Image: nist.gov · rights & removal
Lithium Batteries May Help Unravel Mystery of Ancient Papyrus Scrolls
Reporting by NIST Taking MeasureRead the original at nist.gov
Executive Summary
Facts Only
* NIST researcher Cyrus Daugherty developed an algorithm to virtually unroll the tightly wrapped layers of a cylindrical lithium-ion battery using X-ray images taken at different angles.
* The study aims to extend the life of lithium-ion batteries.
* Volcanic material from Mount Vesuvius buried ancient papyrus scrolls in 79 C.E.
* In 1750, workers found charred papyrus scrolls in Herculaneum.
* Initial attempts to unroll Herculaneum scrolls resulted in crumbling or destroyed papyri.
* Researchers used X-rays and CT scans to examine some inked scrolls and develop algorithms for deciphering them.
* A collaboration involving Brent Seales, including computer scientist Seales, developed algorithms similar to Daugherty’s.
* An X-ray spectrometer was proposed as a method to identify lead salts in ancient ink.
* Replicas of the scrolls were created using modern papyrus inscribed with carbon-based ink mixed with lead sulfate.
* Daugherty's algorithm was adapted to examine the replicas, resulting in successful deciphering within days.
Full Take
The narrative demonstrates a powerful thread connecting abstract computational methods—like multi-layered structure analysis in battery science—to tangible historical reconstruction challenges. The core implication lies in the translation of methodology: an algorithm developed for analyzing modern physical layering is successfully adapted to address complex material obfuscation in ancient artifacts. This mirrors the recognition that sophisticated pattern recognition, whether in physics or paleography, depends on finding reliable indicators within obscured data, even when traditional methods fail due to confounding variables like ink composition (carbon versus lead salts). The progression from a practical engineering problem (battery unwrapping) to an esoteric historical puzzle highlights how fundamental principles of algorithmic thinking can bridge disparate fields.
The limitation arises in the transition from empirical success on replicas to the actual ancient scrolls; the successful application is contingent on known ground truth, which introduces a dependency on external validation. Furthermore, the acknowledgment that current methods fail due to elemental ambiguity (carbon vs. carbon) forces the narrative toward novel material science integration, such as using X-ray spectroscopy to exploit element-specific interactions with radiation. The tension between applying established AI/ML techniques and navigating the specific material physics of ancient artifacts suggests a pattern of innovation where existing tools are pushed to their limits to reveal new forms of knowledge that resist conventional observation.
Bridge questions: If the challenge of discerning carbon ink from charred papyrus is overcome through advanced spectroscopy, what other material or chemical ambiguities define the barrier to reading ancient documents? How can the successful adaptation of Daugherty’s model inform broader methods for interpreting historical data that lacks direct physical correlation with modern materials science? What are the long-term ethical and epistemological implications when computational power is used to reconstruct histories where the original evidence remains physically inaccessible?
From the original · NIST Taking Measure
Just a Standard Blog Four years ago, NIST researcher Cyrus Daugherty developed an algorithm to virtually unroll the tightly wrapped layers of a cylindrical lithium-ion battery. Daugherty’s software transformed X-ray images of the battery, taken at different angles, into a CT scan.Read the full story at nist.gov
Sentinel — provisional
No strong signs of machine writing were found in the source article. Provisional estimate, not a finding that a person wrote it.
LIKELY_HUMAN (confidence: 0.15)
This looks only at the wording of the original source article, not at this page's AI-written sections. A small local AI model made this estimate. It has not been checked against known human and machine texts, so treat it as provisional. It cannot show who wrote an article.
