique to those areas. So that's where I want
to see when I saw the spatial or the network of these recorders
across the country, like, Oh, oh, that's
a really cool thing. And I think I just need
to spend more time trying to develop how
to access that data. Besides antidotally, when you click
on it one time, and it kind of it's in
real time telling you where these birds are being detected at
the same time. And so that I think
is really cool, but I just need
to kind of help develop that tool.
Like we're using it. It's not just the
data that we have. The real power these I'm seeing is
the data that other points across the country are
developing. [Sage] That's great. And so now, tell me what did your students
think about it? Did they give you any feedback
after your trial, working with students
with the data? [Colin] Yeah, I got some feedback. Some of which
was like, Oh, I don't think most
students don't approach spreadsheets with the same joy and passion that I do. And you got to remember
that about them. But to them, it's the bird that they
want to see in the field and they are here. So they like the idea. They liked seeing it,
but the data processing or the data
interpretation of it was this little
piece about that. So I had a lab activity, which I developed that
did that does help. One thing I
learned to do is I realize in developing
an activity for them, how to interpret the data, how to clean and
prepare the data. How to find essentially the signal versus the
noise in the data, that it was a lot to
try to do remotely. And so I actually flipped it around
a little bit, and I said, Here's
one way to do it. We walk through it. But
then I also said in this modern day
with generative AI, how could AI also help
us? Identify data? There was an
activity whereas, I walked through the
long way to do it, the old school
way of doing it to identify the number of unique species
that we're detecting. And that's an important
thing for biologists. Sometimes we say,
well, we only hear species A here. We don't see it in
any other places. So, maybe this
place is really important if only
specie A is found here. That's a regional level of diversity that we
care a lot about. So I flipped it around and I
said, Okay, well, here's how we did it along with the
old school way. And then I said, Okay, well, let's see
how we could also. And I usually, for my
distance learning, I always make a video
for every activity. I said here's actually
how we can use AI to help us find thee same patterns and trends. So I walked them through a little bit how to take
that data that we've pulled out of
the spreadsheets and now see how AI can help us actually identify those
same trends as well, and actually could do
it in a faster way. So that's like a
learning curve for me, which is still You know, we're getting an era where data shortage is no
longer the challenge. It's great to be able
to create new data, and we do need to, especially when we
think about over time, how species are changing. But there's a lot
of large slum. There are datasets there
that we don't fully utilize because it's
time consuming to do so. So to me, that was another
little aha moment, like, Okay, I Yeah, I need to make
that a part of my class as much as
helping my students learn how to
use AI to help them for not just
data management, but in their own careers. I only became possible or something that was
evident to me when I saw, because these reporters
had the potential to generate a lot more
data than individually, I could do in a class. It became like,
a new threshold in terms of the amount of data now requires a different response
than if I only had 20 rows of data. It was like, Okay, that's a really
cool thing, but helping students learn to work through that. I think a lot of us
are still there, still learning how to
help students learn to use generative AI in a positive part of
their learning process. So that to me was
like a new thing. [Sage] I think that's
terrific. It seems like such a great
opportunity to incorporate AI into their learning
and help them to build some AI
literacy skills, and that's very relevant and meaningful to
the curriculum, they're learning. [Colin] Also, and again, because
these tools, like, again, if you went out, and I've done
this part before, prior to these type
of acoustic tools, which is how many
birds you saw, six birds in
your backyard, because students have
to do an hour of birding every
week in my class. So they're
creating curating their own datasets
for birds that they've
seen identified. And that's a
smaller. That's much more of a manageable data set that you could
work through. Now we're like,
say 20 rows, 50 rows, 100 rows
of observations. But now you have
the potential to have thousands
of rows of data. Okay, gosh, that's a
different challenge for a student learn and overwhelming,
otherwise, right? Yes. So that's how AI AI is helping
us identify this. AI can help us actually learn how to process
that type of data. So that was like
new for me. [Sage] This has been
excellent. I really appreciate you being here. Thank you for your
time and sharing some of your experiences
about your project, and very excited to see where you're
headed next. For more information
about the Oregon Community College Distance Learning
Association and opportunities
for funding to explore emergent
technology, please visit our
website at OCCDLA.net. And thank you, Colin, and thanks for listening. [Colin] Awesome. Thank
you so much Sage. Thanks for the support.