First up on today’s podcast, Jeff talks with Maya Ajmera, CEO of the Society for Science who along with the pharmaceutical company Regeneron, host an annual science contest for high school seniors. Maya talks with Jeff about the contest and...
First up on today’s podcast, Jeff talks with Maya Ajmera, CEO of the Society for Science who along with the pharmaceutical company Regeneron, host an annual science contest for high school seniors. Maya talks with Jeff about the contest and what it takes and means to be a finalist!
Next Kim and Jeff talk with Amara Orth. Amara lives on her family farm in Glenwood Iowa, near Council Bluffs. Her parents have 25 colonies and her grandfather has a couple thousand colonies. She’s been around bees her whole life. Amara graduated from high school this spring and will be attending Stanford college this fall. A couple of years ago, when working bees with her grandfather, he said something that got her attention….the bees sound different today, and that got her to thinking about the sounds of bees.
Amara set up a series of listening devices on her 25 hives to eavesdrop on her bees so she could identify the vibroacoustic patterns she detected in each hive. She measured the sounds and vibrations in each hive and developed a mathematical model that analyzed colony health with an accuracy of 92%. She entered her research in Regeneron Science Talent Search Awards program and placed 9th out of several thousand entries, earning a $50,000 award.
Listen in today and find out more about the sounds of colony health.
We hope you enjoy the episode. Leave comments and questions in the Comments Section of the episode's website.
Thank you for listening!
Links and websites mentioned in this podcast:
We welcome Betterbee as sponsor of today's episode. Betterbee’s mission is to support every beekeeper with excellent customer service, continued education and quality equipment. From their colorful and informative catalog to their support of beekeeper educational activities, including this podcast series, Betterbee truly is Beekeepers Serving Beekeepers. See for yourself at www.betterbee.com
This episode is brought to you by Global Patties! Global Patties is a family business that manufactures protein supplement patties for honey bees. Feeding your hives protein supplement patties will help ensure that they produce strong and health colonies by increasing brood production and overall honey flow. Global offers a variety of standard patties, as well as custom patties to meet your specific needs. Visit them today at http://globalpatties.com and let them know you appreciate them sponsoring this episode!
Thanks for Northern Bee Books for their sponsorship of Bee Books: Old & New with Kim Flottum. Northern Bee Books is the publisher of bee books available worldwide from their website or from Amazon and bookstores everywhere. They are also the publishers of The Beekeepers Quarterly and Natural Bee Husbandry. Check them out today!
Thanks to Strong Microbials for their support of Beekeeping Today Podcast. Find out more about heir line of probiotics in our Season 3, Episode 12 episode and from their website: https://www.strongmicrobials.com
We want to also thank 2 Million Blossoms as a sponsor of the podcast. 2 Million Blossoms is a regular podcast featuring interviews with leading bee and insect researchers in the world of pollination, hosted by Dr. Kirsten Traynor.
We hope you enjoy this podcast and welcome your questions and comments in the show notes of this episode or: firstname.lastname@example.org
Thanks to Bee Culture, the Magazine of American Beekeeping, for their support of The Beekeeping Today Podcast. Available in print and digital at www.beeculture.com
Thank you for listening!
Podcast music: Be Strong by Young Presidents; Epilogue by Musicalman; Walking in Paris by Studio Le Bus; Original guitar background instrumental by Jeff Ott
Beekeeping Today Podcast is an audio production of Growing Planet Media, LLC
Jeff Ott: Welcome to Beekeeping Today Podcast presented by Bee Culture. Beekeeping Today Podcast is your source for beekeeping news, information, and entertainment. I'm Jeff Ott.
Kim Flottum: I'm Kim Flottum.
Global Patties: Hey, Jeff, and Kim. Today's sponsor is Global Patties. They're a family-operated business that manufactures protein supplement patties for honey bees. It's a good time to think about honey bee nutrition. Feeding your hives protein supplement patties will ensure that they produce strong and healthy colonies, by increasing brood production and overall honey flow.
Now is a great time to consider what type of patty is right for your area and your honey bees. Global offers a variety of standard patties, as well as custom patties to meet your needs. No matter where you are, Global is ready to serve you out of their manufacturing plants in Airdrie, Alberta, and in Butte, Montana, or from distribution depots across the continent. Visit them today at www.globalpatties.com.
Jeff: Thanks, Sherry, and thank you Global Patties. Each week we get to talk about how much we appreciate our sponsor's support, and we know you'd rather we get right to talk about beekeeping. However, our great sponsors are critical to help making all of this happen. From the transcripts, the hosting fees, the software, the hardware, the microphones, the subscriptions, the recorders, they enable each episode. With that, thanks to Bee Culture magazine for continuing their presenting sponsorship of this podcast. Bee Culture's been the magazine for American beekeeping since 1873. Subscribe to Bee Culture today.
