While the robotics team worked to design and build the system, experts in artificial intelligence and machine learning developed algorithms to restore the frescoes.
After being shattered and buried for centuries, an ancient Roman fresco in Pompeii has been given a second life thanks to a pioneering robotic system designed to help reassemble fragmented artifacts, one of archaeologists’ most arduous tasks.
The technology, developed under an EU-funded project called RePAIR, combines advanced image recognition, AI-driven puzzle-solving and ultra-precise robotic hands to speed up traditionally slow and often frustrating repair tasks.
The robot project, launched in 2021 and coordinated by Venice’s Ca’ Foscari University and unveiled in Pompeii on Thursday, brings together an international research team that used archaeological sites as testing grounds.
The experimental project “actually began with a very specific need to reconstruct fragments of frescoes destroyed during World War II,” said the site’s director, Gabriel Zustriegel.
Researchers believe this technology has the potential to change restoration practices around the world.
A robotic system that combines advanced image recognition, AI-driven puzzle solving, and ultra-precise robotic hands to help archaeologists reconstruct fragmented frescoes that have been buried for centuries was tested at the ruins of Pompeii, Italy, on November 27, 2025. (Credit: REUTERS/Guglielmo Mangiapane TPX IMAGES OF THE DAY)
The robot uses two sizes of flexible hands and twin arms with visual sensors to identify, grab, and assemble debris without damaging delicate surfaces.
The explosion of Mount Vesuvius in 79 AD sank Pompeii into volcanic ash.
Pompeii, a once prosperous city near Naples, and its surrounding countryside were submerged in volcanic ash by the explosion of Mount Vesuvius in 79 AD.
The researchers focused on frescoes preserved in fragments in Pompeii’s warehouses: two large ceiling paintings that were damaged in the first eruption and then shattered by bombing during World War II, and frescoes in the so-called Gladiators’ House, which collapsed in 2010.
A replica was created during this initial testing phase to avoid putting any original parts at risk.
While the robotics team worked to design and build the system, artificial intelligence and machine learning experts developed algorithms to reconstruct the fresco by matching colors and patterns that may be invisible to the human eye.
Experts say the task is similar to solving a giant jigsaw puzzle, but with additional challenges such as missing pieces and no reference image of the final result.
“It’s like buying four or five boxes of jigsaw puzzles. You mix everything together, then throw the box away and try to solve four or five puzzles at the same time,” said Marcello Perillo, a professor at the University of Venice who coordinated the project.
