Answer extracted from the Digital Construction Podcast — listen to the full episode below.
Why does scan-to-scan navigation work better than free 3D flight for mesh data in web browsers?
Most users prefer scan-to-scan navigation because flying around in free 3D space causes disorientation and confusion. This mode—similar to Google Street View, jumping between fixed camera positions—feels intuitive and prevents the dizziness associated with free 3D flight. A teleport feature then extends this by allowing teams to create virtual scan locations at positions where no original camera existed, giving full navigation flexibility without sacrificing user comfort.
When streaming high-resolution 3D mesh data in a web browser, the navigation experience becomes critical to how users actually engage with the data. As Dominique Pouliquen explains in the Digital Construction Podcast episode, the platform offers two fundamentally different modes, but user behavior quickly revealed which one works.
Scan-to-scan navigation mimics the familiarity of Google Street View: users jump from one pre-captured scan position to another without interpolation or continuous movement. This approach feels anchored and predictable—users always know where they are because they're always standing at a real camera position that was physically captured.
Free 3D flight, by contrast, lets users fly anywhere through the mesh in true 3D space. While this sounds powerful in theory, it creates a concrete usability problem: disorientation. Without a fixed frame of reference, users lose their sense of position and become confused about where they are relative to the scanned environment. This confusion translates to lower engagement and higher cognitive load.
3D Mesh vs. Point Cloud
A point cloud is a collection of 3D dots in space—you can see through the gaps. A 3D mesh converts those dots into continuous surfaces, matching the actual walls, floors, and objects in the real world. Meshes compress data more efficiently and are much easier for humans to interpret visually.
The breakthrough came with a feature that solved both problems: teleportation to virtual scan locations. Rather than restricting users to real camera positions, the platform lets teams create teleport points at any location within the mesh—even where no original scan was captured. Users still get the comfort of snap-to-point navigation without losing the flexibility to explore the full environment.
This design insight illustrates a larger principle in reality capture workflows: the technical capability to stream high-resolution 3D data is only half the battle. How users navigate and interact with that data determines whether the platform actually gets used. The teleport feature turned navigation from a binary choice (confined vs. lost) into an intuitive, scalable solution.
"Everything takes longer than expected, and usually it's three times longer than what you think."
Dominique Pouliquen — Co-founder and CEO of Cintoo. Pouliquen led reality capture at Autodesk, where he co-founded the ReCap team in San Francisco to democratize point cloud access across Revit, AutoCAD, and other construction software. He later left Autodesk to co-found Cintoo with PhDs specializing in point cloud compression technology, bringing that compression and streaming capability directly to web browsers for construction teams.
The shift from scan-to-scan to the teleport model also reflects a shift in how construction teams actually use mesh data. Rather than passive viewing, users need to annotate, measure, and communicate problems on site. The episode explores how these navigation modes evolved in response to real user feedback, turning a potential weakness of web-based 3D into a strength by matching navigation to the way humans naturally explore space.
See also
What business model decision proved most effective for a platform targeting large construction enterprises?
The founders initially based their business model on tokens, similar to Autodesk's approach, but this created a problem: users felt afraid to consume tokens because they were worried about running out, paralyzing adoption.
Why did the founder leave a senior role at Autodesk to start a new company?
While leading ReCap at Autodesk, Pouliquen identified a major headache: the workflow was very desktop-centric and there was no easy way to upload, share, and collaborate on point cloud data across teams in a cloud-based environment.
What is the core principle behind compressing massive point cloud data into formats suitable for cloud sharing?
The technology converts huge point clouds into 3D meshes or 3D surfaces, which solves the problem that point clouds are like 3D dots in space where you see through the dots. Surfaces are continuous and match how the real world actually looks.