HBIM between Antiquity and Industrial Archaeology: Former Segr è Papermill and Sanctuary of Hercules in Tivoli

: Industrial heritage with secular production activity constitutes a speciﬁc ﬁeld of application to reﬁne digital tools for knowledge within the HBIM (Heritage Building Information Modeling) process. Industrial sites are traditionally linked to the exploitation of local resources, and, not infre-quently, are settled by recovering the ruins of ancient buildings and monuments. The Sanctuary of Hercules in Tivoli represents a signiﬁcant case study that moves between classical and industrial archaeology, in particular the “Cartier à Segr è ”, hereinafter referred to as former Segr è papermill, to test ArchaeoBIM concepts and to investigate current and lost heritage. Starting from the documents and the digital survey with the use of UAV videogrammetry, the aim is the construction of the informative model with particular attention to the 4D management to describe the evolution phases and the exploration of the construction speciﬁcities of buildings and machines between pre-modern techniques and industrial age. The results show the possibility of creating a diachronic HBIM to investigate a complex industrial heritage, its evolution and production phases, modeling components for this type of architecture, with the deepening of the LOD of BIM (Building Information Modeling) instances applied to machines. The application represents an augmented knowledge process appli-cable on industrial heritage through modeling instances of machines and industrial processes that would allow regional and transnational cross-sectional studies and the enhancement of fruition and reuse of these sites.


Antiquity from Secular and Enduring Production and Antiquity from Reuse
When the palimpsest of urban and suburban industrial landscapes are observed [1], two modes emerge in which historical industrial heritage is rooted in the historical built environment. Objects may possess their own antiquity from secular and enduring production or be the bearers of an antiquity from reuse. Either way, cultural identity and landscape "heritage" can be expressed in both material and nonmaterial evidences, through historical, political, social, and constructive histories [2,3].
On the one hand, the first category is found in the longevity of productive districts linked to specific resources and deep-rooted skills. To give some examples, we can cite the activity of mining sites, such as the Colline Metallifere [4,5], or mining sites, such as the Apuan marble basins of Carrara, or the Tiburtine travertine basins of Tivoli. From a manufacturing point of view, we can mention the paper districts of Fabriano, Amalfi [6], or the Liri river basin [7].
Concerning the latter, an exemplary case is the paper mill set by the Montecassino Abbey in Sant'Elia Fiumerapido [8], active from 1516 to 1985, in which the mechanical innovations of the Fabriano masters were immediately applied, such as the pestle hammer, and with time, the various organizational and industrial innovations followed one another, including the introduction of the Hollander beater, the wood departments, and the modern the verification the potential that digital tools [17], as they have long been used in archaeology, having to build effective models of representation of palimpsests, a digital model that allows for a comprehensive view of classical archaeology and industrial archaeology. With this in mind, the complexity of information modeling of requests for cultural heritage must necessarily follow the BIM method so that information aspects from manual and archival sources must be associated with a geometric modelling: key parameters, instances, timing, and algorithms will allow for the emergence of phases, groupings, and relationships. Historical studies therefore come into play upstream and downstream of the BIM process. -the illustration of the specific objectives that with these premises have been set to this research. In the case study, a process that contemplates at least the fourth and sixth dimensions of BIM (time and management) certainly allows for passing from the construction of the instances to the return of the phases from the BIM process.

"Italian Way" of the Industrial Heritage
As we have already mentioned [18], in 1980, Renato Covino traced in the journal "Quaderni storici" [19], a first balance of the pioneering years of industrial archaeology in Italy. With two exhibitions in 1978, industrial archaeology had landed in Italy, and the attitude of British scholars was accompanied by "an awareness of Britain's progressive loss of a leading position among industrialized countries and therefore a desire to preserve the signs of a glorious past, which constituted a powerful factor of national identity" [19]. All this took place while a process of diffusion of industrial archaeology was underway. In the space of a decade, it spread to Mediterranean European countries, with particular vitality in France, Italy, Spain, and Portugal [19][20][21][22].
In Italy, the starting point is a greater sensitivity towards social phenomena, and the country's specific cultural background "meant that the debate on its disciplinary field, its chronological limits, its methods and its aims was [ . . . ] very lively and often saw divergent positions being confronted" [19]. Just on the social studies front, Aldo Castellano in 1977 notes, that in Italy, already at the beginning, the limitedness of the analyses dedicated to the industrial monument from the point of view of social history or technology is cautiously approached, making the study of the monument "scientifically unproductive". He therefore proposes that industrial archaeology should lead to the "reconstruction of the history of civilization and industrial culture through material documents" [23].
