Driving Quality Forward - A Unified Platform for Factory Insights
Designed for factories, built for speed — track metrics, manage updates, and align teams across sites, all from one powerful interface.
Quick Snapshot
Here’s the project in a nutshell: what we aimed for, how we got there, and what came out of it — all at a glance
Client :
Japanese car manufacturing company
Goal :
Build a modern, user-friendly system from the ground up to replace a fragmented legacy platform and improve operational efficiency nationwide.
Project Type :
New internal platform development covering factory performance dashboards, user management, data input, bulletin posting, and parts backtracking.
The Chaos We Navigated
Started without direct user access — relied on second-hand feedback to pitch the proposal
Relied on legacy system as reference and indirect stakeholder feedback
Tight 2-week sprints with shifting developer feasibility (some features went from "yes" to "no" mid-sprint
Language barriers caused delayed or second-hand communication
What Users Were Dealing With
Monthly-only data updates made quick decisions nearly impossible
Clunky legacy system scattered data everywhere — nothing was centralized
No clear audit trail or tracking for part backtracking workflows
Lacking of intuitive interfaces made daily use frustrating for workshop engineers and factory managers
How I Tackled It
Secured buy-in with a high-fidelity concept even before user interviews were allowed
Established scalable design system components upfront
Delivered comprehensive feature screens covering dashboards, user management, data input, bulletin board, and part backtracking.
Adapted fast across 9 sprints, translating shifting client needs into solid UX
Where It Ended Up
Stakeholders said they were "Fans" of the new design (their words!)
Engineers looked forward to ditching the old system
Client changed vendors mid-way to better match our UX vision — designs are still being used today
"Despite the challenges, this project showed me how to create clarity and alignment through design — even in complex, high-stakes environment"
Curious how it all came together? Let’s break it down step by step.
1. The Big Picture - Setting the Scene
Every story needs a backdrop. Here's who we worked with, why this project mattered, and who was on the team.
Redesigning Efficiency for a Japanese Car Manufacturer
We partnered with a leading Japanese Car Manufacturing company, they are a global leader in automotive innovation, to design a desktop-based internal platform from the ground up—focused on streamlining KPI tracking, user management, data input, and bulletin postings across multiple factories. While a legacy system existed, it served only as a rough reference.
Working within a fast-paced Agile framework with 2-week sprints, I led the end-to-end UX process and delivered a fully client-approved design package, collaborating closely with developers, product managers, and stakeholders across the 6-month timeline.
Duration :
6+ months
Platform :
Desktop web application
Process :
Agile (2-week sprints)
Team :
UX/Product Designer (me) | Developers | PMs | Client Stakeholders
2. Behind the Legacy - Why This Project Was Needed
Before our involvement, the client’s factory teams were relying on a legacy system that had not seen major updates in years. While functional, it created more friction than value — especially for engineers and managers juggling multiple roles across factory operations. Through internal documents and feedback from management, we identified key pain points that shaped our redesign strategy:
Key Pain Points
Outdated System with Poor Usability
The interface was unintuitive, requiring users to reference multiple external documents just to complete a single task.
Data Only Updated Monthly
Users had no access to real-time insights. Production issues couldn't be spotted early, making proactive responses difficult.
Fragmented Data Entry and Review Process
Operational data was scattered across multiple Excel sheets, with graphs either manually generated within Excel or selectively uploaded into internal systems. This lack of a centralized source created inconsistencies and made analysis time-consuming and error-prone.
3. The Proposal That Kicked It All Off
With no direct access to users or stakeholders, I crafted the initial proposal based solely on legacy screenshots, second-hand notes, and initial brief from the client.
The goal? To prove how a cleaner, more modular UX could replace the cluttered, outdated system.
I built a high-fidelity concept that reimagined key features like KPI dashboards and data input flows — focusing on clarity, simplicity, and usability.
Modular
A modular design allows easy addition, removal or modification to features without disrupting the entire system.
Simple
Remove the clutter, and help the users easily complete their tasks by optimizing the processes.
Flexible
Support diverse user needs, this allows the platform to adapt to different workflows and preferences.
The proposal struck a chord with stakeholders. It didn’t just get approved — it set the project in motion.
Early Low Fidelity
The Proposal That Set the Standard
To make sense of complex data without overwhelming the users, I combined grid-based, tabbed, card, and tile layouts into a unified design.
