Wellington's robotics students are celebrating a major win after competing in the National Robotics Challenge this week. Their custom-built autonomous vehicle, created by upper school students Bean Fischer ’26 and Hannah Guffrey ’25, earned 1st place in the high school division of the Autonomous Vehicle Challenge — and received the Honda Innovation Award for their event. Out of more than 500 robots entered across the competition, only 13 were nominated as finalists for the Innovation Award, which recognizes the most inventive and creative designs. Bean and Hannah’s robot was named the Innovation Award winner for the Autonomous Vehicle Challenge, while another team took home the overall competition-wide prize. Notably, their vehicle outperformed the college teams competing in the same challenge.
But this isn’t your average science fair project. Bean and Hannah spent months designing, coding, and building a robot that could navigate a complex obstacle course — entirely on its own.
The challenge, known as the Autonomous Vehicle Challenge, is deceptively simple: complete a course filled with brightly colored buckets, ramps, and a single arch as quickly and accurately as possible. But there’s a catch. Once the robot starts, no one can touch it — it has to operate independently, using only its onboard code and sensors. “You get more points for going over a ramp instead of around it or through the arch instead of avoiding it,” said Hannah. “It’s about speed, yes, but also about strategy and precision.”
To tackle this, the team focused on building a sophisticated vision system — essentially teaching the robot how to "see" the course and react in real time. “That’s been the hardest part,” Bean admitted. “You’re trying to get a machine to recognize a bucket and understand where it is in space. And not mistake someone’s red jacket for an obstacle.”
This year’s success didn’t come out of nowhere. Two years ago, Bean attempted the same challenge solo — and the robot barely made it around one bucket. The camera disconnected, and the robot lost its way. Rather than give up, Bean returned this year with renewed determination, teaming up with Hannah and drawing on what they had learned from past mistakes. That process — trying, realizing there’s a problem, and working to fix it — is at the heart of engineering. It’s also what makes their progress so impressive. Throughout it all, they were supported by Dr. Chris Fischer P ’24 ’26, Wellington’s engineering consultant, whose steady mentorship helped guide their development as both programmers and problem-solvers.
Over the course of the year, the team tested three different approaches to computer vision. The first method used color detection, filtering the camera feed for certain hues and size; it found the largest continuous “blob” of the correct color. The second added edge detection, allowing the robot to analyze the shape and position of what it was seeing. But the third — and most successful — method involved training a neural network using hundreds of tagged images. The result is a machine learning model that can recognize buckets more quickly and more accurately.
All of this is powered by a Jetson Nano — a compact computer designed for artificial intelligence applications. The students also wrote custom code to integrate the Nano with their motor control systems and a gyroscope, allowing the robot to sense its orientation and adjust in real time. During the competition, that feature proved critical: after a wheel clipped the arch and spun the robot sideways, it used its internal guidance to reorient itself and keep going — ultimately completing three-quarters of the course. “We built it all ourselves — the frame, the wiring, even a library to help the motors communicate with the computer,” the students said. “It’s definitely not plug-and-play.”
While the technical accomplishments are impressive, the process stands out the most. Many of these students have been building and iterating on robots for years — learning from failed runs, coding errors, and vision systems that didn’t quite cooperate. This year, their robot (named Fish with Wheels, a nod to the fish-eye camera they initially used before switching to a standard webcam) not only completed the course — it earned top honors. Along the way, they demonstrated real-world problem-solving, including an on-the-spot fix when bright sunlight washed out the camera’s view. With no time to spare, Bean and Hannah improvised a lens cover using duct tape and the tinted lenses from a pair of Honda promotional sunglasses — and just like that, the robot could see again. Their 1st place finish and Innovation Award win are a testament to Bean and Hannah’s creativity, persistence, and problem-solving. They didn’t just meet the challenge — they redefined what was possible.