Hey, everybody, thanks for joining us. We're really happy you're here. Before we get started, just a quick reminder to subscribe or follow Beekeeping Today Podcast and give us a five-star rating. It really does help. Also, we are now adding complete transcripts of each episode on the website after the show notes. Check them out. You can also leave questions and comments online under each show. You can leave a comment, ask a question, reply to a question, ours or a listener's.
Click on "leave a comment" at the top of the episodes' show notes to join the discussion. Have you listened to an episode and thought, "That person sounds really interesting, and I'd like to know more about them"? Well, now you can. Each episode links to a guest profile. Each profile has a guest photo, bio, contact information, including Instagram and Twitter details if they have them. Check it out. Finally, share the podcast with your beekeeping friends, email them links, or mention it at your next beekeeper meeting.
Hey, everybody, thanks again for listening. We know you have many different podcasts to choose from, and we're really honored you've chosen to spend the next several minutes with us. I hope your first week of summer is going well. How are your bees handling the heat? We're hearing reports of honey being pulled across the Midwest, and confidentially, Kim reported that his colonies have three mediums of honey each. Wow. How about water?
Do you know where your bees are getting water? I was asked by Dr. Jim Tew to join him on The Honeybee Obscura Podcast last week while Kim was away to talk about honeybees and water. It is a fun discussion that I hope you listen to. In fact, Honeybee Obscura has 79 episodes available to download. I'm sure you'll find many different topics that stimulate your beekeeping mind. We have a great show lined up for you today with two different guests.
First up, we have Maya Ajmera for the CEO for the Society of Science. Each year, they sponsor a contest for high school seniors from around the United States, including the US territories to submit science projects for review. Now, this isn't any old just high school's contest. The first place senior receives $250,000. Wow. Now, this in itself is pretty cool, but why is she a guest? Because our main guest today is at Amara Orth, the 18-year-old, newly graduated high school senior from Council Bluffs, Iowa.
Her science project involved recording the sounds of the hive and determining its current health through special Artificial Intelligence code she wrote specifically for that purpose. Wow, that's impressive. Amara comes from a family of commercial beekeepers and it is her grandfather who inspired her when they were out inspecting bees. When he said, "Hey, Amara, those bees don't sound quite right. I wonder what's going on." No, he didn't sound like that. I'm sure he is very cool, had a very good voice. Hear how Amara placed in this very prestigious science contest featuring her research using honeybees. Let's get right to our conversation with Maya and the Society for Science.
Jeff: Welcome, Maya, to Beekeeping Today Podcast. In a few moments, Kim and I will be talking with one of the finalists of this year's Regeneron Science Talent Search, Amara Orth, about her competition project that she submitted. Can you tell us a little bit about the Society for Science and about the Regeneron Science Talent Search?
Maya Ajmera: Sure. It's a pleasure to be with you on this really fun, interesting podcast. I am the president and CEO of the Society for Science and publisher of Science News. The Society for Science was founded 100 years ago. It was founded by EW Scripps and William Riter. The reason they founded it is because they were very worried about what they were reading in newspapers, about three-legged martians.
They wanted to create and build evidence-based science journalism, so they founded Science News.That's still going 100 years later, but in 1942, they really were very worried about talent, really, for young people to go out and discover and research on all kinds of extraordinary discoveries out there. This talent would then our generation of scientists and engineers and entrepreneurs in our country.
It was founded as science talent search. For the first 50 years, it was called Westinghouse Science Talent Search. 25 years later, it was Intel, and then after that, Regeneron. The biotech company became our sponsor seven years ago. The Regeneron Science Talent Search really searches for the best scientific talent and leadership in the country.
Jeff: At the high school level.
Maya: Yes, they are seniors only, but they may have been doing the research over high school or during the summers and they go through a very intensive process of applications turning in a research paper and we get about 1800 applications.
Maya: We have judges from around the country and they whittle it to the top 300 scholars and those top 300 are whittled to 40 finalists. Amara was one of those finalists this year. She also competed in the Regeneron International Science and Engineering Fair, which is the largest pre-collegiate STEM competition in the world as well. Amara competed in both, and she is an extraordinary young person.
Jeff: You mentioned STEM and that's science, technology, engineering, and mathematics.
Maya: That's correct?
Jeff: Oh, that's a great program. How does a high school student junior decide and enter the competition on a local level?
Maya: No, it's an open funnel. Any kid can apply. They just need to know about it. We work with STEM teachers throughout the country. We have a suite of outreach and equity programs. We want to make sure that every young person can become a scientist or engineer if that's what they want to do. We've created a suite of these programs from a mentorship program called The Advocate Program to high school research teachers conferences to STEM action and research grants. That in itself then goes at the grassroots and really gives young people the information that they need to be able to compete in this competition.
