How Deutsche Bahn Is Digitizing Its Rail Network with 360° Video

• Hundreds of small stations impossible to document with LiDAR alone
• No consistent pipeline from raw video to usable 3D data
• Spatial data siloed from operational systems like SAP PM
• Technical expertise required to capture and process site conditions

• Hundreds of stations digitized using affordable 360° cameras
• Significant reduction in on-site time for briefings and asset documentation
• Non-technical staff can now create and share 3D models independently
• SAP PM integration links each spatial asset to its operational metadata
• One unified platform combining point cloud, mesh, and 360° imagery across teams and departments
• Virtual site access for safety briefings and contractor onboarding before physical visits

Europe’s largest rail operator manages over 5,000 stations across a continent — and needed a way to digitize all of them without deploying specialist scanning teams to every site.

DB Systel GmbH, Deutsche Bahn’s digitalization subsidiary, turned to Prevu3D to build a reality capture pipeline that any employee could run with an affordable 360° camera.

What followed was a scalable 3D digital twin program that cut on-site time, connected spatial data to SAP PM, and gave maintenance and operations teams a virtual window into every station on the network.

The Challenge: Turning 360 Video Capture into Usable Reality

DB Systel GmbH’s Immersive Technology team needed a scalable way to digitize stations and facilities quickly, using a two-track approach suited to the diversity of the rail network.

Germany’s rail infrastructure ranges from major urban hubs to hundreds of small, rural stations—and not every site warrants the same level of scanning investment.

For large, complex stations, high-fidelity mobile LiDAR scanning remains the right tool. But for the long tail of smaller stations, the team needed a faster, more accessible method: 360° video capture.

Rather than replacing LiDAR, 360° represents the first milestone in a broader capture strategy—enabling panoramic visualization, spatial understanding, and asset location at scale, without the need for specialized equipment or technical expertise.

At the same time, DB Systel GmbH needed:

• A consistent 3D data pipeline from 360° video to E57 and mesh.
• The ability to combine scans from multiple devices such as 360° cameras, mobile LiDAR, and drones into unified, scalable environments.
• A process to keep visual data synchronized with operational systems like SAP PM, ensuring that spatial information supports daily maintenance and safety workflows.

The Solution: Prevu3D’s Reality Capture-to-Mesh Pipeline

Partnering with Prevu3D, DB Systel GmbH introduced a new 360° camera-based capture workflow to transform raw video footage into reliable 3D environments within RealityPlatform™.

How it works:

Capture – Anyone at DB can now document stations using affordable 360° cameras such as the Ricoh Theta X or Insta360.

This low-barrier workflow makes reality capture accessible to non-technical users, allowing them to record walkthrough videos across hundreds of stations.

Processing – Prevu3D’s pipeline converts uploaded 360° footage

into point clouds (E57) and meshes through an optimized Capture-to-Mesh Engine—eliminating the need for manual user manipulation. Users upload once; the system handles processing automatically.

Integration – The generated RealityTwin™

Can be integrated with SAP PM equipment IDs, linking each asset to its corresponding metadata once setup is performed.

After configuration, data updates are managed manually through RealityAssets™ to ensure accuracy and control.

Visualization & Editing – With RealityPlan™

Teams inspect, understand, and annotate environments for safety instructions, equipment maintenance, and planning tasks.

Scalability & Access – Through RealityPlatform’s enterprise cloud

Visual data is streamed securely and shared across departments—supporting remote collaboration, onboarding, and enterprise-scale site management.

The Result: Improved data fidelity, usability, and scalability

Prevu3D enables DB Systel GmbH to unify data fidelity, usability, and scalability—key pillars repeatedly emphasized by the team’s technical leads.

The RealityTwin™ open architecture supports the integration of reality data, CAD/BIM, and enterprise systems, empowering DB Systel GmbH to connect the 3D visualization layer directly with operational tools like SAP PM once initial linking is configured.

Key benefits realized:

Automated 360° Processing – Reduced manual intervention and re-processing time.

Enterprise Scalability – Data bundling allows multiple representations (point cloud, mesh, 360° imagery) to coexist within a single dataset, making it easy to share and reuse across projects and teams.

Low Training Threshold – Non-technical teams can navigate, add, and share 3D models from any computer, democratizing access to spatial data.

Improved Data Quality in SAP PM – 3D visualization enables faster asset identification and updates to master data.

The Impact

Through this 360° capture initiative, DB Systel GmbH has processed a large and growing number of stations using 360° videos and Prevu3D’s Capture-to-Mesh engine.

The program has delivered measurable improvements in operational efficiency, including significant reductions in on-site time for tasks such as briefings and asset documentation.

Safety and training workflows have also improved, allowing staff and contractors to explore stations virtually before visiting them physically.

Prevu3D helps us simplify the capture of stations. Now every employee can create basic 3D models themselves—scalable across hundreds of stations.

— Jonas Heidecke, Requirements Manager, DB Systel GmbH

Looking Ahead

As DB Systel GmbH scales the program, the Prevu3D ecosystem will serve as the visual foundation for its enterprise-wide 3D visualization strategy, linking Reality Capture, Object Library, and SAP data into a unified operational environment.

Future phases include leveraging RealityComposer™ to facilitate the integration of meshes captured from different scanning equipment.