Overall, at the basis of the birth and development of industrial archaeology in Italy, there is a constitutive multidisciplinary nature, as evidenced by the different disciplinary fields that have been interested in it since its origin, from the historical-architectural and technological disciplines to the historical-economic and social ones.
In Italy, as in France and in Central Europe, the processes of mechanization, the complexity of work organization, lead to an extension of the boundaries of interest, not only of the contexts. Eugenio Battisti [24] draws the attention towards the conquests of the Italian Renaissance, both in the genius, the machines, and in the constitution of the capital, of the connected financial system, and in the organization of the job, with Franco Borsi alongside him [14]. If we abandon the theoretical discussion and move among the research related to the understanding, protection, and enhancement of individual sites, we find ourselves operating in a reduced, regional, and local context.
In this context, we can happily record how it is easier to succeed the rigorous experience of the theory, also because we directly experience the collaboration between several disciplines with their own methods. As Covino writes, "the local dimension becomes an element of guarantee of control over the research, whose result can easily be put in circuit and communicated through the associative structures operating in the territory" [19].
As we will see later, the industrial archaeological presences of Tivoli, in the ancient structures of the Sanctuary, are an exemplary case study to experiment with tools of investigation and representation aimed at an integrated and complex concept of industrial archaeological heritage, aware of the long-term paths that over half a millennium have affected the industrial evolution: machine, organization, and energy.

BIM and Palimpsests
The use of technology and the BIM process applied to the built heritage is now well established, although this still represents a challenge, both in research and in design practice. The concept of HBIM (Historic BIM) has been declined in different forms starting from its introduction by Murphy [25], going from time to time to detail the specificities of the process adapting it to the single matter. The HBIM methodology has been applied over the years to different types of built heritage: churches, monuments, open spaces, and abandoned industrial sites [26].
Some new acronyms are thus defined, such as AHBIM in the case of BIM application to architectural heritage [27], or the concept of ArchaeoBIM in the case of BIM developments for archeology [28]. In the first case, Brusaporci et al., developed a workflow purely focused on the stratigraphic analysis of the masonry for the architectural heritage, investigating the construction history of the building through the identification of the chronological phases, integrating direct observations of the material structure with knowledge construction techniques and the study of documentary sources [27]; in the second case, Garagnani et al. instead addressed the specificity of the BIM application to archaeology, focusing the workflow on information aspects on the one hand, and on virtual reconstructive modeling on the other [28,29]. This last approach is strongly based on the aspects of cognitive investigation, both material and documentary sources, thus inserting itself in the HBIM research, but being devoid of constructed contexts to be surveyed in their entirety, instead establishing the foundations for a new method of experimental archaeology that allows for a virtuous cycle of control of all the steps: "from the starting data used to formulate the model itself, up to the simulation and post-figuration, that is to the virtual conjectural restitution of a building that actually existed in the past" [28].
The problems of modeling, both informative and geometric, of the built heritage are obviously linked both to the complex system of historical-constructive knowledge, on the basis of manuals and documentary sources, and so not representative for typical software libraries [30], and by the constructional specificities and the evolution over time of the building object, such as, for example, the irregularities in the realization and the manifested degradation. These two specific areas of HBIM pose the greatest challenge of using the BIM tools and process, wherein the standardization logic typical of contemporary construction adapts with some problems to the unicum represented by the individual buildings of the built heritage [31,32].
The topic of information modeling plays a fundamental role when it comes to the built heritage, and even more so in archaeology applications. The information necessary to manage constructive hypotheses and virtual reconstructions can take different forms, can require a greater amount of space, and comes from extremely heterogeneous sources (i.e., reports, drawings, previous surveys, material surveys) [27]. Therefore, the databases currently available on the market do not always adapt optimally to the objectives set at the basis of the HBIM process, and several experiments are under development constituting a central topic of debate.
In general, the shared approach is to implement the information within the model, distinguishing two main research lines: through the development of an ad hoc database, with or without specific interfaces, but which is better suited to the case and the information that they must be questioned [33], or through the semantic implementation of parametric objects enriched by heterogeneous information and the relationships that bind the objects themselves, typical of the definitive building knowledge management (BKM) approach where the central node is that of ontological modeling [34,35]. These approaches represent the basis for the application of the BIM process to the specificity of industrial archaeology. It can be said, paraphrasing the title of a recent volume, that the center of interest of ArchaeoBIM is simultaneously the primary material records and the lost heritage, and this challenge coincides exactly with the research of industrial archaeology [29].