Jeff: The selection process is highly competitive, as you said.
Maya: That's correct.
Jeff: How does that work from a local level, from a local high school to perhaps a regional down to the- you said the 300.
Maya: It's all gone to the national level. Anybody who applies gets to the national level. Those 1800 applications are looked at with judges from across the country. Those judges then look at those 1800 and whittle it down to the top 300. Then we bring an acclaimed group of judges to Washington DC and rock them up for four days and then they go through the top 300 and go through and select the top 40 finalists, who come to Washington DC and spend a week.
Then they go through a very intensive interview process. What's interesting about that is that we're very interested in scientific leadership. If Amara, for example, her project is about bees, they may ask her a lot of physics questions or mathematics questions. They're really looking for that well-rounded scientific leader, not only that has great research, but is also really understanding and fascinated by science and everything that is science and engineering.
Jeff: Where can students find out more information about entering into this competition?
Maya: Sure. You just need to go to societyforscience.org and just scroll down and you'll see Regeneron Science Talent Search. You press that button, and you'll find out everything you need to know. The application deadline is in early November.
Kim: Does the project need to be completed by early November or well on the way?
Maya: Well, there's no project that's ever completed, right?
Maya: There's no research project ever completed, but there is a place where I feel I have good data to present or not-so-good data to present, but I've gone through the research exploration of a question I'm trying to answer, and I have something to say about it. That's important. I think the student makes that decision with their mentor or teacher as to when that stomping point is, when is a good time to let the world know?
Jeff: I applaud what you're doing. I applaud the projects you do and the opportunities you're given to young people to explore, learn how to think and be critical thinkers and present their findings in a logical and straightforward fashion. This is a great program.
Maya: I would also like to say that Amaras project is not the first project in researching bees. This is a very popular topic with our students. Just a couple of years ago we had a young woman named Raina Jain who did a project on bees. She was a 2020 science talent search finalist the Regeneron science talent search finalist. She is extraordinary, and her work was actually featured in Bee Culture magazine. She's just fabulous. She also founded Queen Bee, wellness supplements infused with honeybee byproducts. A pollinator tree is planted for every bottle sold.
Jeff: Oh, wow.
Maya: She's recently been launched at Whole Foods Market.
Maya: It's pretty extraordinary, right?
Jeff: Yes, it is. Bee Culture is a key sponsor of the podcast, so we'll have to look that up. Thank you for joining us on today's episode, we look forward to talking to Amara, and in the future, be looking forward to talking to you about other science projects that come to you about bees, beekeeping, pollination. We cover the gambit here at Beekeeping Today Podcast, and, again, I applaud what you're doing, especially, I don't know. Sometimes it seems science is under fire these days. This is really good work. Thank you so much for doing it
Maya: Absolutely. It's a pleasure to be here. Thanks so much for having me.
SuperMicrobials: Hey, beekeepers, many times during the year, honeybees encounter a scarcity of floral sources. As good beekeepers. We feed our bees artificial diets of protein and carbohydrates to keep them going during those stressful times. What is missing though, are key components, the good microbes necessary for a bee to digest the food and convert it into metabolic energy.
Only SuperDFM-HoneyBee by Strong Microbials can provide the necessary microbes to optimally convert the artificial diet into energy necessary for improving longevity, reproduction immunity, and much more SuperDFM-HoneyBee is an all-natural probiotic supplement for your honeybees. Find email@example.com or at Fine Bee supply stores everywhere.
Jeff: While you're at the Strong Microbials site, make sure you click on and subscribe to The Hive, their regular newsletter full of interesting facts and product updates. Welcome back to the show. Sitting across the virtual Zoom Beekeeping Today Podcast table is an Amara Orth, and she is from Council Bluffs, Iowa. She is a freshly graduated senior from high school, and her science project will blow your mind. Amara, welcome to beekeeping today, podcast.
Amara Orth: Hi guys. Thank you for letting me on your show today.
Kim: It's a pleasure having-- It's nice to meet you, Amara.
Jeff: Yes. Amara, your project is an award-winning project. It's not just any old project. It's an award-winning project. We will have links, just heads up to folks. You will want to check out the video links to Amari's project, Amara. I know that you come from a family of beekeepers. Why don't you just introduce yourself to the listeners, and let them know about your background and your family?
Amara: Yes, sure. My name's Amara Orth. I'm from Council Bluffs Iowa. I grew up on a honeybee farm. My family raises about 50 colonies, and my grandparents are commercial beekeepers with about 2000 colonies. Wow. For my project, should I go through that or should I wait for that?
Jeff: Sure. Yes, tell us what, what was the purpose of the project? Yes, what's a little bit about the background on the project.
