In preparation for the 350th anniversary of the completion of the Basilica di San Petronio (1663), the project Felsinae thesaurus began in 2010 as a major 3D digital survey to inform and analyse the conservation of Bologna's cathedral. Previous restoration projects, such as one carried out between 1972 and 1979 under the supervision of the Superintendent of Bologna, attempted to address the erosion caused by weather and pollution, but the resins used were found to be damaging the stone of the façade.

One of the main tasks of contemporary conservation practice is the high-resolution recording in both colour and three dimensions to establish and monitor the rate of both past and future decay. Factum Arte (and, later, Factum Foundation) was tasked with the high-resolution 3D scanning of all the sculptures on the Basilica's three main doors of its unfinished façade. The recording work, both white light scanning and high-resolution photography was carried out alongside the restoration work undertaken by the Opificio delle Pietre Dure and the Bolognese restoration company Leonardo.

The documentation work began in April 2013. A previous 2011 test scan of the figure of San Petronio by Jacopo della Quercia, on the right-hand side above the central door, made clear the aim of pushing the limits offered by high-resolution scanners and establish a protocol that would provide data for both virtual/screen-based uses - and the production of exact facsimiles if ever required. The scaffolding needed to be stable enough for the relatively long exposures required to record data at different resolutions (135 microns and 250 microns) using a NUB3D Sidio white light scanner.

Physical rematerialisation tests were also made following several conversations with architect Roberto Terra (Cavina-Terra Architetti) and Monsignor Oreste Leonardi about the best way to preserve the statues on the façade of the church.

Factum's recording work lasted over six months and involved a group of ten people working in two shifts both during the day and the night. Four different 3D scanning systems were employed: three white light scanners and a medium-range laser scanner. A NUB 3D Sidio scanner was mainly used with the support of a Breuckmann Smart Scan 3D and a DAVID SLS-1 to reach difficult areas of the niches. A major part of Factum's work and expertise is to recognise and understand both the potential and limitations of each scanning technique in order to integrate the resulting data during the workflow.

Afterwards, the post-processing work involved a team of three people to clean, merge, align, render and prepare the files in order to create a vast 3D and colour archive documenting the beguiling beauty of the works adorning the façade of San Petronio. PolyWorks was selected as the best software for this part of the job. Where physical limitations of the non-invasive recording process resulted in missing data, the surface was remodelled following extensive photographic references by Grégoire Dupond and Gabriel Scarpa. 

In order to correctly record the position of each sculpture on the façade, Factum Arte worked with ScanLAB London to use a FARO Focus X330 system to accurately record the spatial measurements. When integrated with the other data, this produced a complete 3D model of the three doors of the façade of San Petronio. 

More than 20 carvers worked on the doors of San Petronio. The sculptural groups above each of the doors were carved between 1430-35 by Jacopo della Quercia (main door - the Virgin Mary and Christ and the patron saint of the city, San Petronio), Domenico da Varignana (Sant'Ambrogio), Alfonso Lombardi (left door, The Resurrection of Christ), Francesco da Milano (two soldiers on left and right side of Christ), Amico Aspertini (central group on the right door depicting the Deposition of Christ), Niccolò Tribolo (Madonna on the left) Ercole Seccadenari (San Giovanni on the right). With them, a host of other artists that defined the Bolognese Renaissance worked on the lunettes and carvings: Zaccaria Zacchi da Volterra and sons, Giacomo Silla, Lazzaro Casario, Girolamo da Treviso, Diego Sarti, Antonio Solosmeo and Properzia de' Rossi. 

The data recorded by Factum Foundation helped to assist and inform the restoration process carried out and its condition monitoring, ensuring the façade of San Petronio will survive for many generations to come.

A 1:5 prototype of the Central Door, with its reliefs and statues, was initially produced at a resolution of 100 micros using a sintering 3D printer. This proved it was possible to accurately rematerialise the surface of each statue in case of need. 

