UI Design

UX Design

EFACT AND MORE

Designing Enterprise Software for Vehicle Fault Analysis

TIMELINE

February 2026 – Now

ROLE

UX/UI Designer (Intern & Working Student): User Research, Concept Design, Wireframing & Prototyping

TEAM

TI-223

Context

Internal Diagnostic & Vehicle Fault Analysis Software

To protect proprietary information, all designs use synthetic data. If you’d like to see an interactive walkthrough or talk through my design process, please drop me a message.

Long Story short

I streamlined complex vehicle fault analysis into a unified digital workspace for BMW Group specialists.

I worked on internal software tools at BMW Group, focus area: rebuilding and evolving EFACT – the central platform used by specialists to navigate fragmented diagnostic tools, tickets, and vehicle data. Designing for expert power-users taught me to handle high information density and bridge disparate systems into a cohesive, intuitive workflow under real-time production pressure.


Starting with feature improvements on the legacy system, I conducted user interviews with analysis specialists to unveil core UX bottlenecks. I then co-facilitated team workshops to realign the product vision, ultimately designing a completely new web application—including the seamless integration of a standalone sub-app I created.

I worked on internal software tools at BMW Group, focus area: rebuilding and evolving EFACT – the central platform used by specialists to navigate fragmented diagnostic tools, tickets, and vehicle data. Designing for expert power-users taught me to handle high information density and bridge disparate systems into a cohesive, intuitive workflow under real-time production pressure.


Starting with feature improvements on the legacy system, I conducted user interviews with analysis specialists to unveil core UX bottlenecks. I then co-facilitated team workshops to realign the product vision, ultimately designing a completely new web application—including the seamless integration of a standalone sub-app I created.

Phase 1

Working within Legacy Constraints & Continuous Feature Design

For several months, my work focused on evolving the legacy version of EFACT. Rather than designing in isolation, I worked in a continuous feedback loop with key users—our vehicle analysis specialists—during regular interviews and weekly refinement sessions.


  • Gathering Requirements: In these sessions, specialists shared pain points, feature requests, and workflow bottlenecks based on their daily plant operations.


  • Iterative Concepting: I translated their technical needs into conceptual UI proposals, presenting early wireframes and interaction concepts back to the users for validation.


  • Handoff to Development: Once a feature design was refined and approved, I collaborated closely with my team to prepare clean assets and specifications before handing them over to the developers for implementation.


This foundational period gave me a deep understanding of the domain, but it also exposed the structural limits of the legacy tool—paving the way for the major redesign that followed.

Creating a complex filter system over all systems

Feature left: create column groups, feature right: creating vehicle groups and showing last viewed vehicles

Phase 2

Uncovering Bottlenecks through Deep User Research

As features were continuously added, it became clear that incremental updates were treating symptoms rather than the root cause. To understand the bigger picture, I conducted structured user interviews and observational sessions with vehicle analysis specialists.


My goal was to map out their entire diagnostic journey, identifying friction points and structural limits within the existing system landscape.

Key Findings:

  • Fragmented Tool Ecosystem: Specialists spent critical time constantly context-switching across 5+ independent databases, analysis tools, and ticket systems.

  • Cognitive Overload & Redundancy: High information density without clear hierarchy forced users to manually correlate vehicle data from different sources.

  • The Case for a Paradigm Shift: The interviews revealed that minor UI tweaks in EFACT wouldn't suffice. The specialists needed a unified digital workspace built around their actual workflow.

This research served as the pivotal foundation to stop patching the legacy system and advocate for a ground-up redesign and rebranding.

Research into workflows and dynamic interfaces

Phase 3

Conceptualizing the Next-Gen Tool & Context-Aware Workflows

With clear research insights in hand, I co-facilitated alignment workshops with engineering and product teams to lay the foundation for the new application architecture. Instead of just redesigning screens, we fundamentally rethought how analysis work is structured.

Building a workflow-based tool

Key Architectural Concepts:

Adaptive, Context-Aware Workflows:

The new UI dynamically adapts based on the user's entry point and task scope. Investigating a single isolated ECU fault demands a different tool layout than conducting an end-to-end full-vehicle audit. The interface guides the specialist through tailored workflow steps accordingly.

Rebranding & System Alignment:

We modernized the tool’s visual language, establishing a scalable, modular UI framework that seamlessly unifies external shortcuts, ticket contexts, and diagnostic tools under one modern design system.

Integrated AI-Assisted Diagnostics:

To reduce manual data correlation, we introduced an AI analysis assistant with dual functionality:

  • Interactive Knowledge Agent: A dedicated chat interface for quick data querying, history lookups, and documentation checks.

  • Active Workspace Copilot: An AI layer that directly interfaces with the live analysis screen, proactively highlighting anomalies and suggesting next diagnostic steps within the ongoing workflow.

Phase 4

High-Fidelity UI Design & Modular Sub-App Integration

With the new architecture defined, I transitioned into designing the high-fidelity web application while adhering closely to BMW’s official Design System. The primary goal was to eliminate context-switching entirely and support the end-to-end vehicle analysis process within a single, unified interface.


To ensure visual consistency and seamless integration across BMW’s enterprise software ecosystem, I built the interface strictly using official design tokens and component libraries. Instead of static dashboards,

With the new architecture defined, I transitioned into designing the high-fidelity web application while adhering closely to BMW’s official Design System. The primary goal was to eliminate context-switching entirely and support the end-to-end vehicle analysis process within a single, unified interface.

the application orchestrates specialized tools – such as ticket integration, vehicle check-in analysis, and causation and effect chain analysis (Wirkketten-Analyse) – into a tailored sequence that dynamically adapts to the specialist’s specific diagnostic problem. Additionally, I designed a collapsable side-panel featuring a context-aware AI assistant, allowing specialists to toggle interactive guidance or query system data on demand during complex analysis steps.

To ensure visual consistency and seamless integration across BMW’s enterprise software ecosystem, I built the interface strictly using official design tokens and component libraries. Instead of static dashboards, the application orchestrates specialized tools—such as ticket integration, vehicle check-in analysis, and causation and effect chain analysis (Wirkketten-Analyse)—into a tailored sequence that dynamically adapts to the specialist’s specific diagnostic problem. Additionally, I designed a collapsable side-panel featuring a context-aware AI assistant, allowing specialists to toggle interactive guidance or query system data on demand during complex analysis steps.

Current Status

With the new architecture defined, I transitioned into designing the high-fidelity web application while adhering closely to BMW’s official Design System. The primary goal was to eliminate context-switching entirely and support the end-to-end vehicle analysis process within a single, unified interface.

The project is currently evolving across multiple development stages. While the core application infrastructure and the AI assistant integration are in active implementation, the individual tool modules span various maturity levels – ranging from conceptual wireframes to high-fidelity prototypes and near-final code.

To ensure visual consistency and seamless integration across BMW’s enterprise software ecosystem, I built the interface strictly using official design tokens and component libraries. Instead of static dashboards, the application orchestrates specialized tools—such as ticket integration, vehicle check-in analysis, and causation and effect chain analysis (Wirkketten-Analyse)—into a tailored sequence that dynamically adapts to the specialist’s specific diagnostic problem. Additionally, I designed a collapsable side-panel featuring a context-aware AI assistant, allowing specialists to toggle interactive guidance or query system data on demand during complex analysis steps.

Concept Landingpage