Amara: My project was inspired by this past, I guess, almost two years ago, I was talking to my grandpa. I was trying to come up with a project and then my grandpa said something. He was, "The bees sound a little different today." I thought to myself, "I wonder if there is truly a way I can go into the hive and see how the sounds that the bees make can be indicators of their health." I started my project in early 2020 during the COVID lockdown. All the universities near me were closed.
I had to try to come up with a project I could do at home that wouldn't take very many materials. It took almost two years of trial and error to get the right microphone and the right setup. I didn't have very much coding experience. During my lockdown, I had to teach myself how to code and just to get the right setup and everything. I finally was able this year to get the right setup and the right microphone to be able to get data I could use.
Jeff: Wow. This is just the intro, but I'm amazed. You spent your COVID lockdown years working on this instead of perfecting your TikTok videos. This is something amazing. I like this.
Amara: Thank you.
Kim: Yes. You said your parents have about 25 colonies, and it sounds like you spend some good deal part of your time working with them, right?
Kim: How long have you been working bees, your whole life?
Amara: My family started raising our own colonies back in about 2015, but I've been involved with bees ever since I was really little. I can remember when I was probably about five years old, we had visit my grandparents' house. I just remember being fascinated with the bees and being like, "I wonder if someday I could truly be in the colonies and tinker with them to see if I could make research projects to help them out."
Kim: That's pretty cool. One of the things I'm wondering, back up half a step. The contest that you entered is run by-- Who has sponsored this contest?
Amara: It is sponsored by the Regeneron, but also the Society for Science. It's called the Regeneron Science Talent Search.
Kim: How did you find out about it?
Amara: My mom was actually a finalist in the Regeneron Science Talent Search back in, I think it was 1999 and she placed, I think, she got fourth place. She was like, "I think you should try out for this science competition. You can get a lot of scholarship money with it. It was just a truly fascinating, incredible experience for her. She wanted me to see if I could get in there too. It was a lot of fun, it was really cool to meet other really talented scientists like myself that are working on these incredible projects.
Jeff: Oh, that is so cool.
Kim: One of the things on the webpage, Jeff, if you notice it, were the, I want to say the top 10 winners and what the projects they did. You were ninth on that list, if I recall.
Amara: Yes, I was.
Kim: Do you have any idea how many people entered that contest?
Amara: I think it was somewhere around 2000 people nationwide. Wow. Then they made it to the top third or 300, and then out of the top 300, they did the top 40 that got to go to Washington DC and compete for the top award of $250,000.
Kim: What was your reward for placing ninth?
Amara: I got $50,000.
Kim: That's going to go a long way paying for college, that's for sure.
Jeff: We're jumping to the end, but where do you want to go to school?
Amara: I'm actually attending Stanford this fall. I'm going to major in earth systems. It's environmental science.
Jeff: Congratulations, Stanford. Thank you. Yes, that's fantastic. I'm flabbergasted. I'm half torn between wanting to hear more and just feeling sorry for myself for my wasted high school years. [chuckle] I'm just teasing. Tell us about your project. You sit there and you say, "Boy, these bees are making different sounds. Grandpa's even heard it." What was your first step? What did you do first?
Amara: First was trying to figure out what microphone to use. I first started with, it's called a shocked microphone. It's a cylindrical microphone. I originally tried using that microphone, but I found it picked up too much background noise like our chickens and airplanes and the wind, everything would just be part of it. I was, "I got to find the right microphone." After trying probably dozens of different microphones, I finally landed on, it's called a piezo microphone or a shotgun microphone.
It's a flat dish-shaped microphone that I could just put in the top of the hive. The lid could go on. I wouldn't have to open it up in the winter to expose cold air to the bees. After that, then my next step after that was trying to figure out how long I should record. I originally tried thinking of doing 24-hour surveillance, but that took a lot of storage space. It's, "Ooh, maybe not." I went to 15 minutes, and I still find that was too long. I ended up doing a five-minute recording about three times a day. I do, like, in the morning at 7:00 AM then around lunchtime or so, and then I'd do it in the evening.
After I had those recordings, then I had to figure out how I was going to process the data. Originally, I tried looking at spectrograms, which are visual outputs of the audio file, but it's just like colorful lines, essentially. I was like, I can't exactly isolate individual variables because, for example, wingbeat harmonics. I tried looking at that, but I found that isolating the individual variables made it so the overall processing was really difficult, and the data wasn't really useful.
I was like, "I think I have to use machine learning." I was like, "Oh, my goodness, how am I going to do this? I don't have any coding experience." My project is one of the first ones that uses this machine learning early warning detection for beads. I took a lot of research on different articles that did it for I believe it was, like, whales and birds and other animals. I eventually found the right machine learning system to use. It's called Hidden Markov models.
Jeff: Wait, let's say that again. Hidden H-I-D-D-E-N.