Sculpted by Amico Aspertini (1474–1552) and his workshop between 1526 and 1530, the lunette tableau above the right-hand portico of the façade shows Christ being lowered by either Joseph or Nicodemus while the two Marys lament to either side. Famously described by Giorgio Vasari as ‘an eccentric man of extravagant brain, whose figures… are equally eccentric and even mad’, Aspertini was a skilled and prolific painter and sculptor; Vasari writes that ‘there is no church or street in Bologna which has not some daub by the hand of this master’.

By 2010, when the team from Factum Foundation undertook to scan the three doors on the unfinished façade, Aspertini's sculptural group was in a state of extreme fragility. Sections of the marble had denatured and dissolved as a result of time and exposure, a process probably accelerated by modern airborne pollutants and acid rain, and the arms of Jesus in particular, projecting from the main bulk of the statue, were in need of significant consolidation work.

Given these circumstances, there was extensive discussion at this time as to whether the most deteriorated statues of the façade should be left in situ or whether they should be brought down from the façade and exhibited in a museum setting within the church complex. This latter solution would have allowed the sculptures to be seen up close and at eye-level in a far more controlled setting, with facsimiles or copies replacing the originals on the front of the Basilica. The Opificio delle Pietre Dure ultimately decided to preserve Aspertini’s original sculptures in situ. But the use of copies, far from being radical, is one which has long been used in church restoration schemes and is deeply embedded in historical practice – one thinks, for example, of the replicas of the horses from Constantinople that now occupy the place of the fragile originals on the loggia of San Marco in Venice, or of Viollet-le-Duc replacing the damaged 13th-century sculptures of the kings of Judah in Notre Dame, decapitated during the French Revolution.

During cleaning, it was decided to remove previous restorations from the left arm of Jesus in order to safeguard the group, leading to a further difficult decision: whether to replace the deteriorating stone with a lightweight facsimile, which would place less stress on the fragile upper arm, or whether to reattach the original using a metal pin which would add to the stress on the marble. Scans made by Factum (using a NUB3D Sidio Pro 3D Scanner, with additional data provided by a Breuckmann Smartscan 3D) provided the restorers with detailed information to aid them in making this decision. The original arm was carefully removed and a ‘prosthetic’ acrylic resin arm was made and tested, but a collective decision was ultimately made by the restoration teams to keep the original sculpture on display. This involved reattaching the arm with a visible metal support structure – a solution that, although it avoids the stress which would have been caused by a single pin, remains aesthetically awkward.

The scan data turned out to be important for condition monitoring sooner than expected. During the restoration process the right arm of Christ also broke as a result of the weakness of the stone, resulting in damage to the hand and fingers. Fortunately, as scanning had been carried out before the breakage, the data could be used to inform the reconstruction, and the restored arm was also replaced using a metal exoskeleton. Although the main aim of the recording was to create an accurate report for monitoring the condition of the sculptures, the scan data will also provide future restorers with a firm basis from which to replace the damaged original arms with digitally restored facsimiles should they ever wish to do so.

Another place in which scan data has resulted in substantive changes to our knowledge of the sculpture is the turban of Joseph of Arimathea. This is carved as a separate unit from the head of Joseph and is held in place by a square locating pin. The turban was removed during cleaning and was scanned independently of the sculpture, allowing both the visible surface and the connecting join to be recorded. Once removed, it became evident that there were two feasible positions for the headpiece, one with the knot at the front and one with the knot at the back. Following the 1972–79 restoration campaign the turban was placed with the knot at the front, although the scan data suggests that it fits the sculpture better with the knot at the back.

As Amico Aspertini’s Deposition clearly illustrates, the state of an original artwork changes significantly over time. The choice faced by the restoration team is one faced by all guardians of buildings and works of art: whether to leave external sculptures in their original location to slowly deteriorate, treating them as objects with a quantifiable life-span, or whether to preserve the originals and to replace them with replicas (to borrow the opposition presented by Jean Clair in L’hiver de la culture, Paris: Flammarion, 2011). While there is no one-size-fits-all answer, the fact that objectively accurate facsimiles are now a possibility makes this choice a more real and immediate one than ever before.