It started with a throwaway comment during a cleanroom tour on my first week.
My supervisor was walking me through the assembly process for the hexblock. The geometry is awkward. It doesn't sit flat, it doesn't hold itself in a convenient orientation, and assembling it requires you to constantly reposition it by hand. I watched the technician work through it and, without really thinking, said something like: "This looks really tedious. Is there a way to hold it better so assembly is easier?"
I thought I was just making small talk. My supervisor paused, looked at me, and said something I didn't expect: "That's a good question. Why don't you figure something out?"
A few days later, I had a project.
The Problem Worth Solving
Before jumping into design, I spent time just watching. The hexblock's odd geometry meant operators had to hold it in one hand while assembling with the other, constantly adjusting their grip to access different faces. Nothing about the process was dangerous or broken, it just had a lot of unnecessary friction. Every step required a small mental decision: How do I hold this right now? Is this aligned? Did that seat correctly?
Those aren't big questions, but they add up. In manufacturing, small inefficiencies don't stay small. When you repeat a process dozens or hundreds of times across a production run, every wasted second compounds. A task that takes thirty seconds longer than it needs to, repeated a hundred times, becomes nearly an hour of lost time. That realization, that the stakes of "small" inefficiency are actually pretty high at scale, was one of the first genuinely useful things I learned on the job.
Iteration, Not Inspiration
The first version of the fixture was basic by any standard. It was essentially a holding block: something that gripped the hexblock and kept it stable so an operator could work with both hands free. No moving parts, nothing clever. It was designed to prove that even a static fixture would reduce the constant repositioning problem.
It helped. But watching operators use it made something clear pretty quickly, holding the part steady was only part of the problem. To assemble the hexblock fully, you need access to multiple faces. A static fixture helped with one orientation and then became an obstacle for the next.
So the fixture evolved.
Version 1
A simple holding fixture. Stable grip, both hands free. Proved the concept had value, but exposed the need to rotate the part mid-assembly.
Version 2
Added spinning functionality so operators could rotate the hexblock without removing it from the fixture. Smoother transitions between assembly steps, but the part would shift slightly under load.
Version 3
Introduced a pin-lock system to index the hexblock at defined rotational positions. Snap into place, assemble, rotate to next position. Consistency improved significantly.
Version 4
Reinforced the pins after operator feedback flagged that the hexblock could shift under tighter assembly forces. The final version held firm across the full range of assembly steps.
Each version came out of conversations with the operators who were actually using it. That feedback loop was genuinely the most important part of the process. I could have designed something more elaborate from the start, added more features, more constraints, more degrees of control, but the fixture got better because I kept asking what's still annoying about this? and then went back and addressed it.
Measuring What Actually Changed
Once we had a version that felt stable and complete, we ran a formal time study. Stopwatch testing, comparing assembly with the fixture against assembly without it. It's a simple method, but it gives you a number you can point to.
Assembly that had taken around six minutes came down to two. Not from moving faster or working harder, but from removing the repositioning, the guesswork, the constant small decisions. The operators were doing the same task, just without the friction that had been baked into it for however long the process had existed.
That number felt significant. Not because of any individual design choice I made, but because it showed how much hidden inefficiency had been sitting in a process that everyone had just accepted as normal.
What I Actually Learned
I came into this internship expecting that the interesting engineering problems would involve something technically demanding, complex calculations, difficult tolerances, sophisticated systems. And maybe those problems exist elsewhere. But the thing that actually made an impact during my time there was something much simpler: paying attention to how people were working and asking whether it could be easier.
The fixture itself isn't a complicated piece of engineering. It's a holding device with a rotation mechanism and a pin-lock system. A mechanical engineering student could sketch the concept in an afternoon. What took time, and what actually made it useful, was the iteration. Understanding the process well enough to know what was worth solving. Watching operators and listening when they said "this part still catches." Going back and changing things instead of defending the original design.
Manufacturing engineering, at least from where I was standing, is largely about reducing friction. It's not always about building something new, it's about finding the places where a process asks more of people than it needs to, and then designing something that asks less. Consistency matters enormously on the floor. When a process is inconsistent, quality becomes dependent on individual skill and attention rather than on the system itself. A good fixture doesn't just save time. It makes the outcome more predictable, regardless of who is assembling on a given day.
I also learned that an offhand comment in a cleanroom can turn into a real project. That the problems worth solving are often already visible, they just need someone to notice them and ask if they can be better. I didn't arrive with a solution. I arrived with a question. And that, more than anything else I did, was what got the project started.
The fixture is still in use. I find that genuinely satisfying, not because it's technically impressive, but because it's useful. And in a manufacturing environment, useful is everything.