15-Year-Old Builds Bionic Turtle Robot That Swims Without Propellers

Evan Budz’s idea began at a campsite in Ontario, where he watched a snapping turtle move through a pond. The turtle left virtually no wake or ripples; the water closed behind it unchanged.

Budz then considered how a research robot would move through the same pond—by spinning propellers. That mismatch between the turtle’s quiet passage and a machine’s noisy propulsion became a science project.

Conventional underwater drones rely on propellers, which are noisy, create turbulence, and stir up sediment. The stirred mud and mechanical noise can scare fish, cloud the water, and damage the very ecosystem the drone is meant to observe. Budz recognized that the propulsion system conflicted with the mission, so he copied the turtle’s answer.

The robot is called BURT, short for Bionic Underwater Robotic Turtle. It swims like the snapping turtle Budz observed, propelled by a pair of flapping front flippers rather than propellers. The design copies a real sea turtle’s division of labor: the front flippers produce forward thrust, while smaller hind limbs handle steering and stability and do not participate in propulsion.

Budz modeled BURT in SolidWorks, 3D-printed its parts, and built it around a transparent tube holding electronics, with a thumb-sized camera at the front. The machine weighs about 11 pounds, roughly the size of a large cat; its lithium battery lasts up to eight hours, extendable with solar panels. Cruise speed is about half a mile per hour, similar to a real sea turtle. Every biomimetic choice serves one purpose: letting BURT pass through water without stirring it.

BURT is not merely a quiet swimmer; it is a quiet searcher for trouble. Its forward camera connects to a Raspberry Pi, a poker-card-sized computer running machine-learning models trained to recognize signs of ecological stress: coral bleaching, invasive species growth, and plastic trash. The robot follows a preset search grid autonomously, with no tether and no remote control. It swims its route, records what it sees, and can transmit data.

The real world later forced design changes, a part of the story many retellings skip. When Budz moved BURT from clear water into murky conditions, currents and shifting light defeated the camera. He added two elements not on the original drawings: front lights and an ultrasonic transducer that uses high-frequency sound to detect obstacles the camera cannot see. The problems of a real water column, not a tidy plan, produced the final machine.

The 96% figure is real, and it comes with the boundaries a skeptic should expect. In tests, BURT correctly flagged replicated coral bleaching 96 times out of 100. The word “replicated” is key. The corals were 3D-printed models; the water was a backyard pool at his grandparents’ house, just over eight feet deep, with stable light and fixed depth, and the machine had been shown beforehand what bleaching looked like. BURT later went to Lake Ontario, which is nothing like a pool, but the 96% belongs to controlled testing, not an open lake or a real reef.

That is not a dismissal; it is how a prototype should be reported. A 15-year-old established a clean baseline under controlled conditions, hit 96%, then took the machine into messy enough places to expose what the baseline missed. This is the shape of real instrument development, except it happened in a pool and was run by a tenth grader. Judges saw more than a school project. BURT won the Best Innovation Project award at the 2025 Canada-Wide Science Fair, competing among about 25,000 students nationwide, and later won one of the top awards at the EU Contest for Young Scientists in Riga against dozens of countries.

The idea beneath the awards is the durable part: a genuine critique of how nature is observed. Humans build noisy, wake-dragging machines to study animals that flee noise and wakes, then trust the data those machines bring back from an environment already disturbed by their arrival. A teenager watched a turtle that solved the problem 200 million years ago and copied the answer. The best way to observe a pond without changing it was apparently patented in the Mesozoic: fins, not propellers.