Within the research on HBIM, there are, therefore, numerous current research groups [36] that are interested in the organization of knowledge and ontological modeling [34,35], as well as in the information modeling of data management through interactions multidisciplinary that integrate digital sources and data into a broader BIM process [30,[37][38][39][40] whose primary purpose is to guarantee an applied knowledge and collaborative design/management tool. The case presented in this paper falls within the second line of applied research. The research is part of the context, already foreshadowed for many years, for the digital visualization of cultural heritage, with the aims of research, management, and fruition, as highlighted in the London Charter [41] and updated in the Seville Principles [42]. In particular, the need to use the complex digital tool arises from the various factors of inaccessibility, intricacy of the different phases, and disappearance of many components of the heritage of different periods, with the effective opportunities that the digital can offer a simultaneous diachronic vision of even overlapping phases.
Given the challenge of disentangling the virtuous uses of digital representation of cultural heritage from those that are redundant if not ineffective, we are convinced of the preeminence of the second principle, Aims and Methods: "A computer-based visualization method should normally be used only when it is the most appropriate available method for that purpose" [41]. First of all, it was evaluated as to whether the use of digital tools was the most appropriate for the purposes of research and study related to the presence of a complex palimpsest of lost heritage, as well as for the general objective, in the context of a multi-value site, of being able to extract the specificities of the archaeological-industrial evidence and values.

Main Objective: An ArchaeoBIM for Industrial Archaeology
This paper takes up the challenge of declining ArchaeoBIM, of which it captures experiences of application to the legacies of antiquity, for industrial archaeology objects and systems. The issue of defining information models for the study, investigation and recovery of built heritage, is a very current and rapidly developing line of research. Within what has been presented in the previous section, it is necessary to focus the attention on a particular type of built heritage, the industrial one.
In the course of 30 years of studies, it has become evident that even in Italy, the centrality of industrial archaeology is the anthropology of a recent past. However, one has to deal with the numerous disciplines that are interested in it, in the historical, economic, productive, technological, architectural, and sanitary fields. The research shows how digital tools allow for the organization of such layered data to increase the possibility of more cognitive synthesis for the various disciplines involved.
Although, from the point of view of the systematic approach to the generation of informative models, the applications on cultural heritage could share various aspects (i.e., multiscalarity, integrated surveys, scan-to-BIM process, and the automation of processes semantic recognition of geometric elements), there are others instead strictly characterizing the industrial heritage: possible decontaminating the area from toxic materials; use of industrial elements and/or experimental technologies; and industrial production nature reflected both in function layout and spatial form, then in the presence of machines.
We do not dwell in this paper in the discussion of known problems on the definition of informative models for cultural heritage in general, now a subject of daily debate in research, but as regards the common themes, we certainly need to mention the relevant possibility of using technologies (i.e., UAV) that allow for the convergence of large distances, usually characterizing industrial sites, as well as the possibility of accessing unsafe or partially collapsed areas, as often happens with this type of heritage [43,44].
Analyzing instead the points that characterize the industrial heritage, we need to start from the first point, concerning the contaminated areas. Biagini et al. [45] consider the need to decontaminate the area from polluting materials (i.e., toxic materials and oils, possible unexpected "burials" of hazardous substances) as one of the most recognizable aspects to be managed through the BIM process. This aspect is fundamental in reducing the impact of the industrial site on neighboring areas during the redevelopment processes, as well as for the reuse of the asset itself, since the products treated when the industry was in operation (i.e., machine oils or other polluting liquids) may have spilled into the ground, causing possible risks for future users and for the environment (i.e., groundwater pollution, difficulties in growing new trees). The 4D and 5D management can realize full control over the shape of the terrain and the quantity of material, so as to allow the possibility of planning and controlling the removal of small or large quantities of them, correctly integrating these operations into the design [45].
The second characterizing point, concerning the construction area, represents a critical aspect of dual significance. While on the one hand the use of industrialized components goes well with the construction logic of BIM models and the standardization of the components, on the other hand, these constructions are often characterized by their experimental techniques, for which it is complex to find exhaustive information.
