01 — Project Overview
Enterprise Digital Twin
Manufacturing Platform
Barry Callebaut operates one of the world's largest chocolate manufacturing networks, producing thousands of tons of chocolate products across multiple factories worldwide.
Manufacturing operations involve hundreds of production parameters—including ingredient ratios, equipment utilization, production schedules, machine temperatures, throughput targets, quality checkpoints, and logistics constraints. Production planners and quality teams often need to answer complex "what-if" scenarios before executing changes on the factory floor.
"Traditional planning methods relied heavily on spreadsheets, disconnected manufacturing systems, manual calculations, and delayed reporting. This made production planning reactive instead of predictive."
The goal of this project was to design an enterprise Digital Twin platform capable of simulating factory operations before changes were deployed in the real world. The platform combined real-time production monitoring, predictive simulations, AI-assisted recommendations, and governance workflows into a single unified experience.
02 — Business Context
Balancing Trade-offs in
Chocolate Production
Manufacturing organizations generate enormous volumes of operational data every minute. However, data alone does not enable better decisions. Production planners constantly balance competing priorities:
03 — Problem Statement
The Reactive Planning
Bottleneck
Planners and QC teams faced mounting pressure to maintain high throughput and product quality under volatile real-time conditions. Without simulation tools, predicting downstream consequences of a line adjustment or demand spike took hours or days of manual calculation.
04 — Strategic Objectives
Primary Goals
05 — Success Metrics
Measured Outcomes
06 — Responsibilities
My Design Role
As the Product Designer, I owned the end-to-end user experience for the Digital Twin platform:
07 — Target Users
User Personas
Production Planners
Increase throughput, balance production load, and allocate resources efficiently across factory lines.
Quality Controllers
Maintain product consistency, detect quality deviations early, and prevent defective batches.
Operations Directors
Monitor global factory performance, evaluate risks, and optimize regional asset utilization.
08 — Interactive Prototype
Barry Callebaut Chocolate
Factory Simulator
Adjust the tempering temperature and conching duration in the sidebar to run simulated Digital Twin projections.
Ghent Plant Simulation Output
SYSTEM STATUS: OPTIMALCurrent parameters are within optimal crystallization limits. Crystal type V formation is stable. No downstream packaging bottlenecks detected.
09 — Design System
Component Architecture
The platform leveraged a modular component library customized with Barry Callebaut brand aesthetics:
10 — Research & Discovery
User Research Methods
We conducted qualitative interviews and quantitative surveys with production planners and quality controllers across multiple European and North American facilities.
11 — Opportunity
The Opportunity Statement
"How might we transform complex manufacturing telemetry and simulation models into an intuitive decision-support tool that empowers plant managers to predict quality issues and optimize factory throughput before execution?"
12 — UX Design Principles
Core Design Principles
13 — Product Ecosystem
Information Architecture
The platform architecture is structured into three main layers:
Layer 1: Enterprise Overview
Global factory status, plant comparisons, and critical deviation alerts.
Layer 2: Operational Monitoring
Live production metrics, equipment health, line efficiency, and quality monitoring.
Layer 3: Decision Support
What-if simulations, AI-generated recommendations, and structured approvals.
14 — Design Decisions
Key Architectural Choices
Separation of Concerns
Separating Watch Tower monitoring from the active Simulator workspace to prevent cognitive overload.
Progressive Disclosure
Global View → Factory View → Machine View → Simulation Run → Approval Workflow.
15 — Technical Considerations
Performance & Governance
Engineered to handle high-frequency IoT sensor telemetry streams while maintaining a responsive, fluid user interface during complex simulation runs.
16 — Final Solution
Unified Manufacturing Hub
The final solution unified disparate manufacturing systems into an intelligent enterprise platform combining operational monitoring, predictive simulation, and AI-assisted planning within a single closed-loop ecosystem.
17 — Business Impact
Measured Outcomes
18 — Future Roadmap
Platform Expansion
Predictive Maintenance
Integrate machine health IoT sensors to forecast equipment failures before line downtime.
Sustainability Analytics
Monitor energy consumption, carbon emissions, and water efficiency across plants.
AI-Powered Scheduling
Automatically recommend optimized daily production schedules based on raw cocoa constraints.
19 — Reflections & Summary
Key Learnings
Working on the Barry Callebaut Digital Twin Platform reinforced that successful manufacturing products are built around decisions, not data. Trust in automated AI recommendations is earned through transparency and explainability.