A Hybrid Layout Built for Clarity and Control
This hybrid approach brought modular structure and visual clarity — allowing users to:
Instantly spot high-priority KPIs
Effortlessly switch between datasets
Dive deeper into trends with detailed drilldowns
By blending familiar UI patterns with purpose-driven hierarchy, the final design made navigating factory data feel simple, fast, and intuitive — even for first time users.
The client responded positively to the proposed design and approved it quickly. With their support, we moved on to user interviews to validate our ideas and refine the design further.
4. Sprint 0 - Building the Foundation of Scalability
With full access to users and stakeholders, Sprint was our chance to get things right from the ground up. We rolled up our sleeves and dove into research, aligning user needs with business goals, and shaping a design direction that could scale across factories and future use cases.
This phase wasn't just about planning — it was about translating insights into action. Every persona, flow, and requirement we crafted served as a building block for a platform designed to simplify complexity and empower users on the ground.
Groundwork in Action
User Personas
Real people, real needs. We crafted personas based on interviews with engineers, team leads, and factory managers — highlighting their pain points, motivations, and how they interact with the system daily.
Step 1: Defining user groups
Mapped out every key user across the manufacturing process, organizing them by department to ensure focused and meaningful interviews.
Step 2: Question banks
Developed role-specific question banks to guide interview, helping us surface real user needs even before the feature set was fully locked in.
Step 3: Defining User Persona
Turned interview insights and department structures into detailed user personas that grounded our designs in actual workflows and goals.
Feature & User Flows
From dashboards to data entry, we mapped the critical paths users take across the platform after defining the features to be developed with the client. These flows shaped how we structured screens and prioritized functionality.
Feature Requirements
Working closely with stakeholders, we translated feedback and business goals into a clearly defined feature list — giving us a focused, validated scope to move forward with.
Design System
With the feature set finally locked in, I began building a robust design system to ensure consistency, scalability, and visual clarity across all screens and components.
Key Actions :
Conducted stakeholder interviews and user research to uncover core points and expectations
Define feature requirements based on business goals and user needs
Mapped out key workflows for major features (KPI dashboard, user access, data input, bulletins)
Outline feature flows to visualize end-to-end interactions
Established a design system to ensure consistency across screens and components
Created high-fidelity mockups for all key feature, setting a clear direction for development
5. Building in Motion - What Happened During the 9 Sprints
While the designs were fully approved at these start, what followed across the 9 development sprints wasn't a smooth rollout — it was a live balancing act of shifting feedback, technical blockers, and scope pivots. Here's how the journey unfolded, phase by phase:
The calm before the storm.
Development started strong with clear handovers and direction.
Business Analyst insisted on starting with the Data upload function even though I recommended starting with the dashboard and the data visualization features.
Although the dashboard was initially prioritized, the business analyst redirected focus towards the data input page — specifically the upload flow and pop up validation — for early development.
Small design tweaks started creeping in: alignment issues, misunderstood logic, and feature expectations that weren't surfaced earlier.
Took initiative to dig into the existing dashboard experience — spotting gaps, untangling friction points, and uncovering fresh opportunities to elevate the UX.
Version 1
The dashboard originally merged live and historical data on one page, but there would overlapping filters needed which would have made it confusing. After a few iterations, I split both of them into two tabs to make it a cleaner, more intuitive experience.
Version 2
In the second version, the client also requested a live data of the incident free days within the factory.
Version 3
After creating Version 2, I noticed the incident tracker was taking up more space than necessary, leading to layout inefficiencies. Around the same time, the client informed us of two separate production lines and requested that live production data be split accordingly — prompting a layout redesign to better optimize space and clarity.
Version 4
In Version 4, the business analyst shared that one of the KPI displays was no longer needed by the client, so it was removed from the dashboard. I also used this opportunity to rethink the live production display — drawing inspiration from audio buses in digital audio workstations to create a more structured, scannable layout.
Version 5
In the final iteration, the PM shared that the client didn’t have images for each factory workshop — prompting a shift away from visual-based layouts. I also learned that the live production data reflected overall performance, not per workshop, so I adjusted the design to surface key workshop-specific insights more clearly at the top level.
"The screens were approved, but some details only became 'wrong' once they were built."
