Let’s talk about a future without microplastics with Anna Du
At only 12 years old, Anna Du invented a robot to detect microplastics in our oceans. On this episode of “Let’s talk about it”, Michael sits down with Anna to find out where the idea for her robot began, what challenges she’s faced along the way, where she finds inspiration and what she has planned for the future.
ML (Michael Londesborough): Today’s guest is 13-year old Anna Du from Massachusetts, USA. She is a middle school student who is tackling microplastics in our oceans using her own innovative ROV, a remotely operated underwater vehicle, to detect and locate microplastics under water. Anna started out being concerned about this issue and is now playing an active role in its solution.
How did you become invested in the microplastic problem?
AD (Anna Du): I love going to the beach for sea grass. New England sea grass is extremely beautiful and makes a great gift for friends. One day while picking it up I noticed that a piece didn’t feel right. It was shinier and felt different from the others. I thought it was a piece of plastic. Then I started looking around and seeing plastics everywhere. I couldn’t stop thinking about it so I decided to do something.
ML: You decided to make a change. First by picking up that plastic. Then you realized you could do it more effectively? Thought of a more efficient solution to this problem?
AD: Yes, I began by trying to clean up pieces of plastic. Then I thought it would make more sense if I were able to locate where microplastics are aggregating before trying to clean them up. This would allow us to target our resources into areas where we already know they’re located. In the future I also plan on determining common environmental factors in places with high microplastic aggregation. I’ve been studying factors such as turbidity, pressure and temperature to see if they have an effect on their aggregation.
ML: Your robots can go to the bottom of the ocean and find the plastics. How do they do that?
AD: I have three different detection systems. The main one uses an infrared detection mechanism. It works by exposing the sample to different wavelengths of infrared light, comparing the images to thermographs and creating a color map which then clearly shows the differences between plastics and non-plastics. I’ve also developed two supplemental systems which use artificial intelligence to make the main system more adaptive. One of them is called the unnatural color detection system. It identifies microplastics by identifying colors indicative of the presence of synthetic dye. The second is is a morphology classifier, which identifies shapes common in mircoplastics and classifies them into various categories.
ML: The gathered information allows you to create a picture of the factors influencing plastic accumulation and then you can guide the robot there? How does that work?
AD: The ROV takes pictures, stores them on a USB drive and takes them to a computer back on shore which analyzes the collected data. Once I add a sensor to the ROV, it’ll analyze that data as well, allowing me to determine where microplastics are aggregating.
ML: Why should we be concerned about microplastics in our water ways?
AD: Microplastics have already been shown to have an effect on the environment and they’ll affect many types of species in the future. Not only are microplastics devastating to the ecosystem, they are also very detrimental to human health. Scientists have been studying how microplastics affect human health and have found a possible correlation to autism.
ML: Are you working with people on developing technology to also collect this plastic waste?
AD: I’ve been looking at collection methods, but right now I’m mainly focused on the identification process.
ML: What was the most difficult engineering problem you had to solve?
AD: One big problem I faced was a lack of data. Thousands of data points are needed to train artificial intelligence, even for a very simple classification area. It would have been extremely difficult to collect thousands of photos with the ROV, so I ended up using a process known as data augmentation. By using data augmentation, I was able to emulate numerous micrographs and train my AI with those.
ML: Sounds like quite a big project. Who has helped you along the way? What network of people enabled you to realize your project goals?
AD: My parents were very supportive and encouraging. I’ve also been in contact with field experts who specialize in ocean research as well as speaking with people at spectroscopy companies about how plastics would look under infrared light.
ML: If you were addressing other young people who are looking to make a difference in the world, what advice would you give them?
AD: When trying to solve some of the world’s biggest problems, you have to stay motivated. It doesn’t matter how hard it may seem or what difficulties you may encounter along the way, allow yourself to think about all sorts of crazy ideas and one of them might just work.
ML: So you’re using infrared light to distinguish between different types of material. And it has to work in a water environment, where there is a scattering of light. I can imagine a whole host of engineering problems to overcome. What’s the engineering process like?
AD: I started by researching what might work best and came up with the infrared detection system- a lot of recycling companies use infrared to identify plastics. After that I created my own chip using a copper etch piece of plastic and etched my design onto it.
ML: Can we expect thousands of your robots going around ocean beds and determining where microplastics are accumulating? Is that the future?
AD: Once I identify where microplastics are aggregating on the ocean floor I plan on using a swarm of ROVs to look at them and potentially clean them up.
ML: Do we have an idea how it might work? Would a ship above, which could communicate with the robot, send and receive information to generate a picture of the ocean floor and process the data? How difficult will it be to send and receive information from the robot?
AD: At this point the ROV would have to communicate with some device, either on shore or on a boat. One of my main goals, however, is to make it more autonomous in the future. Ideally it could identify and clean up microplastics on its own.
ML: What source of energy powers these devices?
AD: Currently, because it’s tethered, I am using a power source on shore. For the future I’ve been thinking about using solar or maybe even wind power.
ML: Tell me a bit about your plans for the future.
AD: I hope to be able to go to MIT to study environmental engineering. The environment is very important and I want to do my part in making the world a better place. I’ve also always loved engineering so I thought I would combine these two things.
ML: What would you wish for your robot in terms of technological improvement?
AD: Right now I’m really working on the artificial intelligence aspect of my project and I’ve been running into some issues as to how well classifications are working because in some areas it’s working really well but in others it still needs a little bit of work.
ML: I wish you all the luck in the world with all of your ideas for the future sustainability of our planet. Thank you very much for speaking with me today and I wish you the best of luck in all that you do.