Amara: Hidden Markov. It's like M-A-R-K-O-V.
Jeff: Okay. Hidden Markov.
Amara: I found that the Hidden Markov models are also abbreviated to HMMs. I found that they had fewer false-positive rates compared to other machine learning models and was overall easier to use and fit the sound of the bees a bit better. It was less like-- A lot of the other models were very linear, and it didn't really match the sounds of the bees. After I decided on the Hidden Markov model, I had to train the model. For that to collect my data, I first started with a Colony Hill state assessment. For this, I would go into the hive, and I would determine one of eventually, nine Hill states that I tested for. It'd be like queen right, queen right broodless, queen right honey flow, virgin queen, volatile chemical exposure.
For that to be, you're given a dawn toxic mite treatment called formic acid. I did queenless, dwindling hive, robbing dwindling hive, and then robbing no hive. I would assess for one of those nine health states, and then I would go into the sound collection. I would record the audio for five minutes. The microphone was connected to a sound acquisition thing called a Scarlet 18i20. It was just like a sound interface. It was connected to a computer running the application Audacity. Then in Audacity, I would label the audio file with the identified health state.
Then I would train the Hidden Markov model using, it's called a Train HTK command, it's like one of the lines of codes, to produce nine different directories or health states. That's a lot of steps. I would call upon those directories to help recognize unknown audio files using another command called Recognize HDK to produce an annotated label file. For this, it would be like a spreadsheet, and I would look for the highest number of occurrences for the different health states. The audio file had the highest number of occurrences of these different health states, that would be to identify health state.
Jeff: Why did this only take you two years? I'm just kidding. That's amazing. What you're doing is phenomenal. I want to go back, though, because I just get tripped over some of the technology. The microphone. What frequency range were you listening for? Were you just looking at for the 20,000 MHz, or did you try to go outside the human hearing?
Amara: I believe the shotgun microphone or the piezo microphone rather had a range of about 100 to 16 KHz. I found that range picked up the bee noises, but not the super high or low-frequency background noises unlike the shotgun microphone.
Jeff: Okay, Kim, did you catch that?
Kim: I am just listening, Jeff, because this is over my head. My listening to bees goes as far as taking off the cover. I'm enjoying listening to this. I don't understand a lot of it just because I don't have training in this, but the story you're telling is amazing. The things that you're looking for and the things that you were measuring. If I recall, the data you were collecting allowed you to be somewhere over 90% accurate in terms of predicting what was going on in the hive.
Amara: Yes. The training was 92% accurate with the training files, and it was 100% accurate? Oh, no, it's 100% accurate with the training files and 92% accurate with the unknown files.
Kim: I wish I could be 92% accurate in anything, let alone listening what's going on in a beehive.
Betterbee: Get your high-quality Lyson beekeeping equipment at Better Bee. As the United States distributor for Lyson, Better Bee is proud to offer honey extractors that range in size from holding two to 88 frames, insulated polystyrene hives and nuke boxes, wax and honey processing equipment, and over 250 styles of licensed, highly rated, and easy to use silicone candle molds. Visit betterbee.com to learn more and shop for high-quality licensed products. Visit Better Bee today, your partners in better beekeeping.
Kim: You were working with your parents' 25 hives, and it sounds like from what you said, you set up a colony to bee queen right, or queen right or being robbed. You made the colonies fit and then measured the sound so that you could say, "Okay, this sound says this colony is queen right. Is that how this works?
Amara: Sort of. It was rather like my mom. She grew up on a honeybee farm, so she knew all these different health states by heart. We didn't necessarily alter the hive to have these health states. We would just go into the hive and identify what we thought was the best health state. Sometimes the model would correct us. It would be like, Oh, it's actually not queen right. It's actually queenless. We found we were like, "Why does it say it's queenless? Then the colony died a few weeks later. It sometimes corrected us, but we'd look into the hive and we'd go through and see what the hell state was. Some of them were easier to train than others.
I originally started with just two health states, so doing queen right and queenless. As I had more data, I was able to correct it and get the right health state and branch out to eventually get nine different health states.
Kim: Yes. The things you were looking at were listed on the release that they put out. Like I said, I'm impressed. It's good you had a beekeeper working with you. You had two minds working on identifying what the colonies were doing. It wasn't maybe it was pretty much pretty sure that this is a queen-right colony getting ready to swarm.
Kim: Yes. That's good. You would just work with the 25 hives that your parents have. That was your whole data set.
Amara: Yes, that's all I was able to get done, but I hope to do more hives in the future. My mom and I are actually working. We're leaving for New Zealand in about 24 days. My mom, she does honeybee microbiology research at Creighton University in Omaha, Nebraska. She got the opportunity to continue her research in New Zealand to look at varroa mites. She was telling them about my research, and they're actually really interested in it. I might have the opportunity to do some of my sound collection on the hives in New Zealand. We're trying to get some control group of my grandparents' hives as well.