Given the high degree of industrialization of the components for industrial building, modular elements are often used, some prefabricated, and this can represent an advantage in the modeling and definition of parametric elements. Much of the heritage of industrial archaeology is made up of elements in iron or ghisa, concrete, stone, and brick, sometimes decorated, for which it is necessary to adapt the modeling of ad hoc parametric elements. Among these, it is certainly worth mentioning the reinforced concrete structures, often built with experimental technologies to cover the large spaces of the factory, and that today often show advanced signs of decay [45].
The third aspect concerns the spatial configuration of these buildings, strictly interrelated with the production layout. The definition of information models is a fundamental tool for understanding this aspect through the possibility of investigating on the one hand the geometric complexity and on the other a reconstruction of the production process that has defined its shape and space.
Function and shape are two important factors in the evaluation of this type of architectural space, given that the factory is an environment created for industrial production and therefore intrinsically linked to the productive activity that took place within it [33]. Furthermore, if the machines and materials of the production are still, at least partially, present on site, the survey can encounter significant difficulties linked to large, occluded areas that cannot be measured. These complexities can be overcome with a careful study of the "industrial" nature of the building, wherein the regularity in the composition of the spaces or the understanding of the production processes can integrate the measurements for the modeling of the building elements [45].
With regard to the second and third aspects, it should be immediately highlighted that the challenge of the present research becomes somewhat more arduous as it is confronted with spatial configurations and building systems representative of that category of antiquities for reuse illustrated in the introduction. As will be seen in the discussion, this makes it possible to highlight central themes and characteristics for an ArchaeoBIM of industrial heritage. Next to the basic instances of the industrial heritage, there are in fact the instances of the lost heritage of the sites within which it falls, as well as those of the many phases of life in the centuries or millennia of anthropization, abandonment, and reuse.
Finally, in all cases, the fourth aspect, central to all industrial heritage, comes into play: machines. The possibility of creating AS-BIM (As Built) models must necessarily include the implementation of the information relating to the machinery that characterized the industrial process of those places, defining their architecture. The machines are not always kept inside the places, and thus in this approach, the management of the fourth dimension of BIM becomes fundamental, allowing for the defining of the evolutionary phases of the building and therefore of the presence of the machines up to their decommissioning, degradation, partial ruin, or complete absence. After these specifications are defined, the industrial heritage can therefore be evaluated as a special field of experimentation for HBIM applications, capable of passing from degradation to a resource or, if neglected, of being lost forever [45]. In this aspect, it is possible to read a parallelism with the application of ArchaeoBIM, in the purposes of digital and informative reconstruction of an aspect, in this case, the industrial process, which has been lost over time.
In virtue of the opportunities offered by the collaborative digital platform, the research has been focused on the second, third, and fourth aspects of the industrial heritage, a deepening opportunity methodologically allowed by the environment in which we are working that enables each operator to implement the model according to his expertise.

Tivoli, a Prominent Site of Industrial Heritage in a Context of Millenary Archaeological and Artistic Presences
The reputation of the ancient city of Tivoli goes beyond the narrow circle of archaeologists, art historians, or architects interested in the dense historical and urban events of a history longer than 3000 years. In fact, it is from an immemorial time that Tivoli has a reputation for its antiquities and for the quality of its landscape. The masters of the Renaissance definitively sanctioned the primacy of the monuments and the landscape context, investigating, portraying, and emulating its imposing ruins. Thanks to them and to those who followed in the following centuries, the city became a favorite destination of the Grand Tour of the scions of the British, French, and mittle-European ruling class, bringing out, alongside the specialized interest, a relevant tourist cultural vocation.
Over the years, in the complex patrimonial palimpsest, next to Villa Adriana, which represents and mythologizes the places of the empire for the delights of the emperor or under which it reached its maximum extension, new sites of incredible interest emerge. Among all, we mention again two villas: Villa d'Este, an undisputed prototype of the Renaissance villas, and the restoration of otium and the Villa Gregoriana, a poignant combination of the horrid and the sublime made possible by a majestic work of hydraulic engineering through the Gregorian ducts, which, diverting the river Aniene to protect the city from flooding, generated the picturesque waterfalls that provided a fundamental theme to the painters of Vedutism and the art of the romantic garden.
