An Algorithm for Preserving Art
A medieval art museum models the building’s climate using wireless sensors to better maintain its collection.
Paolo Dionisi Vici’s life’s work is preserving aging art made from wood. He gets misty-eyed encountering rare artifacts from his hometown in Tuscany. He looks strikingly like Frank Zappa.
In other words, Dionisi Vici seems like an unlikely person to get excited about the wireless sensors that are typically used to monitor the temperature of busy computers packed into IBM data servers.
But that’s exactly what he was excited about on April 6 in the busy halls of the Cloisters, the Metropolitan Museum of Art’s upper Manhattan branch, which contains some 3,000 medieval works.
With technical assistance from IBM, Dionisi Vici, an associate research scientist for the Met, has deployed 120 low-power temperature and humidity sensors there since June of last year in his quest to determine the ideal environmental conditions for priceless wood works.
The project could provide better models for art conservators, who have historically relied on intuitive observations and more generalized, slower measurements to prevent cracking, warping, and mold. One day it could provide a model for maintaining the Mona Lisa, which has an infamous 11-centimeter crack through its poplar wood panel. Dionisi Vici has been among the few who have had access to the painting to study its condition.
Currently, most artwork at the Cloisters is kept at a standard range around 70 °F and 50 percent humidity, but Dionisi Vici’s goal is to arrive at more nuanced formulas based on how individual art reacts to changing conditions.
Right now, he’s in his lab looking at old scrap wood. His long-term plan is to use real-time sensor data combined with a gauge of the art’s response, such as if the wood swells in certain conditions, to improve art conservation. After all, with visitors moving in and out the building’s open courtyards, climate control is not always easy.
“The idea is to see how right is the climate for a certain object. My real dream would be to have a system where the parameters we adopt are based on the real sensitivity of the objects,” says Dionisi Vici.
Those findings could one day program a self-controlling heating and cooling system that responds to climate fluctuations measured by the sensors, he says.
His search for the perfect climate is enabled by inexpensive sensors scattered unobtrusively about the Cloisters’ high-ceilinged rooms.
The sensors collect temperature and humidity data once per minute, and use radio frequencies to transmit it back to a router that sends the data to IBM’s cloud servers. The result is a nuanced, 3-D building temperature and humidity map that both Dionisi Vici and the museum’s maintenance engineers can view from their desks. The low-power microcontroller has an expected five-year battery life.