Kim: Well, we have quite a few listeners in New Zealand. You will down there. They'll already know a lot about what you've done.
Jeff: We've also learned along those lines from our guests that there's different dialect differences. You'll have to retrain your AI to recognize the different dialects. For the New Zealand bees.
Jeff: That's fun. That is so fun.
Kim: Amara, you collected all of these sounds, and you begin to be able to recognize them. How can that help me?
Amara: Yes, so how it works is that it tells you what the health state is. Sometimes as a beekeeper, you look at your hive and you're like, "I wonder what the health state is. I see that the bees are declining and there might not be a clear way for you to know what's wrong with your hive. You might have some indications like, "Oh, I see a group of dead bees outside the hive. Maybe there was pesticide exposure or something, but sometimes it is really difficult to know what the health is.
Especially in the wintertime, when it's cold, it's really difficult to open up the hives because then you're letting out all the warmth inside the hive. My machine learning model can tell beekeepers what the health state of their hives are. Sometimes before it's even visible to them. It allows beekeepers to know what's wrong with their hive and allows them to take mitigation strategies to protect their hives before their hives collapse.
Kim: You got a microphone in a bee hive, does it live there all the time, or do you move them around, or how does that work?
Amara: It depends on the setup that if you as a beekeeper would be using, you can keep it inside the hives. That's what I had done. I had quite a few different microphones just hanging out of all my 25 hives, so you'd walk and really, "I wonder what's over there," but yes. I had all the microphones inside the hive and I kept them there and they had a pretty long range that they were good for so they probably were active for about 4 months. Then we had to replace them because the health issue of the microphone deteriorated, but overall, they're able to withstand all the weather conditions for about 4 months or so.
Kim: The microphone's in the bee hive, you only for a short of time morning, afternoon, and evening?
Kim: That's the data you collected? How does the data get from the microphone to your computer?
Amara: The microphone was connected to a couple of different cables that eventually led up to the computer. A few different cables that ran up to the computer. Then on my computer, I had an app open that would record the audio for 4-5 minutes. Then after I record the 5-minute recording on the computer, I would save it as a certain file with the label that I gave it, that tool that, "Oh, this is a queen right hive. Then I would then go into the machine learning model on my computer and I would do the different training commands.
Kim: Each of these microphones is hardwired?
Kim: Does the technology, I am asking a question I know the answer to that, a microphone could transmit data to your cell phone?
Amara: Yes. I'm trying my next step for my project is to have a public platform for beekeepers around the world to be able to upload their audio and get the results. I'm eventually hoping to get an application on your phone that you could hook up the microphone to. That'd be my ideal next step.
Kim: It'd be wireless then and I could walk into a bee yard and my phone would pick up data from the hives that I pointed it at?
Amara: I don't think I'm quite to the step of being wireless. I would be hooked up, how you have your headphones on your iPhone. It'd be that setup.
Kim: Jeff, have you ever tried anything like this?
Jeff: Yes, actually I have, not at all like Amara's doing, but you I'm sure you've heard. In your preparation and research before you started your project, you've probably read about Dr. Jerry Broman Shank out of the University of Montana and his work was sound and bees.
Jeff: He has the app, and we've had him on the podcast, that does much the same, but he uses a mobile device, Android or iPhone, and just shoves it in the front and records 60 seconds of sound. I don't know the current status of that project, or if you've had any correspondence with him in preparing your project, have you?
Amara: No, I wasn't, I didn't get any I wasn't really able to find very much public information on his stuff.
Jeff: I would think that you would benefit, he would probably benefit from your research [laugh] and it'd be a great collaborative resource to look into at some point, but yes, this is really exciting. Were you listening primarily to the wing vibrations or were there other sounds the bees were making that you were picking up?
Amara: I originally tried just isolating those individual sounds, but I eventually just decided to do all the noises that the bees were making to train the model using all the noise.
Jeff: You might have said this early on, and I apologize if I didn't catch it. Was one of your model's a level of varroa infestation?
Amara: I hope to get that in the future. It's difficult to train for the varroa mite levels. It takes a lot of, you have to the screening thing, the screen to catch the mites at the bottom of the hive. That's my ideal next step is to try to connect the Vibroacoustic data to the varroa mite levels. Because that's something that my grandpa was really interested in because he has a lot of varroa-mite issues with his hives and he really wanted to see if there was a correlation between the acoustic patterns and the varroa mites.
Jeff: Absolutely the might and even small hive beetle for the areas of the country that have to deal with that. Boy, this sounds like an exciting project. Your next step is to New Zealand and an experiment in New Zealand and see where that takes you. Then in the fall, I assume, Stanford.