The common thread, which also emerges from monumental studies of archaeology and more specifically of ancient topography, is certainly water. Tivoli rises on the cliff that the river Aniene has modeled over the millennia and from which in the past it found a whirling flow towards the plain below. This jump was the object of wealth and ruin for the city. Wealth has been an intense productive exploitation, at least since the late Middle Ages, which over time has seen the establishment of paper mills, ironworks, and tanning, up to the production of hydroelectricity, with some exceptional records, such as the first power plant, the Acquoria power plant, which in 1892 began to provide electricity for the lighting of Rome at 30 km away. Therefore, we refer to specific studies on the city and its archaeological heritage [46], while the present contribution focuses precisely on the archaeo-industrial theme [7] and on a case in which classical and industrial archaeology intersect. It is very complex to draw an overall picture of the manufactures such as tanneries, felting, ironworks, and paper mills-particularly numerous-mentioned in different sources that do not always agree. Initially, they drew their impulse and benefit from the network of pipelines that had already been dug under the city in Roman times and from the imposing jump that the Aniene river makes at Tivoli. Starting from the 16th century, this network thickens and complicates, announcing the remarkable hydraulic constructions of the 19th century and the records of energy exploitation of the 20th century ( Figure 1).  There are many testimonies attesting to the leather tanning, the oil mills, the wool wagons, or the production of paper [47], already in the 15th century, and which also report some proprietary events, such as those of G. M Zappi (1519-1596) reported by Pacifici [48]. The report of G. B. Fontana of 1590 testifies that the past floods of the Tiber had torn up a dam that provided at least 65 factories including grain mills, oil mills, ironworks, and paper mills [49]. There are many testimonies attesting to the leather tanning, the oil mills, the wool wagons, or the production of paper [47], already in the 15th century, and which also report some proprietary events, such as those of G. M Zappi (1519-1596) reported by Pacifici [48]. The report of G. B. Fontana of 1590 testifies that the past floods of the Tiber had torn up a dam that provided at least 65 factories including grain mills, oil mills, ironworks, and paper mills [49].
The size of the production was craftsmanlike, but often assisted by mechanics already refined by the end of the 15th century. For example, Scavia reports, with regard to paper mills, that in the 17th century, "most [ . . . ] had only one or two vats with very few workers and a daily production of paper limited to a few dozen kilograms; a maximum of 4500 sheets for each vat, corresponding to little more than 50 kg" [50].
At the beginning of the 19th century, among the areas corresponding to the present-day Latium, in the Liri Valley, there was a circularity of applications, figures, and innovations with French and northern European experiences. Also in Tivoli, during the 19th century, an entrepreneurial and technological leap took place, thanks certainly to the availability of energy, which was added to that of water quality [51], but also to the activation of entrepreneurial synergies with more mature industrial basins such as those of Terni and of the municipalities along the Liri river.
The site of the Sanctuary of Hercules is particularly representative of this path, because there one will find several plants from the Napoleonic era to the present day, for iron; for paper; and, with great impact, for the production of electricity.

Case Study of the Paper Mill Segrè Papermill in the Sanctuary of Hercules: Antiquity from Reuse
Because of its position and configuration, the Sanctuary of Hercules Victorious was one of the areas most affected by industrialization in Tivoli. As we will observe, many industrial structures, hydroelectric power plants, paper mills, and even an ironworks from the Napoleonic era, found their place in the ruins of the complex, a Roman Sanctuary, a depositary of a long history.
Its birth is placed at the turn of the second and first centuries BC, just outside the fortifications of the city, to accommodate travelers of the Via Tiburtina Valeria. The model for the sanctuary and for the monumentalization of many sites of worship of the Italic area are the great structures of the Hellenistic period of the Greek world. Therefore, large structural elements of modeling and containment of the land, such as substructures and cryptoporticus, give rise to a system that seeks a strong rationality in the disposition of the functions in an imposingly scenic architecture [52].
The large new structures are linked to the territory, to its geography, and are built on consolidated streets up to encompass, as in the case of Tivoli, with entire sections through hollow structures, necessary due to the steep terrain, where it was of fundamental importance the use of opus caementicium applied by Roman workers.
In the imperial age, there was the maximum urban expansion of the city that saw the fabric-building reach and surrounding of the sanctuary. They realized the covered market, the aqueducts, the thermal baths, the Amphitheater, and numerous villae of the Roman patriciate, as Villa Adriana. Villas fed by the four aqueducts directed towards Rome are the Anio Vetus, Anio Novus, Marcia, and Claudia [53].
