How Viscan turned a failing reality capture workflow into a landmark success

For Nicolai Nolle, CEO of Viscan GmbH, the goal was to deliver a high-accuracy digital twin for infrastructure construction—capturing and updating site conditions daily using advanced reality capture workflows. They were the digital backbone for the B29 highway expansion, a landmark project set to become the first 100% digital construction site in Germany’s Baden-Württemberg region. Their job was to provide a constant, up-to-date digital twin of the entire site, documenting progress every single day using the latest in reality capture technology.

With a complex project involving nine new structures, bridges, and retaining walls, the pressure to succeed was immense. Every stakeholder, from the planners to the construction teams, relied on Viscan’s data to be fast, accurate, and, above all, reliable.

However, in the critical early stages, the team hit a wall. The cloud solution that they had chosen, the very engine that was supposed to power their digital dream, began to fail them.

When your tools can’t keep up with your vision

The Viscan team had a brilliant workflow for capturing data. Using a combination of drones and their own innovative scin® (SC – anning IN-tuitively) app on handheld scanners, they made it easy for anyone on site to capture the reality of their progress. The problem wasn’t their method; it was the back-end processing.

When they sent the massive amount of daily data to their initial cloud solution, they were met with frustration. They faced massive performance and quality problems. The accuracy was so poor that the team had to manually rework individual scans, completely defeating the purpose of an automated workflow and putting the project’s tight schedule at risk. Their digital backbone was cracking under the strain.

Choosing a scalable digital twin partner

The team at Viscan didn’t want to compromise their high standards. So, they made the tough but necessary decision to find a new engine for their workflow, one that could deliver the reliability and performance they demanded. They chose to partner with Bentley Systems because they needed a high degree of reliability and a strong partner who can address problems and issues promptly and competently.

This strategic pivot was the turning point.

“Bentley provides us with the iTwin Capture solution as an OEM partner, in turn this is reducing costs and significantly speeding up processes in the documentation of construction sites.“

— Nicolai Nolle, CEO and BIM Professional, Viscan GmbH

With their new partnership, the Viscan team re-engineered their workflow for maximum efficiency. They implemented iTwin Capture as their new cloud processing engine. By doing this, they took direct control over their reality data pipeline. The performance and quality issues vanished. The team could now reliably process the massive datasets, automating the production of high-fidelity reality meshes and clean, georeferenced point clouds from the raw image data. The need for manual rework disappeared.

Next, they leveraged Bentley’s Orbit 3DM to create a single, cohesive project environment. This allowed them to combine their own mobile mapping data with other sources, giving every stakeholder a unified view of the project without cumbersome data conversions. This powerful visualization environment became the launchpad for their AI workflow.

Powering a human-in-the-loop AI workflow

This reliable data foundation from iTwin Capture and the unified viewing environment from Orbit 3DM were the keys to unlocking the power of their custom artificial intelligence.

To analyze the millions of images, Viscan employs sophisticated imaging AI methods to automatically detect objects and identify changes. But the AI’s findings must be verified by experts. To empower them, Viscan built their scalable annotation platform around the Orbit 3DM viewer. This enabled their team to virtually navigate any part of the site, overlay engineering designs on top of the reality data, and visually verify the AI’s findings.

When a correction was needed, they could quickly annotate assets directly within the 3D environment. This created a powerful feedback loop: every human verification was fed back into the system, making Viscan’s own AI models smarter and more accurate for the next dataset.

By actively integrating these powerful tools, the Viscan team bypassed their earlier roadblocks and focused their resources on what they do best: delivering exceptional value to their client.