Jeff: Have they expressed interest in what you've done in the project?
Amara: Yes, I was talking. I did get a few emails talking about someone doing acoustic work at Stanford, but I'm also working with this company out of Wisconsin. They look at the temperature and the weight data of the hive and I'm trying to work with them to get my acoustic data part of their technology as well so that the beekeepers can get not only the temperature and the weight of their hive, but also the acoustical aspect to draw it all in together.
Jeff: Nice. We happen to know that company very well. They're good folks there. This is really exciting. Besides going to New Zealand, what else are your on your plans for the summer? You have anything besides bees in New Zealand?
Amara: No, not really. My family's going to be in the South Island of New Zealand. We're going to they're going to be staying there for a year, I'm really excited to be able to explore and go bungee jumping. I'm quite excited about that.
Jeff: [laughs] The audio sound of that would probably be really fun too just to analyze that and train your app for fear. [chuckles] This is the sound of abject fear. All right. Is there anything we haven't asked you that you'd like to tell us about your project? There's so many aspects of it. I want to make sure we cover what you wanted to cover.
Amara: I think other than just making sure, I've highlighted everything so that people can understand it. Sometimes I get lost and all the technical aspects that I don't quite make it down to everyday person level. I make it too fancy.
Jeff: The one question I wanted to ask you is what was the biggest thing that you learned in this project?
Amara: Biggest thing I think I learned was the unique aspects that you can hear from the bees. One of my favorite parts of doing my project was going out to my family's hives and putting on the headphones and just listening to the sounds of the bees were making. It was almost, it gave me goosebumps hearing how the bees sounded. Especially at night, I could hear these different, unique clicking noises that the bees were making. Just being able to listen to the bees and hearing how they sounded, especially before a storm or just in general at night in the day. It was just a magical fear feeling hearing how they sounded
Jeff: The biggest thing you learned, biggest takeaway?
Amara: Just that I feel the biggest thing I learned was that you can hear the different health states of the bees. I didn't really know that you could actually highlight the unique aspects of the sound bees make to be able to get the health state. I never knew that there was that much depth in the sound that the bees are making so that you could highlight the health aspects. That was truly-- When I finally got it this year and I was able to map out my data on my board.
One of my favorite graphs shows a collected swarm that my family got in early August and it showed the health fully transitioned to like queenless and then queen then robbing no hive. It was just truly, truly fascinating to be able to be like, "wow, I can actually use machine learning to get the health state of the hive." You can draw this much information from the sounds that the bees are making to be able to create a graph to see their health. That was truly a fascinating part for me.
Kim: I'll bet your grandpa can do it without any help.
Kim: Long-time beekeepers. I've known a lot of them over the years and sound and smell of a hive. They're pretty right on about what's going on just because they've been in so many for so long, but you picked it up after two years, so let's. A lot easier and a lot faster.
Jeff: I like the Rosetta stone that you've developed with the audio. That's I look forward to hearing big things about your project and future growth of the work that you've started and you've created here. It's really fun.
Amara: Thank you.
Jeff: Amara, thank you for joining us on Beekeeping Today Podcast learned a lot from you, and I think it'll be fun for you in the coming months, coming years. Good luck at Stanford.
Amara: Thank you.
Amara: Thank you.
Jeff: Now, that was really invigorating. I enjoyed talking with Amara. She is really on the ball.
Kim: Yeah, the next generation, but I'll tell the things that I found comforting was that she was introduced to a project by her grandfather, saying the bees sound different today. People. Who've been keeping bees for lots and lots of years know that something sounds different, know that something smells different when they open a beehive. I like the way it started, it started feet on the ground and then got technical. It started with somebody, with experience with bees. That was nice.
Jeff: Then someone else hearing that and saying, "Yes, what can I do?" That's a great thought, and then having the drive and the ability to pick it up and take it somewhere. That's exciting.
Kim: Take it a long ways and do it well. 50 grand. That's a nice surprise.
Jeff: [laughs] That it is. That is a nice price, and like I told her, it just really puts my senior year to shame.
Jeff: I think I was playing guitar in a street corner somewhere.
Kim: Well, look, you ended up okay.
Jeff: Okay. We'll leave it there.
Well, that about wraps it up for this episode. Before we go, I want to encourage our listeners to rate us five stars on Apple podcast, wherever you download and stream the show, your vote helps other beekeepers find us quicker, even better. Write a review and let other beekeepers looking for a new podcast. Know what you like. You can get there directly from our website by clicking on reviews along the top of any webpage. As always, we thank Bee Culture, the magazine for American beekeeping for their continued support of Beekeeping Today Podcast. We want to thank our regular episode sponsor, Global Patties. Check them out @globalpatties.com.