The Sanctuary was already affected by the presence of two water channels, the Forma, for the supply to the sanctuary itself, as well as the Brizio. To these, in the second half of the 16th century, the enormous quantity of water coming out of the monumental fountains of Villa d'Este was added, about 800 L/s [54].
At the turn of the 1600s, these profitable supplies induced the Apostolic Chamber to begin the productive exploitation, with a foundry and an arms factory. Since then, the industrial history of the site is rich in events and enterprises. Among the most significant subsequent uses is the one that Luciano Bonaparte, Napoleon's brother, made of it. A promoter of numerous industrial initiatives, he built in the early 19th century a metallurgical plant that included a smelting furnace, two ironworks, a shovel factory, and a nail factory. All this was possible thanks to the large amount of water that flowed in the sanctuary, then called Villa di Mecenate, at the time owned by the same Bonaparte [55,56].
In the second part of the 19th century, particularly invasive interventions were those made for the passage of water pipes, not only for the supply of factories, but especially for that of the penstock of hydroelectric power plants that over the years were built downstream of the jump of the Sanctuary. A first work was the passage of the hydraulic conduit named after the engineer Raffaele Canevari in 1886. As illustrated by F. Cairoli Giuliani, the duct exploited the various cavities already present in the archaeological complex and the possibility of earth movements in order not to excessively disturb the present structures [54].
The basin of decantation equipped with grid and sluice was almost certainly in the center of the sacred area, on the structure of the podium of the ancient temple. From here, the canal continues crosswise in a northerly direction, feeding the turbines of a paper mill and a pasta factory with a 1.6 m pipe, until it reaches a spillway inside the north portico. Then, after having served these factories, the water was conveyed from the spillway into the penstocks of the Mecenate jump, in a drop tower called Canevari Turret [57] (Figure 2). The second hydroelectric phase is connected to the strengthening of the exploitation through the realization of the new Acquoria power plant. It involved the installation of several new components of the plant, including the Mecenate loading basin, built in 1901 and decommissioned in 1993 at the expense of ENEL; the penstocks from the mouth of the basin to the hydroelectric plant; and the piezometric well, still visible among the industrial ruins. A 1.3 m diameter metal conduit branched off from the basin and served two separate branches, the first for cooling the rotating machines of the power plant, passing through the Pantanella settling tank, and the second feeding the power plant [58]. downstream of the jump of the Sanctuary. A first work was the passage of the hydraulic conduit named after the engineer Raffaele Canevari in 1886. As illustrated by F. Cairoli Giuliani, the duct exploited the various cavities already present in the archaeological complex and the possibility of earth movements in order not to excessively disturb the present structures [54]. The basin of decantation equipped with grid and sluice was almost certainly in the center of the sacred area, on the structure of the podium of the ancient temple. From here, the canal continues crosswise in a northerly direction, feeding the turbines of a paper mill and a pasta factory with a 1.6 m pipe, until it reaches a spillway inside the north portico. Then, after having served these factories, the water was conveyed from the spillway into the penstocks of the Mecenate jump, in a drop tower called Canevari Turret [57] ( Figure  2). The second hydroelectric phase is connected to the strengthening of the exploitation through the realization of the new Acquoria power plant. It involved the installation of several new components of the plant, including the Mecenate loading basin, built in 1901 and decommissioned in 1993 at the expense of ENEL; the penstocks from the mouth of the basin to the hydroelectric plant; and the piezometric well, still visible among the industrial ruins. A 1.3 m diameter metal conduit branched off from the basin and served two separate branches, the first for cooling the rotating machines of the power plant, passing through the Pantanella settling tank, and the second feeding the power plant [58].   The 20th century was the century of the most impressive industrialization of the area; in addition to the power plants, the entire sanctuary was used internally and expanded with numerous structures in masonry and reinforced concrete to accommodate a very large paper mill the Segrè paper mill ( Figure 3). Thus, we reach the definition of what is now a vast archaeological area within which the sanctuary descends. The 20th century was the century of the most impressive industrialization of the area; in addition to the power plants, the entire sanctuary was used internally and expanded with numerous structures in masonry and reinforced concrete to accommodate a very large paper mill the Segrè paper mill ( Figure 3). Thus, we reach the definition of what is now a vast archaeological area within which the sanctuary descends.

