Traditional ocean and lake monitoring is shockingly outdated. Scientists typically drop heavy nets into the water, pull up samples, ship them to a laboratory, and spend days or weeks squinting through microscopes to count microplastics. It's expensive, painfully slow, and misses vast stretches of our waterways.
Evan Budz, a 16-year-old student from Burlington, Ontario, decided there had to be a better way.
Instead of waiting around for traditional research labs to modernize, Budz spent nearly 4,000 hours in his workshop creating an autonomous bionic turtle named BURT. Equipped with a custom 3D holographic camera and edge-computed neural networks, his underwater robot swims through lakes and rivers, identifying microplastics in real time with 94% accuracy down to particles smaller than 10 microns.
His work earned him the $50,000 Gordon E. Moore Award at the 2026 Regeneron International Science and Engineering Fair. Beyond the competition accolades, his prototype offers a blueprint for how environmental monitoring needs to evolve.
Why Nature Outperforms Traditional Underwater Drones
Most aquatic survey tools look like bulky torpedoes or tethered ROVs. They require heavy batteries, churn up sediment, and startle local marine life with aggressive thrusters. Budz took a different path after watching a snapping turtle glide through freshwater during a 2024 camping trip in Ontario's backcountry.
Turtles move with incredible hydrodynamic efficiency. They use their front flippers for primary propulsion while employing their back legs as rudders for precise steering. Budz replicated this exact mechanics in SolidWorks and 3D-printed a custom flipper propulsion system.
By mimicking turtle locomotion, the robot achieves several key design advantages:
- Extended operational lifespan on a single charge, running up to eight hours at roughly 0.5 miles per hour.
- Low acoustic and mechanical disturbance, allowing the robot to slip through fragile aquatic ecosystems without disrupting habitats.
- A wide, stable shell geometry that easily houses internal processing hardware, power systems, and specialized sensors.
Early prototypes weren't an instant success. Initial pool tests resulted in the chassis sinking straight to the bottom due to buoyancy control failures. Budz spent months consulting local aquarium experts, running computational fluid dynamics simulations, and tweaking the flipper oscillation frequencies before getting the balance right.
Identifying Plastic Particles on the Fly
Detecting microplastics under water without bringing physical samples back to a lab is a massive engineering hurdle. Standard cameras fail because murky water, floating biological matter, and tiny plankton confuse visual sensors.
Budz solved this by building a custom 3D holographic imaging device encased inside an acrylic tube.
Digital holography doesn't capture flat 2D images. Instead, it records three-dimensional light interference patterns, granting high spatial resolution across a large depth-of-field without requiring complex motor-driven refocusing lenses.
To turn those raw holograms into actionable data, Budz integrated an onboard microcomputer running custom machine learning models trained on polystyrene, polyethylene, and PET microplastic samples. The AI categorizes particles on the spot, separating synthetic polymers from natural organic debris and microorganism cells.
During field validation across 10 different Canadian lakes and waterways, the platform maintained its 94% accuracy rate while executing pre-programmed search grids autonomously.
The Bigger Shift in Environmental Science
What makes this project stand out isn't just that a teenager built it in his spare time. It demonstrates a major shift toward decentralized, real-time environmental tracking.
For decades, water quality data has suffered from massive gaps because gathering physical samples costs too much money and takes too much human effort. Deploying small, low-cost autonomous agents like BURT makes continuous tracking viable for municipal governments, conservation groups, and independent research teams.
Beyond microplastics, Budz has already adapted the platform to flag coral bleaching with 96% accuracy and log invasive species through automated image analysis.
What Comes Next for Low-Cost Aquatic Sensing
If you want to track water contamination in your local ecosystem, waiting for federal environmental agencies to run multi-year surveys isn't your only option anymore. DIY bio-inspired robotics combined with localized machine learning models are rapidly dropping the barrier to entry.
To apply this kind of technology in your own environmental projects:
- Shift focus from physical sample collection to in-situ optical detection to cut laboratory overhead.
- Utilize digital holography rather than traditional lenses when imaging particles under 50 microns in variable water clarity.
- Design bio-inspired propulsion systems when working in shallow, plant-heavy, or sensitive freshwater habitats where propellers tangle or cause disruption.