Thanks to Strong Microbials for their support of this podcast. Check out their probiotic line @strongmicrobials.com. We want to thank Better Bee for their long-time support. Check out all their great beekeeping supplies @betterbee.com. Thanks to Northern Bee Books for their support of Bee books, old new with Kim Flottum. Check out all of their books at northernbbooks.co.uk. Finally, and most importantly, we want to thank you, the Beekeeping Today Podcast listener for joining us on this show. Feel free to leave us comments and questions that leave a comment section under each episode on the website. We'd love to hear from you. Thanks a lot, everybody.
[00:43:14] [END OF AUDIO]
Amara Orth, 18, of Glenwood, Iowa, developed a method to identify vibroacoustic patterns of honeybees which reflect the health of the hive, to predict hive collapse and help protect her family’s bee farm for her Regeneron Science Talent Search computational biology and bioinformatics project.
Working at home and in the family barn, Amara measured the sounds and vibrations from the bees in 25 hives from August to November 2021 and then analyzed the data using a mathematical model she developed. Her system predicted bee colony health with 92% accuracy. She hopes this will provide beekeepers with an early warning for hive collapse and give them time to intervene. She plans to expand the sound library and make her system available to other beekeepers.
Amara also founded a nonprofit organization that informs beekeepers of the pollen and chemical composition of their honey.
An avid soccer player, Amara attended Lewis Central High School in Council Bluffs, where she played on the varsity and club teams. Amara has also traveled internationally for bee conservation. In 2019, she used scholarship funds received from the Broadcom MASTERS competition to attend a bee conservation program in the Costa Rican cloud forest.
*This Biography has been modified from one posted on the Regeneron Science Talent Search (STS) Website regarding the 2022 STS finalists
President and CEO of Society for Science/Publisher of Science News
Maya Ajmera is the President and CEO of Society for Science and Publisher of its award-winning magazine, Science News. Founded in 1921, the Society works to promote the understanding and appreciation of science and the vital role it plays in human advancement. The Society is best known for its world-class science competitions, including the Regeneron Science Talent Search, Regeneron International Science and Engineering Fair and Broadcom MASTERS.
As an alumna of the Westinghouse Science Talent Search, now sponsored by Regeneron, Maya has helped to transform the nearly 100-year-old Society into a dynamic, entrepreneurial organization. During her second year as President and CEO, Maya spearheaded a $100 million, 10-year sponsorship for the Science Talent Search with Regeneron. In addition, Maya has brought Science News Media Group from a decade long deficit into a bullish enterprise by developing a new education pillar and diversifying SN Media Group’s income stream. She founded a new series of outreach and equity programs to reach more underserved STEM students in the United States with an $11 million investment.
Maya served as the inaugural Social Entrepreneur in Residence for Duke University and Visiting Professor for the Practice of Public Policy at The Sanford School of Public Policy at Duke. She was a Visiting Scholar and Professorial Lecturer at The Paul H. Nitze School of Advanced International Studies (SAIS) at Johns Hopkins University and continues to serve as an adjunct on the faculty of the International Development Program and teaches the course, “Social Innovations in International Development for Children and Youth.”
Maya founded The Global Fund for Children (GFC), a nonprofit organization that invests in innovative, community-based organizations working with some of the world’s most vulnerable children and youth. Under Maya’s 18 years of leadership, GFC grew from a vision into one of the largest networks of grassroots organizations working on behalf of vulnerable children. To date, GFC has awarded nearly $50 million to over 700 grassroots organizations in over 80 countries, touching the lives of over 11 million children worldwide.
Maya is also an award-winning children’s book author of more than 20 titles, including Back to School, Every Breath We Take, Children from Australia to Zimbabwe and To Be a Kid, with more than 5 million readers worldwide.
Maya is the recipient of the 2020 National Science Board Public Service Award, recognized for her tremendous contribution to increasing the public’s understanding of science and engineering. Maya is also a recipient of the Henry Crown Fellowship at the Aspen Institute, the Echoing Green Fellowship, the William C. Friday Fellowship for Human Relations and the Rotary International Graduate Fellowship. She is sought out nationally and internationally to address audiences on STEM education, local and global philanthropy, global children’s rights, international development and social entrepreneurship. Her work and life story have been profiled by such media outlets as CNN, The Oprah Winfrey Show, Financial Times, NPR and many others.
Maya is a board member for Echoing Green, Kids in Need of Defense, Sibley Memorial Hospital and the North Carolina School of Science and Mathematics Foundation. She served as a member of the Innovation and Civil Society subgroup of the Obama presidential transition’s Technology, Innovation and Government Reform Policy Working Group.
Maya holds an A.B. from Bryn Mawr College and a M.P.P. from The Sanford School of Public Policy at Duke University. She lives in Washington, D.C. with her husband and daughter.