2024
Siemens Healthineers
siemens-healthineers.com
The Challenge: Complex Lab Errors, High Costs
In high-throughput laboratories, every minute counts.
Technicians working with the Atellica® Solution — a sophisticated immunoassay and clinical chemistry system — often faced assay errors that were difficult to diagnose.
Troubleshooting was a nightmare:
Error logs were scattered across subsystems.
Many technicians had low IT literacy.
Downtime meant delayed patient results and increased costs.
Whenever a problem occurred, labs had to escalate issues to expert engineers, leading to expensive site visits and hours (sometimes days) of lost productivity.
Our Mission
Design a web-based assistant that simplifies troubleshooting, empowers technicians globally, and reduces unnecessary escalations.
My Role
I was the UX Designer for the Atellica Troubleshooting Assistant. I led research, designed workflows and prototypes, and applied the SHUI Design System for consistency.
I collaborated with requirement engineers, PMs, and lab teams across global markets to ensure the product was intuitive, technically feasible, and aligned with real lab needs.
Target Audience & Stakeholders
Primary Users:
Lab Technicians: Perform day-to-day troubleshooting; often low IT literacy
Lab Managers: Monitor overall system health, track recurring issues, and oversee resolution success
Secondary Stakeholders:
Field Engineers & Experts: Receive escalated issues for deeper analysis
Business Leadership: Interested in reducing downtime, service costs, and improving efficiency
Stakeholder Goals:
Enable technicians to self-resolve issues
Reduce unnecessary escalations
Provide actionable insights for lab managers and engineers
Research & Planning
Methods:
Contextual inquiry with field service engineers
Interviews with lab managers and support leadership
Analysis of historical error logs and resolution records
Key Findings:
Engineers struggled to interpret raw error logs remotely
Lack of structured workflows caused repeated escalations
Global differences in lab setups and regulations required flexible yet consistent solutions
Planning Outcome:
Defined guided flows for troubleshooting based on error types and severity
Prioritized dashboard metrics: recurring issues, resolution success, escalation history
Designed system to integrate seamlessly with LIS
User Persona
Design Process
Step 1: Ideation & User Flows
Mapped current troubleshooting workflow → identified bottlenecks
Designed dual-path flows: guided for juniors, direct access for seniors
Created wireframes for dashboard, step-by-step troubleshooting, and escalation reports
Step 2: Prototyping & Iteration
Tested low-fidelity wireframes with engineers for usability
Iterated high-fidelity prototypes based on feedback
Refined metrics visualization and troubleshooting guidance
Step 3: Final UI Design
Clear, intuitive dashboards with trend charts and severity indicators
Step-by-step guided troubleshooting panels
Automated escalation reports compatible with LIS
Application of SHUI Design System
Used SHUI (Siemens Healthineers UI) for enterprise-grade consistency
Standardized components, typography, and accessibility patterns
Integrated with LIS workflows for seamless adoption
Benefits
Consistent UI across dashboards, troubleshooting steps, and escalation flows
Faster development via reusable components
Improved user confidence through familiar patterns
Design Showcase
Outcome & Impact
Results
25% reduction in Mean Time to Resolution (MTTR)
15% fewer escalations to expert engineers
Improved confidence among technicians with low IT literacy
Standardized troubleshooting workflow across global labs
Reduced unnecessary site visits → cost and time saving
Reflections & Learnings
Key Takeaways
Simplifying complex data is critical in high-stakes environments
Dual-path flows effectively balanced novice and expert needs
Global usability testing was essential for adoption
Accessibility and visual clarity built user trust
Future Opportunities
Integrate AI-based predictive troubleshooting
Introduce smarter escalation recommendations and auto-suggestions