User exposure reveals misalignments
Started implementation of daily KPI dashboard and data visualization features.
Once client-side testers saw the working version, feedbacks picked up rapidly.
Technical challenges surfaced mid-development that weren't clear at the start, prompting fast design tweaks to keep pace with the sprint deadlines.
"Feasibility feedback sometimes come just days before the sprint end, forcing me to make design compromises fast."
Rework, clarification, and chasing alignment.
Several features had to be partially redesigned — not because the UX was wrong, but because tech constraints or client made change request.
Further addition of data visualization features.
We relied on async feedback loops, but second-hand communication often led to delays or misinterpretations.
Last-minute changes and design survival mode.
Final features like user management and bulletin posting were built, but not without hiccups.
Bugs, edge cases, and "one more change" requests stacked up as the finish line approached.
Some UI polish had to be deprioritized in favor of functionality and delivery speed.
"At this point, the goal shifted from 'perfect' to 'stable and usable enough to test in production."
What It All Led To
After weeks of iteration, collaboration, and adapting to real-world constraints, the final design brough together all the moving parts — clarity, usability, and flexibility.
The redesigned platform delivers a unified experience for internal factory users, enabling them to:
Monitor both live and historical production KPIs with clarity
Input, validate, and correct data more efficiently
Manage user access and roles across multiple sites
Publish and receive bulletins to stay aligned across departments
Every screen was built using a modular design system to ensure future scalability and consistency. The result: an experience that feels less like a system to fight with — and more like a tool teams actually want to use.
7. Roadblocks Along the Way - Constraints We Worked Around
No project is without its bumps. Here what made the journey messy — and what we learned from it:
Misalignment from Second-Hand Feedback
Communication with the client was often filtered through intermediaries, leading to delayed or unclear feedback. This made it challenging to validate design decisions quickly or clarify expectations early.
Feasibility Shifts Mid-Sprint
Some features initially marked as "doable" turned out to be complex during implementation. Dev constraints surfaced late, forcing rushed design adjustments under tight sprint timelines.
Language & Cultural Barriers
Working with a Japanese client meant navigating language differences and cultural communication styles, which sometimes led to misunderstandings or slower feedback loops.
Shifting Roles & Scope Creep
With most communication going through a business analyst, it became unclear if design intentions were accurately conveyed to the client. Feedback often came back vague or altered, and last-minute request to add or remove features became frequent. At times, the project felt less like a collaborative design process — and more like executing a vision that kept changing
"It reached a point where I wasn't sure if I was designing for the user — or redesigning for the client."
Tight Timelines, No Room to Breathe
Sprint cycles left little buffer for iteration. Any feedback or blocker that arrived late meant immediate prioritization, design trade-offs, and minimal breathing room for refinement.
Fragmented Data, Limited Testing
With operational data only updating monthly, we couldn't test live scenarios in real time. This meant designing based on assumptions and internal feedback rather than ongoing user input.
"I wasn't just designing screens — I am constantly adjusting, communicating, and navigating the unknown."
8. Outcome & Reflections
This project wasn't a straight line — it zigzagged through shifting requests, second-hand feedback, and real-world constraints. But despite the chaos, the impact was clear.
The client openly expressed that they were fans of the design work. Even though development is still in progress, the feedback we've received has been validating — not just of the design itself, but the process behind it.
More importantly, this experience reminded me that good design isn't just about polished screens. It's about navigating uncertainty, aligning moving parts, and staying grounded in the user's reality — even when that reality keeps shifting.
9. Key Takeways
Early approval ≠ fixed scope
Designs will evolve — sometimes drastically — even after sign-off. Being open to change is just as important as being confident in my direction.
Second-hand communication dilutes intent
When working through intermediaries, clear documentation and annotated designs become essential tools for preserving my thinking.
Scope creep needs active boundaries
Without clear negotiation or product ownership, design risks becoming reactive. It's critical to ask "why" and reframe changes in terms of user value.
Designing with imperfect data is part of the job
Even with outdated systems and limited testing windows, it's possible to create meaningful improvements by designing flexibly and iterating forward.
Real-world UX is rarely neat — and that's ok
Messy processes, shifting feedback, and unclear requirements are normal. What matters is how I adapt, communicate, and keep advocating for the user.
