AI-Powered Troubleshooting Portal for Industrial Controller
1. Overview
Role: Senior UX Designer (Lead)Domain: Industrial UX + AI Systems (Semiconductor Manufacturing)
Users: Fab Technicians, Remote Service Engineers, SMEs, Managers
Deliverables: Research → Personas → Affinity Map → IA → Flows → Wireframes → UI → Prototype → Testing
Duration: Approx. 6 months
Goal: Reduce troubleshooting time and empower technicians to resolve eligible issues quickly using AI + a safe Quick Fix action.
3. UX Research
3.1 Research Activities
I executed a 4-stage research plan:
1. Stakeholder Interviews
- Engineering managers
- Product owners
- AI/ML team
- SMEs
- Process engineers
2. User Interviews
- 12 remote service engineers
- 6 fab technicians
- 3 maintenance managers
3. Contextual Inquiry
Observed:
- Controller log interpretation
- Parameter cross-checking
- Trend analysis
- Ticket lifecycle
4. Artifact Review
- 200+ sample log files
- Error code tables
- SOPs & maintenance manuals
- Past service requests
User Personas
User Personas
User Personas
Persona 1 — Remote Service Engineer “Daniel”
- Age: 32
- Experience: 4 years
- Motivation: Resolve tickets fast, reduce dependency on SMEs
- Pain Points:
- Logs take too long to analyze
- Too many repeated low-severity issues
- AI predictions lack clarity
- Needs:
- Transparency in AI reasoning
- Structured troubleshooting steps
- Better visualization of logs
User Personas
User Personas
User Personas
Persona 2 — Fab Technician “Mei”
- Age: 26
- Experience: 1–3 years
- Motivation: Keep the tool running and prevent downtime
- Pain Points:
- Must wait for engineer response
- No visibility into progress
- Simple issues take too long to fix
- Needs:
- Clear status updates
- A safe “Quick Fix” option
- Easy-to-follow guidance
Empathy Maps
Empathy Maps
Empathy Maps
Engineer Empathy Map
- Thinks: “I need reliable predictions; guessing wastes time.”
- Sees: Large logs, inconsistent ticket notes.
- Does: Reads logs line-by-line; compares with history.
- Feels: Overloaded with repeated small issues.
Empathy Maps
Empathy Maps
Empathy Maps
Technician Empathy Map
- Thinks: “I wish common issues were instantly fixable.”
- Sees: Downtime pressure
- Does: Files ticket, waits, follows steps
- Feels: Frustration when issues are repetitive
6. Affinity Mapping
I synthesized all research into 4 major themes:
Theme 1 — Trust in AI
Users need explainable predictions.
Theme 2 — Standardization Gap
Engineers follow different troubleshooting processes.
Theme 3 — Communication Gap
Technicians lack visibility into engineer actions.
Theme 4 — Repeated Low-Risk Problems
27% of tickets were solvable with known, simple actions → opportunity for Quick Fix.
8. Ideation & Concept Development
I conducted structured ideation using:
1. Crazy 8s
Generated 8 variations for:
- AI Insights
- Quick Fix UI
- Error state handling
- Step guidance
2. “How Might We…”
- HMW automate simple fixes?
- HMW make AI explainable?
- HMW standardize troubleshooting?
- HMW support technicians without risk?
3. Co-Design Sessions (with SMEs)
Reviewed:
- Which issues can be safely automated
- Which actions technicians can perform
- Required safety validations
10. Information Architecture
Technician Portal
- Dashboard
- Create Ticket
- Quick Fix (Eligible Issues Only)
- Step-by-step Guidance
- Ticket Timeline
- Resolution Summary
Engineer Portal
- Ticket Workspace
- AI Insights
- Controller Health Snapshot
- Troubleshooting Flows
- Knowledge Base
- Past Resolutions
User Flows & Task Flows
User Flows & Task Flows
User Flows & Task Flows
Technician Flow
- Notices tool issue
- Opens portal
- Creates service request
- Portal evaluates eligibility for Quick Fix
- If eligible → Quick Fix shown
- Technician executes Quick Fix
- System verifies resolution
- Ticket resolves automatically or escalates
User Flows & Task Flows
User Flows & Task Flows
User Flows & Task Flows
Engineer Flow
- Ticket assigned
- Upload log → AI parses patterns
- Review root cause suggestions
- Examine explainability panel
- Enter guided troubleshooting flow
- Complete steps
- Add resolution summary
- Ticket closes
12. Wireframes
My Background
My Background
My Background
Low-Fi Wireframes
- Created rough sketches for Quick Fix screen
- Log insight module
- Root-cause dashboards
- Ticket overview.
My Skills
My Background
My Background
Mid-Fi
- Data density planning
- Hierarchy planning
- Navigation model
- Step-by-step flow layouts
My Projects
My Background
My Projects
High-Fi
- Industrial dark mode + light mode
- Final components
- Visual anomaly graphs
- Interlock warnings
- Progress states
13. Prototypes
I created interactive prototypes in Figma for:
- AI Insight Dashboard
- Quick Fix Flow
- Troubleshooting Stepper
- Ticket Timeline
- Controller Health Summary
Included:
- Interaction states
- Edge cases
- Error screens
- Success/failure validation
14. Usability Testing
Participants
- 8 engineers
- 5 technicians
Tasks Tested
- Using Quick Fix
- Understanding AI suggestions
- Navigating troubleshooting flows
- Reading anomaly graphs
- Completing resolution steps
Outcomes
- Quick Fix highly appreciated
- AI became “trustworthy” only after adding explainability
- Technicians needed simplified language
- Engineers liked dense data view
Iterations Based on Findings
✔ Added rationale under Quick Fix
✔ Improved trend graph clarity
✔ Added technician-friendly vocabulary
✔ Used collapsible sections to control density
15. Impact
Quantitative
- 27% of tickets resolved using Quick Fix
- 42% reduction in troubleshooting time
- 30% reduction in manual log analysis
- 2× faster onboarding for new engineers
Qualitative
- Engineers felt empowered by structured flows
- Technicians gained autonomy
- SMEs approved knowledge retention from system
16. Final Deliverables
- Personas
- Affinity Map
- Empathy Map
- Journey Map
- Information Architecture
- User Flows
- Task Flows
- Wireframes (low-mid-high)
- UI Design
- Clickable prototype
- Design documentation for dev teams