Context

GBM Tech & Control builds port management software. Terminal operators use it to track containers, coordinate yard movements and keep a port running through shifts that do not stop.

I joined as a product designer and within a few months was leading two others. The work ran across two very different surfaces: enterprise dashboards for planners in the office, and tools used by people standing outside, on the yard, in the weather.

Constraints

Port work does not pause for a usability session. Getting real input meant going to where the work happened and watching it. Slower than a call, and the only thing that produced answers worth acting on.

The two audiences had almost nothing in common. A planner has a large screen, both hands, and time. A yard operator has a device, one hand, and a truck waiting. The same data had to work in both places.

The team was small and the product was large. Consistency could not depend on three people remembering the same thing.

Key decisions

I went to the terminal. I ran usability research in context, with operators, on the job. What came back changed the product in ways a remote session would never have surfaced, because the real problems were physical. Glare on the screen. Gloves. A device held at arm's length. A step that made perfect sense at a desk and none at all on the yard.

I built the design system before the team needed it. Three designers producing screens for two device classes will drift within a month. Putting the system in place early made consistency a byproduct of the tooling, so nobody had to police it in review.

I handed over reasoning, not mockups. With two designers on the work, the useful thing to pass along was the criteria behind a decision. Once those were written down, reviews got shorter and the second decision no longer needed me in the room.

I designed for the hardest position first. The yard, before the office. A pattern that holds up on a phone in direct sunlight, operated with one hand, holds up everywhere else.

What it produced

A shared system that survived three people working in parallel, and a research habit that put design decisions on top of observed behaviour. The whole story, with the field findings and the numbers, is in the animated deck linked below.