Field note
A Prettier Version of the Same Bad Day
A customer is paying me to build them a modern operator experience, and recently they asked me to recreate the exact screens that make their current experience miserable. Same cramped layout. Same wall of data. New colors.
This one bugs me to no end, so let me explain why.
The point was a better day, not a prettier one
The whole reason you bring in new tools (AI, live data operations, a modern UX layer) is to give the operator a better day at the machine. Not a prettier version of the same bad day. The old HMI screens I keep being handed all have one thing in common: they try to show everything about a step at once. Every tag, every counter, every readout an engineer could think of, crammed onto one screen for a single route step. Half of it has nothing to do with what the operator is actually doing right now.
That is not an operator screen. That is a fat client from fifteen years ago wearing a fresh coat of paint.
Picture the operator mid-step. They need the next action, the one value that tells them it worked, and a clear way to flag a problem. Instead they get forty fields, three of which matter, and they have to hunt for those three every single cycle. We call that "complete." The operator calls it noise. Over an eight hour shift, that noise is fatigue, missed signals, and errors we then blame on the operator.
The belief I want to kill
Here it is: that operators want dense, complicated screens, and the best we can do is rebuild them faithfully. They don't, and it isn't. A good operator screen shows you what you need to finish the step in front of you, and it gets quiet when things are going well. Information density should be inversely proportional to how well things are going. When the line is healthy, the screen is nearly empty. When something breaks, that is when the detail and the AI recommendations show up, right where the hands are. That is the point of the new tools. If the best thing I build is your old screen in new colors, you didn't need me, and you definitely didn't need the AI.
You don't need me to move a button
There is a second thing on this project that gets under my skin, and it is related. The customer wants me to make every change. Move a button, relabel a field, reorder a step. In the old world that meant a developer, a ticket, a sprint. In the world we actually live in now, it means you say, in plain English, "move this button from here to there," and it moves. You don't even need a full sentence. A phrase does it. You are not a developer and you do not have to become one. The platform handles that part for you.
So when I say "you can do this yourself," I am not pushing work back onto the customer. I am handing them the thing they actually bought: the ability to shape the operator's screen in the moment, without me in the loop. That is a feature, not a chore.
Too far removed from the work
I am too far removed from this one. I haven't met the operators. Not one of them. I have not stood on this plant floor. Everything I know about what these operators "want" is reaching me secondhand, through one person who talks to me and is interpreting what they believe the operators need. And their interpretation is: rebuild the old screens.
I was an operator (I'll admit, nearly thirty years ago). I know what it feels like to be handed a screen designed by someone who has never run the machine. So this is what I would ask of any customer: let me talk to your operators. Directly. I do not need you to translate them for me. The translation is where the design goes wrong. The person on the line will tell me in thirty seconds what they actually look at and what they ignore. No requirements document ever captures that.
I guess there is a trust issue I need to address here, and I have not cracked it yet. When a customer will not open that door, I cannot promise the best possible operator experience, because I am designing for a person I have never met, from a conference room, through a game of telephone. I can build exactly what I am told. I just cannot call that operator-centered, because no operator was ever in the room.
Where that leaves me
A little frustrated, mostly determined. I will keep making the case, because it is the right one. The new tools exist to simplify the operator's world, not to reproduce its clutter. Real simplicity at the glass takes real sophistication behind it, and it takes talking to the people standing at the glass.
Give me the operators. Let the screens get quiet. That's the whole idea.