UX Redesign for
Field Operators
Redesigning a live operator workflow inside Murphy's Integrated Operations Platform (IOP), for people working on-site, often in gloves, heat, and glare.
Redesigning a live operator workflow inside Murphy's Integrated Operations Platform (IOP), for people working on-site, often in gloves, heat, and glare.
I worked on this project during my summer internship at Murphy Oil Corporation!
Problem: Field operators needed extensive training just to navigate the task management app, and completing common tasks required too many steps, slowing down operational work in the field.
Solution: Redesigned the operator mobile experience around operator priorities: bigger touch targets, clearer hierarchy, fewer steps to common actions, and always-visible workflow status.
Impact: Faster task completion, reduced training burden, fewer errors/rework, and higher operator confidence, validated through stakeholder approval of the proposed workflow enhancements.
Murphy's Integrated Operations Platform (IOP) bridges the Remote Operations Center (ROC) and field operators, making sure tasks get dispatched and tracked reliably from creation to completion. Part of the platform, PAWS (Process Automation Workflow Solution), digitizes forms and automates approvals to streamline these operational processes.
In practice, the existing tool created friction instead of removing it:
Methods used:
Key insights, mapped against industry standards:
| UX Category | Industry Standard | What Operators Were Actually Experiencing |
|---|---|---|
| Information Hierarchy | Critical info identifiable in <5 seconds (NNG); focus order preserves meaning (WCAG 2.4.3) | Important actions were mixed in with secondary options |
| Cognitive Load | Limit competing information, emphasize primary actions (WCAG 3.3.2) | Duplicate and low-value information cluttered the screen |
| Accessibility | Touch targets ≥24×24px, 44×44px recommended for frequent actions (WCAG 2.5.8) | Controls were too small for gloved operators to use reliably |
| Navigation Efficiency | Common workflows take ≤3 interactions to begin (WCAG 2.4.5) | Extra clicks were required to reach common actions |
| System Feedback | Users always understand workflow progress and next steps (WCAG 4.1.3) | Limited visibility into workflow status |
Problem statement: How might we redesign the operator mobile experience so that field operators, including those working in gloved or hands-limited conditions, can complete common tasks quickly, without requiring training to navigate the app?
Primary user: Field operator using the app on-site, often in physically demanding conditions, needing to complete and update tasks dispatched from the ROC.
Success metrics: Reduction in average task completion time, reduction in training time/onboarding sessions needed, reduction in support tickets or reported errors.
| Option | Pros | Cons | Outcome |
|---|---|---|---|
| Continue refining admin console | Already in progress | Didn't address the friction operators actually felt in the field | Cut: refocused on operator mobile experience |
| Operator mobile experience redesign | Directly addresses field-reported pain points; higher day-to-day impact | Required re-scoping project focus | Chosen |
Each round moved from a first attempt to a revised, working solution.
Issue
Gloved operators struggled to accurately hit sort/filter icons , and sun glare made on-screen text hard to read outdoors.
First attempt
Enlarged both buttons and text to fix glove accuracy and glare readability at once.
Feedback
Larger text left only 2 task cards visible at a time , and the fix undid the homepage's core function, so it was judged a net hindrance.
Revised approach
Kept the larger buttons , since that solved glove accuracy without a space cost, and rolled font size back down , since text was the real driver of the space problem; glare/vision needs can be handled in device settings instead.
Why this worked: the two problems didn't need the same fix: button size is a fixed UI cost, while text legibility varies by user and already has a system-level solution. Scaling everything up fixed one issue by creating another; separating them let each be solved without the other's tradeoff.
Outcome: Button accessibility increased 16.13% (tap target size), with no loss to task visibility.
Issue
Each task card took up too much vertical space , limiting how many tasks were visible at once.
First attempt
Rearranged existing data within each card to be more compact.
Feedback
Cards got smaller, but stayed data-heavy and hard to scan ; reorganizing addressed the symptom, not the cause.
Revised approach
Confirmed with stakeholders what data operators actually need, then cut fields that weren't functionally necessary , like who dispatched the task and the well's tier.
Why this worked: the real issue was informational, not layout: cards carried data that didn't serve the operator's task-completion goal. Cutting that data directly shrank each card's footprint, which reorganizing alone couldn't achieve.
Outcome: Removing 20% of card metadata drove a 46.67% increase in task visibility.
Issue
Completing one task required 3 separate pop-ups : done → "are you sure?" → "task submitted", for what is functionally one action.
First attempt
Added a persistent progress bar plus a review page before submission, meant to replace the pop-up confirmations.
Feedback
Progress bar was a clear win, kept. The review page failed: operators are task-completion driven, not review-driven , and skimmed past it regardless.
Revised approach
Replaced "Done" and its two pop-ups with a slide-to-submit gesture on the same page, collapsing 3 interruptions into 1 deliberate action.
Why this worked: the review page tried to add a behavior that didn't match how operators actually use the app. Slide-to-submit preserves the pop-ups' original goal (preventing accidental submission) through deliberate physical intent, without a step users would just route around.
Outcome: 3 separate interruptions collapsed into 1 deliberate gesture.
Improvements implemented, mapped to the standards that drove them:
16.13%
bigger touch targets
46.67%
more visible tasks
3 → 1
submit interruptions
This project also doubled as a chance to test where AI tools actually help in a design process, and where they don't.
Figma's new in-file AI
Working on Figma Professional gave me access to Figma's beta, in-file AI as soon as it rolled out, so this project became a live test of what it's actually good for. The pattern that emerged: it's not creative, but it's strong at small, repetitive, menial tasks.
I used it to generate the page organizing sort-card options (lowest to highest, newest to oldest, and so on), a task with a clear, mechanical structure and no real ambiguity to resolve.
It fell short, though, when a task needed critical thought rather than pattern-following: for example, it couldn't come up with good ways to organize the details within a task card. That kind of decision depends on understanding what operators actually need to see first, which isn't something it could reason through on its own.
Copilot for design reviews
Before presenting to stakeholders, I used Copilot as a second set of eyes on the design, specifically to catch small details I may have missed. For example, flagging if a new design had dropped a function that the original page had, something easy to lose track of when a redesign strips a screen down.
Claude for content and presentation logic
I used Claude to help populate task cards with realistic fake details, so mockups felt like real operator tasks rather than obvious placeholders.
I also used it to help put together presentations for stakeholders, not the slides themselves, but the logic of how to sequence and frame the story I wanted to tell.
