From Chaos to Clarity by Arnav

From Chaos to Clarity by Arnav - October 2026 Scholarship Essay

Every successful business has a major problem in year one. Mine was a massive pile of LEGO bricks.

Preparing one rental set required two grueling hours of sorting. The work was tedious, but the larger problem was scalability. My solution was a machine that could identify and sort bricks automatically.

Building it pulled me into mechanical design, electronics, computer vision, lighting, and safety. I constructed a sleek metallic frame, installed a rotating camera, and calibrated the system to distinguish among LEGO colors. In one room, it worked. In another, it misclassified nearly every brick.

The variable was ambient light. I experimented with coverings, camera angles, and recalibration, improving accuracy without eliminating the inconsistency. What began as a business problem became a question about color constancy: how can a vision system recognize the same color when illumination changes the values recorded by its camera?

I am now investigating whether the strongest solution is physical, computational, or both. Should I create a sealed chamber with standardized illumination, normalize camera data against a reference color, or train the classifier on images captured under varied lighting?

I still want to determine which approach produces reliable accuracy without making the machine too expensive or slow to scale.