I am still vibrating with a very specific, very sharp kind of rage, the kind that settles in the base of your skull and makes the glow of a 4K monitor feel like a physical assault. It’s been exactly 11 minutes since that silver SUV swerved into the parking spot I had been signaling for, and as I sit here staring at the terminal curve of a lowercase ‘g’ in my newest typeface, I can’t help but see the parallel. The driver didn’t see me, or didn’t care to. The system-the painted lines, the unspoken social contract of the parking lot-failed. But if you asked the property manager, they’d say it’s a ‘driver dispute’ and go back to their coffee. It is the easiest thing in the world to blame the person at the end of the chain.

Yet, in the industrial world, we still see the same tired conclusion at the end of every investigation: ‘Root Cause: Operator failed to follow established procedures.’ It is a phrase that should make every engineer and manager feel a deep sense of shame, but instead, it is used as a get-out-of-jail-free card.

The Theater of Retraining

Imagine a safety review meeting in a room that smells faintly of floor wax and stale air. The 51st slide of the morning pops up. It’s a graph showing a spike in pressure that led to a system shutdown. The investigator, probably wearing a shirt that needs an iron, points a laser at the screen and says the operator skipped step 31 of the manual. No one asks why the manual is 171 pages long. No one asks why the interface requires 11 separate clicks to confirm a safety-critical action. No one asks if the operator was on the 11th hour of a graveyard shift, squinting at a screen designed by someone who has never stepped foot on a factory floor.

We love to talk about retraining. It’s cheap. You pull a guy into a room, show him some slides, have him sign a paper saying he understands the rules, and you’re legally covered. But retraining is almost never the answer because the error was almost never about a lack of knowledge. It was about a lack of capacity or a failure of the system to align with human psychology.

“

Retraining is often just a ritual of public shaming disguised as professional development.

I once made a mistake in a type specimen where I accidentally used a zero instead of a capital ‘O’ in a technical headline. I knew the difference. I wasn’t ‘untrained’ in the alphabet. I was tired, the two glyphs were optically similar, and my software didn’t flag the discrepancy. Blaming my ‘failure to proofread’ is technically true, but it doesn’t prevent it from happening again. Fixing the glyph geometry does.

The Cost of Cognitive Laziness

This is the core of the problem: blaming the operator is the intellectual endpoint of lazy management. It’s a way to close a file without having to spend $501,000 on a control room overhaul. It’s a way to avoid admitting that the ‘optimized’ process you designed is actually a cognitive nightmare. When we stop at ‘human error,’ we stop learning. We leave the trap set for the next person to walk into it. It’s like designing a door that looks like it should be pulled, but it actually needs to be pushed, and then calling the person who pulls it an idiot. If 11 different people pull that door in a single morning, the door is the problem.

I think about this constantly when I’m refining the x-height of a font intended for industrial displays. If the difference between a ‘3’ and an ‘8’ is only a matter of 11 pixels, I am setting someone up for a catastrophe. I am building a cliff and forgetting to put up a guardrail.

There is a specific kind of arrogance in assuming that a human should be as consistent as a microprocessor. We aren’t. We have blood sugar fluctuations, we have distracted thoughts about our stolen parking spots, and we have physiological limits. A truly resilient system is one that assumes the operator will be tired, distracted, and prone to shortcuts.

It’s why companies like Sis Automations are so vital to the modern landscape; they don’t treat the human as a variable to be controlled, but as a biological reality to be supported through intuitive, error-proof design.

The Swiss Cheese Model Re-examined

If we actually cared about safety, the word ‘retraining’ would be banned from incident reports for the first 31 days of an investigation. We would look at the environmental factors first. Was the alarm sound too similar to the ‘task complete’ chime? Was the critical data point buried under 21 layers of sub-menus? When we look at the ‘Swiss Cheese’ model of accidents, the human is usually just the last layer. There were 11 other holes in the system that had to align perfectly for that one mistake to matter. By the time the operator makes the ‘error,’ the system has already failed them 41 times over.

“

Design is the silent partner in every safety record.

I remember reading about a chemical plant where a valve was left open. The report blamed the technician for ‘failure to verify valve state.’ What the report didn’t emphasize was that the valve handle was identical to 11 other valves in the same row, and it was located in a corner where the lighting was less than 31 lux. The technician had performed that task 1,201 times without an issue. On the 1,202nd time, they were thinking about their kid’s fever. If the valve had been color-coded or had a physical lockout, the fever wouldn’t have mattered.

Symptom vs. Root Cause

We have to stop treating ‘human error’ as a root cause and start treating it as a symptom. It’s a pointer. It’s a big, red arrow saying, ‘Look here, the design is unintuitive!’ or ‘Look here, the pressure to perform is outweighing the capacity to be safe!’ But instead, we fire the tech, hire a new one, and wait for the 51st of next month for it to happen again. It’s a cycle of stagnation that costs lives and millions of dollars, yet we cling to it because it feels good to have someone to point a finger at. It’s the same satisfaction I felt when I imagined keying that silver SUV, even though I knew deep down that the parking lot’s layout is what actually caused the confusion.

“

As I wrap up this font-I’m calling it ‘Sentinel 1’-I’m making sure the aperture of the ‘6’ is wide enough that it can never be mistaken for a ‘5’ or a ‘0,’ even if the screen is cracked. I’m doing my 1% to make the world a little more error-proof.

But design can only do so much if the culture remains obsessed with blame. We need a fundamental shift in how we view the relationship between the tool and the user. The tool should serve the user’s humanity, not demand that the user transcend it.

Designing for Reality

Next time you see a report that ends with ‘operator error,’ I want you to ask why the system was so fragile that a single human mistake could break it. Ask why the 11 safety layers didn’t catch it. Ask if the interface was designed for a human or for a ghost. Until we start asking those questions, we aren’t practicing safety; we’re just practicing theater. And personally, I’m tired of the show.

I’d rather have a parking lot with clear signage and a control room that doesn’t treat me like a machine. It’s not about being perfect; it’s about being real. And reality is messy, inconsistent, and deserves a better typeface than the one we’re currently using to write our excuses.