By moving like the creature it mimics, it minimizes its own footprint
In Burlington, Ontario, a fifteen-year-old named Evan Budz has offered the ocean something rare: a watcher that does not frighten what it watches. By fashioning an autonomous robot in the form of a snapping turtle and training it with artificial intelligence, Budz has created a device that detects pollution, bleaching, and invasive species with striking precision — all while moving as though it belongs. His invention, tested in a backyard pool and honored across two continents, raises a quiet but consequential question about who gets to imagine the future of environmental stewardship.
- Coral reefs and coastal ecosystems are degrading faster than conventional monitoring tools can track them, leaving scientists with incomplete pictures of accelerating harm.
- Most underwater research technology is expensive, disruptive, and inaccessible to the communities that need it most — a gap that has long stalled meaningful ocean surveillance at scale.
- BURT moves like a turtle precisely so it doesn't behave like a machine, allowing it to gather data in sensitive ecosystems without triggering the flight responses that compromise other monitoring methods.
- With 96% accuracy in coral bleaching detection and the ability to identify microplastics and invasive species, the prototype has already demonstrated scientific credibility far beyond its origins in a grandparent's swimming pool.
- Back-to-back wins at Canada's national science fair and the EU Young Scientists Competition in Latvia have thrust Budz and his design into international view, where the real test — scaling the solution — now begins.
Evan Budz was fifteen when he built a robot that swims like a turtle and reasons like a scientist. Working out of Burlington, Ontario, he designed BURT — a biomimetic autonomous underwater device modeled on a snapping turtle — and first tested it in his grandparents' pool. The choice of form was not aesthetic. By mimicking a turtle's four-flippered movement, BURT navigates marine environments without startling the creatures that live in them, reducing the very disruption that makes so much ocean research difficult.
Inside its shell-shaped body, a Raspberry Pi microcomputer runs AI models trained to recognize environmental threats. Cameras and advanced sensors feed data to the system, while a solar-assisted lithium battery provides up to eight hours of autonomous operation. The cost to build remains deliberately low — a detail that matters enormously if the design is ever to move beyond prototype.
What BURT can do is what sets it apart. It detects coral bleaching with 96 percent accuracy, tracks microplastics moving through water columns, and identifies invasive species — all without leaving a mark on the ecosystems it studies. The scientific community responded quickly: in June 2025, Budz won Canada's top science fair honor, and three months later took first place at the European Union Young Scientists Competition in Latvia, competing against hundreds of peers from forty countries. Standing at the awards ceremony, he admitted it still felt unreal.
Budz speaks about his work with unusual clarity for his age. He wants BURT to study real ecosystems and address specific threats like coral bleaching — not by imposing technology on the ocean, but by sending something the ocean won't resist. The deeper promise of what he has built is a model for monitoring our seas through devices that move with marine life rather than against it, designed by someone young enough to imagine what hasn't been tried.
Evan Budz was fifteen when he built a robot that could swim like a turtle and think like a scientist. The Burlington, Ontario student started the project in 2024, designing and constructing an autonomous underwater robot he called BURT—a biomimetic device modeled after a snapping turtle, equipped with artificial intelligence capable of detecting water pollution and environmental degradation. He tested it in his grandparents' swimming pool.
The robot's design is elegant in its simplicity. Four flippers provide stability and mimic the movement of an actual turtle, allowing it to navigate without startling marine life. Inside the shell-like body runs a tube housing the electronics: a Raspberry Pi microcomputer running AI models trained to identify environmental threats, along with cameras and advanced sensors. A lithium battery paired with solar panels gives BURT up to eight hours of autonomous operation—enough time to gather meaningful data from a reef or estuary and return.
What makes the device remarkable is what it can actually do. The prototype detects coral bleaching with 96 percent accuracy, monitors the spread of microplastics through water columns, and tracks invasive species—all without disturbing the ecosystem it observes. By moving like the creature it mimics, BURT minimizes its own footprint on the environment it studies. The cost to build remains low, a detail that matters enormously if the design is ever to be deployed at scale.
The scientific community took notice quickly. In June 2025, Budz won the award for Best Innovation Project at Canada's national science fair. Three months later, in September, he took first place at the European Union Young Scientists Competition in Latvia, competing against hundreds of young researchers from forty countries. The recognition arrived suddenly, almost too much to process. "It still feels unreal," Budz said at the awards ceremony, standing amid the formal reception, still finding his footing in the moment.
When asked about his vision for the work, Budz spoke with the clarity of someone who understands both the problem and his tool's potential. He wants to use BURT to study ecosystems and tackle specific threats like coral bleaching. By designing a robot that mimics a turtle rather than imposing human technology on the ocean, he reduces the disruption his research causes. The animals living in the water won't flee from something that looks and moves like a natural part of their world.
What Budz has built is not just a clever engineering project. It represents a shift in how we might monitor our oceans—through devices that work with marine environments rather than against them, powered by artificial intelligence that can learn to recognize threats humans might miss, and designed by someone young enough to imagine solutions without the weight of how things have always been done. The question now is whether this prototype, proven in a grandparent's pool and validated on Europe's biggest stage, can scale into something that actually changes how we watch over the water.
Notable Quotes
It still feels unreal— Evan Budz, upon winning the EU Young Scientists Competition
By imitating a turtle, I reduce my impact on the ecosystem and don't disturb the animals living there— Evan Budz, on his design philosophy
The Hearth Conversation Another angle on the story
Why design a robot shaped like a turtle instead of, say, a submarine or a drone?
Because a turtle belongs in the water. Fish don't scatter when a turtle swims past. The ecosystem doesn't treat it as a threat. If you send down something that looks and moves like a machine, you change the behavior of everything around it. The data you collect becomes less reliable.
And the AI component—what's it actually learning to recognize?
Patterns of stress in coral, the presence of microplastics, thermal anomalies that signal bleaching events. The system was trained on thousands of images and sensor readings. It's not guessing. It's identifying things with 96 percent accuracy.
That's a very specific number. What happens in the four percent of cases where it's wrong?
You build in redundancy. You send multiple units. You have human verification. No single tool is perfect, but a tool that's right 96 times out of 100 is infinitely better than no tool at all, or than waiting for a human diver to spot the problem.
Why does it matter that a fifteen-year-old built this and not a team of engineers at a research institute?
Because he wasn't constrained by institutional thinking. He didn't know what was supposed to be impossible. He built it in his grandparents' pool with affordable components. If a teenager can do this, imagine what becomes possible when the design is open-sourced, when others improve it, when it's deployed in hundreds of locations.
What's the real problem he's solving?
We're losing our oceans and we don't even know how fast. We lack the tools to monitor what's happening at scale. This is a tool that could actually work.