William Dong

ProblemDrones can't carry heavy payloads for long periods of time.

SolutionCreate an attritable ground vehicle.

ProblemCommercial RC cars aren't designed for modular aero components or custom optimization.

SolutionBuild one myself.

ProblemThe Push Back game rewards tall bots with confident collision capabilities.

SolutionBuild for a low center of gravity, durable design, and speed.

ProblemA NEMA 24 stepper makes modest torque at speed; an actuator wants high torque, low speed, low backlash.

SolutionWrap the motor in a compact 20:1 cycloidal reducer.

ProblemSimulation only gets you so far. FTAD needed to validate aero on real hardware.

SolutionDesign an open-circuit wind tunnel and CFD-validate it first.

ProblemHulls have to wrap real parts, but I had no quick way to tell whether a shape would drag well.

SolutionBuild a sketch-to-curve tool with its own wind tunnel, then check it against published drag data.

ProblemMost IK solvers hand back joint angles without showing where they came from.

SolutionBuild one that moves the arm and writes out the math behind every joint angle.

ProblemCycloidal disk profiles are punishing to model by hand and don't rebuild cleanly when you change a drive parameter.

SolutionWrite a FeatureScript that generates the whole disk from the drive geometry.

ProblemI was doing my materials science homework and wanted to check my work.

SolutionSo I built my own visualizer to see the planes in 3D.

ProblemTeaching spatial concepts is hard. Younger fighters can't process descriptions as fast as speech.

SolutionBuild a simple program that models positioning.