Five engineers across four customer companies: 禾鏵實業 (TSMC-facing load-port maker), 羅昇 FAE (Delta's distribution / after-sales partner), 曜鋅科技 (PLC + CIM / SECS system integrator), and 銓發科技 (PC + PLC integrator who'd given up halfway through self-learning).

Mixed methods. NPS for loyalty, SUS for usability, and semi-structured interviews for the “why.” Baseline cohort of 22 from a Delta training event; same instruments re-run on 17 in the pre-test after the first redesign iteration.

Running this depth of UX research inside a hardware company that ships PLCs for a living. The instincts of the organisation were “just add features.” My job was to make qualitative pain measurable enough that engineering and PM leadership couldn't dismiss it.

One quantified pain point in a stakeholder review changes more minds than ten hours of design critique. Pairing the SUS number (53, with six failures) with the direct quotes from named customers turned an abstract “the UI is hard” into a defensible budget ask. That's the lesson I take into every future project.

Manufacturing Operations Management (MOM) is the operational backbone of modern factories — orchestrating work orders, inventory, equipment status, and quality data across every site, line, and shift. It's the system everyone on the floor depends on, every minute of every day. [Placeholder — replace with your definition]

A single MOM platform serves two radically different contexts: web desktop for managers and engineers analyzing data, and operator tablets and PDAs for line workers executing tasks in noisy, gloved, time-pressured environments. The same data, two completely different experiences. [Placeholder — adjust to your framing]

[Placeholder] Briefly describe how you approached research for MOM — site visits, shadowing, contextual inquiry on the factory floor, etc. Mention any unique constraints (PPE, shift schedules, noise, gloves).

[Placeholder] List the user groups — line operators, shift leaders, production engineers, plant managers — and the questions you were trying to answer for each.

[Placeholder] The concrete outcomes — what users could now do that they couldn't before. Try to include a quantitative metric if possible (training time reduced, error rate down, adoption up, etc).

[Placeholder] The longer-arc effects — design system adoption, replication across plants, influence on subsequent products, or what you'd do differently next time.

[Placeholder] Most enterprise design systems are built for desktop, by people who work at desktops. But the factory floor is a different world — gloves, glare, noise, time pressure, often standing. A design system built for that environment has to start from a different set of first principles.

[Placeholder] Larger touch targets, higher contrast, fewer states per screen, audio/haptic feedback patterns, and every component validated against gloved-hand input. Not a re-skin of the corporate system — a separate system, with its own tokens, components, and documentation.

[Placeholder] Describe how you ran the research that underpins the system — site visits, gloved-hand testing, measuring tap accuracy under realistic conditions, observing actual workflows on the floor.

[Placeholder] Specific findings — e.g. operators missed targets smaller than X mm 30% of the time when gloved, or screen glare made certain color combinations unreadable in parts of the line. Numbers make this section credible.

[Placeholder] How many teams, products, or screens now use the system. The "design once, ship many" outcome.

[Placeholder] Personal takeaways — the gap between "design system that exists" and "design system that's used", or the importance of grounding every token in observed behavior.