The most important failure of my early engineering career: a mechanically ambitious intake system that taught me more about scope, integration, iteration, and design-for-manufacturing than any project that simply worked the first time.
This was my junior year and my first season on Design-Integration, the subteam responsible for CAD and design work for the mechanical team. I had just earned my CSWA certification and was eager to use SolidWorks on a full robot.
The team had also just gone through a major leadership restructuring, and DI was extremely disorganized. I was assigned ownership of the mechanism that would collect and outtake cones and inflatable cubes.
After a design discussion that was far too short, we settled on a claw. My concept used freely rotating claw pads so a cone could automatically rotate with its heavier base downward before being deposited.
That gravity-driven self-orientation is where the mechanism got its name: the Gravity Claw.
The first prototype opened like a cymbal monkey toy: each arm pivoted about one axis, the two sides were synchronized by meshing gears, and pneumatic pistons powered the motion.
It opened and closed as expected, but angular misalignment of the pads ruined the gravity-rotation effect. That failure pushed the design toward four-bar linkages to keep the contact surfaces aligned.
The four-bar linkage fixed pad alignment, but this version was far too bulky.
The final design split the claw into two independent CNC-machined polycarbonate four-bar modules.
Other members of DI were designing the “slurper” intake and the mechanism that would move my claw. The intended arm used two aluminum telescoping stages pivoting from a raised, chain-driven axis.
Deep into build season, it became clear that both the slurper and the telescoping “pink arm” were far beyond what our team could realistically manufacture and integrate in time.
In the final week before competition, almost the entire concept was scrapped. We kept the Gravity Claw and frantically built a simpler articulating arm around it — so late that there was never even proper CAD of the competition configuration.
I was the driver at both competitions, and it became obvious almost immediately that our scoring mechanisms were not competition-ready. The claw and arm were cumbersome to operate and broke after nearly every match.
Eventually we literally tied the arm down. But the rushed build had created one accidental advantage: our robot was almost exactly the maximum permitted size and weight.
In 2023, defense was highly viable. I learned to use the robot's mass aggressively, pushing opponents away from scoring positions and turning a failed scoring robot into an effective defensive machine.
Our programming captain added a wheel-locking function that rotated the swerve modules into an X formation, making the robot extremely difficult to move.
The endgame required alliances to balance multiple robots on a table-sized seesaw. We could balance, drive partially off the edge to clear room for our partners, then lock the wheels and hang there without sliding off.
Despite the failed original design, our alliances reached the finals at both competitions we attended. We lost both finals, but the season became unforgettable for completely different reasons than we expected.
A mechanism can be clever and still be the wrong design if the team cannot manufacture, integrate, test, and repair it within the available time and resources.
Several warning signs were visible long before competition. I learned that noticing a problem is not enough — engineers have to raise it, challenge assumptions, and force difficult conversations before the schedule makes change impossible.
The Gravity Claw improved dramatically across three generations. Each failure taught me something about alignment, packaging, linkages, integration, and manufacturability that CAD alone could not.
