In 2017, the Knight Foundation awarded Vizcaya Museum and Gardens a Museums and Technology grant to turn preservation-grade 3D documentation into public interpretation. I was working at Vizcaya as its digital strategist and helped implement the technology program.

The project focused on two of Vizcaya’s most significant but difficult-to-access features: the Barge in Biscayne Bay, decorated with sculpture by Alexander Stirling Calder, and Robert Winthrop Chanler’s swimming-pool grotto ceiling. Both were important to the site, both were vulnerable, and neither could be experienced closely by most visitors.

Note: This post reflects my own professional experience and perspective. It does not represent an official position of Vizcaya Museum and Gardens.

The problem was not scanning

By the time of the grant, laser scanning and photogrammetry could already produce extremely detailed records of historic places. The harder problem was translation: preservation data was large, technically demanding and created for accuracy rather than visitor interaction. The project asked whether the same source data could support conservation, web access, gallery interpretation, virtual experiences and tactile learning without rebuilding the project from scratch for every use.

That distinction mattered. A preservation-quality point cloud might contain exactly the information a conservator needs and still be far too heavy to manipulate smoothly on a museum kiosk or tablet. Turning documentation into interpretation required choices about file formats, point budgets, graphics performance, hardware, interface design and the amount of detail that could be retained without making the experience unusable.

A network of partners and formats

The work became a multi-institutional pipeline. University partners captured laser scans and photogrammetry. Files moved among preservation and interpretive formats including E57, PLY, OBJ and STL. Daruma Tech developed interactive experiences from those assets. Florida International University helped translate models into 3D-printable objects. National Park Service Historic American Buildings Survey staff reviewed an alpha version and advised on terminology, interaction, software conversion and point-cloud performance.

The kiosk software used open-source tools including Potree, PotreeConverter and three.js. The project documentation and implementation were maintained in GitHub, and the final report explicitly described the development approach as open and replicable, with project materials released under the MIT License.

What actually got built

The grant produced more than one visitor interface. Vizcaya developed a large 4K touchscreen kiosk, portable Surface-based kiosk experiences for the Barge and Grotto, an online Virtual Vizcaya experience drawing point clouds from Amazon S3, virtual-reality assets based on photogrammetry, and 3D-printed replicas for tactile interpretation. Staff evaluated the main kiosk through observation, time-lapse capture and built-in analytics.

The project also exposed the limits of the technology. Even powerful Surface hardware required the number of rendered points to be reduced substantially. Native preservation point clouds did not move cleanly into every virtual-reality environment. Different uses required different conversions. “One scan, many uses” was possible only because a translation layer existed between the archival data and each public experience.

Why openness mattered

The proposal had always framed replicability as part of the project, but the final report shows that this became a real implementation choice. Rather than treating the kiosk as a proprietary end product, the team documented the stack, shared code and recorded the conversions and dependencies another institution would need to reproduce the work.

That was an important shift in emphasis. The useful artifact was not only a kiosk sitting at Vizcaya. It was a documented pathway from preservation capture to public experience.

What I see now

Looking back, this was one of the points where my thinking about preservation technology widened into systems thinking. The scan itself was no longer the end product. Its value increased when the same underlying knowledge could move through conservation, interpretation, accessibility, learning, web delivery, physical replication and institutional partnerships. The project was really about building relationships among data, people and uses.

That also explains why some of the technology has aged while the approach has not. Platforms come and go. File standards evolve. Hardware becomes inadequate. But the architectural question remains: can an institution create digital assets once, preserve them well, and then responsibly adapt them for many forms of public meaning?

If you want a primer on the underlying technologies, see 3D Documentation: A Primer for Cultural Heritage.