Project Overview
The Digital Twin of Switzerland was the concept that won the InCube Challenge 2022, where our team placed first among 20 international participants. The proposal focused on building a country-scale digital twin by using data collected through Swiss Post vehicles as they moved through the national road network.
I led the team and helped shape the system concept, technical framing, and presentation strategy. The core idea was to transform an existing logistics fleet into a scalable sensing platform for geospatial mapping, infrastructure monitoring, and future smart-city applications without relying on a dedicated nationwide scanning fleet.
Technical Implementation
3D Mapping
Designed around large-scale reconstruction and digital twin generation from fleet-collected data
Computer Vision
Vision pipelines for extracting usable scene structure and street-level observations from mobile platforms
Fleet Data Strategy
Leveraging recurring delivery routes to turn logistics operations into a persistent sensing network
GIS Integration
Geospatial reasoning for infrastructure monitoring, planning workflows, and map-centric analytics
Innovation & Impact
The strongest part of the project was the operating model: using a fleet that already covers Switzerland at scale rather than assuming a dedicated mapping operation. That idea made the concept relevant both technically and commercially.
- Scalable Sensing: Reused an existing mobility network instead of requiring specialized data-collection vehicles
- Recurring Coverage: Daily vehicle movement made repeated observation and update cycles realistic
- Street-Level Signal: Ground vehicles provided a practical path to detailed infrastructure understanding
- Platform Potential: The same foundation could support planning, monitoring, and broader smart-city workflows
Technical Challenges & Solutions
Building a nationwide digital twin concept around fleet data raises challenges that go beyond standard mapping demos:
Coverage and heterogeneity: Vehicles, routes, sensors, and operating conditions are not uniform. The concept needed to account for inconsistent viewpoints, changing environments, and a data stream collected over time rather than in a single scan pass.
Calibration and consistency: Any fleet-based mapping solution depends on trustworthy calibration, timestamping, and fusion logic if the resulting map is going to support downstream planning and monitoring use cases.
Privacy and operational fit: The concept also had to respect privacy requirements while fitting into an existing logistics workflow without adding unrealistic operational burden.
Results & Recognition
The project won InCube Challenge 2022 and was recognized for combining technical ambition with an execution model that could plausibly scale. It stood out because it connected perception, mapping, and systems thinking to a concrete real-world deployment context.
For me, it was also an early demonstration of a pattern that still defines my work today: designing perception systems that have to make sense not just in isolation, but inside a larger operational pipeline.