iBASEt Solumina MES: a design practice built from zero, and manufacturing taken off the desktop
Founding designer and first UX hire at iBASEt. I built the UX practice, the team, and the design system from nothing, and took a dense legacy aerospace MES mobile-first onto the shop floor.
The context
iBASEt's Solumina is a Manufacturing Execution System for the most demanding builders in regulated industries: aerospace and defense, nuclear, medical devices. Programs running on it include work for NASA, Rolls-Royce, Cirrus, Lockheed Martin, and Northrop Grumman. The software was powerful and the brand was trusted, but the interface was a 1990s Windows client: 50 or more elements per screen, menus nested deep, thin hierarchy, and a learning curve that needed extensive training. It was also desktop-bound, one step removed from the floor where the build actually happens.
I joined as the company's first designer. There was no design practice, no team, and no system.
What I owned
The whole of UX, starting from nothing. I built the practice, the team, and the design system, and I owned the design decisions end to end.
- I built and led the UX team. I hired a designer straight out of college, on site, and mentored them up to Senior UX Designer, and brought on a second designer in India.
- When I arrived, the product team had no working practice around user stories, and the effort had stalled at the login screen with no agreed path past it. I found a product manager inside the company, partnered with them and mentored them on writing user stories, and together we defined the problems that were actually worth solving.
- I partnered with sales to reach real users: getting onto the shop floors of companies running Solumina, and of companies still doing everything on paper, and spending the annual Solumina user events talking with technicians one on one about where their workflow slowed down and how to speed it up.
- Engineering resisted from the start, pushing back on even implementing the designs. Getting them on board was part of the role, and I worked with them directly until the approach earned its place.
- I built the design system itself: Axure prototypes to pressure-test ideas before any production code, then a tokenized React and Material UI system with a swappable brand layer, so each customer could be given their own theme.
The decisions
Three calls shaped the product.
Density to clarity. The first job was subtraction. Dense data tables became scannable, prioritized cards; a persistent global navigation and contextual side panels replaced nested dropdowns; work screens were rebuilt around clear typography and collapsible sections. Screens that once held 50 or more elements now carry 15 to 20.
Off the desktop, onto the floor. The bigger bet was that the product belonged where the work is. I designed everything mobile-first and scaled it up to tablet and desktop, rather than shrinking a desktop tool down. The base layout assumes a phone in one hand: thumb-reachable navigation, a prominent scan action, full-width cards, 44px targets. The wearable, showing the next operation and a one-tap start, was the smallest surface and the strongest proof of the thesis.
Scan-first for gloved hands. A technician mid-build has hands busy or gloved, works in variable light, and needs an answer in seconds. Before this, they were tied to fixed terminals, walking back and forth across the floor to look up a part or log a defect, relying on memory or scribbled notes in between. So I built the floor experience around one fast loop: scan a part, see its details and history, take the next action (log a discrepancy with a photo, or move on), with a keyboard-operable manual-entry fallback so the flow never depends on the camera. A dark, high-contrast interface suits the floor and large controls survive gloves. This scan-first workflow is where the clearest win came.
What shipped
The result reads consistently from a workstation down to a phone and a wearable: a glanceable dashboard across active programs, a guided step-by-step work instruction, an as-built traceability tree, and the scan entry point that ties it together.
The screens here are wireframe recreations I rebuilt from memory for this portfolio, because the production system runs inside customer environments. The structure, interactions, and design language are what I shipped; the data is fictional.
Accessibility
Held to WCAG AA throughout. Status never relies on color alone; every state pairs a colored indicator with a text label, and text meets contrast minimums. Controls are keyboard reachable with visible focus, the as-built component tree uses a full ARIA treeview with arrow-key navigation, dialogs manage focus, and errors announce through live regions. Camera scanning is paired with a keyboard-operable manual-entry path, so the work is completable without the camera.
The outcome
Lower training burden for new technicians and clearer task focus, reported by shop-floor users.
"Mike built a small UI/UX team and led it successfully from initiation through release. He puts a lot of thought into design decisions, listens carefully to feedback, and collaborates well with developers and PMs."Bob Joyce · VP of Product Development, iBASEt
What I'd carry forward
Modernizing legacy software is mostly an act of subtraction. The power was already in the product; the job was to reveal it and put it where the work happens. Card-based structure, a consistent navigation model, and a real mobile-first stance did more for adoption than any new feature could have. The other half of the job was organizational: building the practice from nothing and bringing a resistant engineering team along was as much of the work as the design.