A midstream oil & gas plant's first piece of equipment can overflow without warning. I replaced ten scattered apps and a phone-call culture with one number, one decision: Go or No-Go.
A "Pig" pushes a "Slug" — a surge of liquid and dirt — through the pipeline toward the plant. If the plant doesn't know it's coming, the Slug Catcher overflows: hundreds of thousands of gallons of oil, a fire hazard, and soil contamination.
Midstream is the industry's middleman — moving product between pipelines, processing plants, and storage. The UX problem lived exactly at that seam: a visibility gap between what the field knew and what the plant could see.
Just to pull current tank data in the field, on a connection that might not hold. Six of seven operators told me SCADA itself was the failure — they'd started requesting Slug Catcher data on iOS through a separate tool instead, because SCADA had become unusable.
I interviewed 7 field operators — including a 40-year veteran — and 7 SCADA IT experts. The mental model of Go/No-Go was already correct in operators' heads. It was the tools underneath it that were breaking.
The insight that reframed the whole project: operators didn't need more data — they needed reachable data. Three tasks mattered above everything else: when to escalate a level to a supervisor, how to escalate a high-risk level, and how to find available hardware like boosters or a second catcher.
I collapsed raw SCADA sensor streams into a single live percentage, with two alarm thresholds baked in: High at 60% capacity, High-High at 70%. Everything an operator needed to launch safely lived in that one gauge.
I designed the gauge once and mirrored the same data across every role that touches a pig launch — nobody works from a different number.