In 2020, RSET teamed up with Seth Parker and researchers at the University of Kentucky on a fascinating challenge: developing a custom photogrammetry scanning system to help document previously unrolled fragments of the Herculaneum papyri.

Some of the historical and research information in this post is referenced from the Vesuvius Challenge article, “New Scans of Herculaneum Papyri at the Institut de France and Bodleian Libraries,” which provides additional background on the ongoing effort to image and recover these remarkable ancient texts.

The Herculaneum scrolls were carbonized by the eruption of Mount Vesuvius in AD 79. Many remain rolled and extraordinarily difficult to read. Others were mechanically unrolled between the 18th and 20th centuries, often using methods developed by Antonio Piaggio.

Previously opened fragments are preserved at the Officina dei Papiri Ercolanesi at the National Library of Naples. Many were mounted to backing material and placed in metal frames known as cornici. The collection includes approximately 1,840 fragments distributed across numerous trays and cabinets.

These fragments presented a unique engineering challenge. The papyrus is extremely thin and fragile. Individual pieces can be crumpled, folded, and layered, while the black carbonized surface makes the carbon-based ink difficult to distinguish with the naked eye.

Our solution started with two X-Carve CNC machines, but we used them in a completely different way than they were originally designed.

Instead of cutting material, RSET modified the CNC systems to precisely position custom camera arrays above the papyrus. The machines provided millimeter-level movement and repeatability while keeping the imaging equipment from contacting the fragile fragments.

RSET modified the CNC machines, designed and 3D printed the camera arrays, developed custom software to control their movement and trigger the cameras, and built two custom transportation containers so the systems could be safely shipped to Naples.

The scanners were designed to quickly and systematically capture large numbers of overlapping photographs from controlled positions. Those images could then be processed through photogrammetry to create high-resolution 3D representations of the surface geometry of the papyrus fragments.

One potential application we explored was using that geometry to 3D print inverse supports precisely matched to individual fragments, providing custom holders capable of supporting these incredibly delicate artifacts.

But building the scanners was only one small part of a much larger effort. Researchers at the University of Kentucky and EduceLab have continued pushing this work much further, combining high-resolution photography, infrared and spectral imaging, photogrammetry, X-ray CT scanning, and machine learning to help reveal writing that has been inaccessible for centuries.

Images of visible ink on previously opened fragments can provide valuable ground-truth data for training models to identify faint ink within scrolls that remain rolled. The ultimate goal is extraordinary: recover writings that have been inaccessible for nearly 2,000 years without physically unrolling and potentially destroying the scrolls.

We are proud of RSET’s small part in this remarkable project — and especially proud of the researchers at the University of Kentucky and the other institutions involved who have done the real magic with the tools, imaging, research, and technology.

For RSET, it remains one of our favorite examples of what can happen when photogrammetry, CNC technology, 3D printing, custom software, engineering, and academic research come together to solve a problem for which there simply isn’t an off-the-shelf solution. It’s exciting to see how far this work has progressed and where it may lead next.

Image credits: Images 1, 9, and 10 are courtesy of the Vesuvius Challenge Substack — scrollprize.substack.com