General Questions from the Specific Case: Antiquity from Reuse
In the area of the Sanctuary, there is a complex of historical and artistic evidence of great value, enhanced by the use of archaeological museum. All this is intersected by the industrial history. We intend to evaluate the opportunities of the application of an archaeological computer modeling with specific attention to the layers of industrial heritage superimposed on the classical and medieval archaeological heritage. It is a relevant challenge because after the paper mill decommissioning, many industrial structures have been demolished, favoring, understandably, the structures of the ancient sanctuary. However, this condition, dictated in many cases by the need for static safety of the artifact, is not fully acceptable if it stems only from a desire for indiscriminate cancellation of industrial memory. In fact, in the industrial memories of Tiburtina, there are a series of fundamental values, technological primates, linked to the electrification of the city; identity characters, linked to the history of paper; and workers and entrepreneurs of the city. Last but not least, the factor of post-industrial landscape definition cannot be neglected, having assumed several elements intersected with the sanctuary, such as the Canevari tower, the industrial sheds, and the structures of the former pasta factory Pantanella.
It is with these objectives that the case study is developed in BIM, an industrial BIM, which, for the reasons of antiquity of the sites considered, on the one hand collects the developments of the H-BIM and considers the simultaneous collection and representation of multidisciplinary data, and, on the other hand, utilizes those of the Archaeo-BIM as a useful tool for analysis and communication of the archaeological lot heritage of any nature.

Materials and Methods
The methodology follows what is reported in Figure 4. The study of the industrial heritage necessarily starts from the constructive investigation, which is divided into archival research, analysis of the evolutionary phases of the building, architectural survey, and investment in materials. In this case, a detail aspect is related to the production machines, which have profoundly marked the formal and distribution aspects of the building organization. The study of the machines must necessarily follow the same investigation process of the building in its entirety, emphasizing the need for a multiscale approach.

General Questions from the Specific Case: Antiquity from Reuse
In the area of the Sanctuary, there is a complex of historical and artistic evidence of great value, enhanced by the use of archaeological museum. All this is intersected by the industrial history. We intend to evaluate the opportunities of the application of an archaeological computer modeling with specific attention to the layers of industrial heritage superimposed on the classical and medieval archaeological heritage. It is a relevant challenge because after the paper mill decommissioning, many industrial structures have been demolished, favoring, understandably, the structures of the ancient sanctuary. However, this condition, dictated in many cases by the need for static safety of the artifact, is not fully acceptable if it stems only from a desire for indiscriminate cancellation of industrial memory. In fact, in the industrial memories of Tiburtina, there are a series of fundamental values, technological primates, linked to the electrification of the city; identity characters, linked to the history of paper; and workers and entrepreneurs of the city. Last but not least, the factor of post-industrial landscape definition cannot be neglected, having assumed several elements intersected with the sanctuary, such as the Canevari tower, the industrial sheds, and the structures of the former pasta factory Pantanella.
It is with these objectives that the case study is developed in BIM, an industrial BIM, which, for the reasons of antiquity of the sites considered, on the one hand collects the developments of the H-BIM and considers the simultaneous collection and representation of multidisciplinary data, and, on the other hand, utilizes those of the Archaeo-BIM as a useful tool for analysis and communication of the archaeological lot heritage of any nature.

Materials and Methods
The methodology follows what is reported in Figure 4. The study of the industrial heritage necessarily starts from the constructive investigation, which is divided into archival research, analysis of the evolutionary phases of the building, architectural survey, and investment in materials. In this case, a detail aspect is related to the production machines, which have profoundly marked the formal and distribution aspects of the building organization. The study of the machines must necessarily follow the same investigation process of the building in its entirety, emphasizing the need for a multiscale approach. and the applications of the ArchaeoBIM concept. What is detailed in the following sections focuses on the specificities of this approach to industrial heritage: application of Archae-oBIM to industrial archaeology, in particular for the definition of a diachronic BIM model (Section 3.1); integrated survey to industrial sites (Section 3.2); HBIM reconstruction model (Section 3.3); and, finally, LOD definition for industrial heritage in the BIM process (Section 3.4).

Application of ArchaeoBIM to Industrial Archaeology
Within the acronym ArchaeoBIM, Garagnani [28] identifies BIM applications with archaeological contexts, a practice that in Italy began being tested in the early 2000s, on the basis of previous experiences in the Anglo-Saxon field. To date, there are numerous teams that are interested in informative modeling of archaeological evidence, more or less well preserved [37,59]. A key point in the application of the BIM process to the existing is certainly linked to the difficulty of transforming unique elements characterized by a high level of formal complexity into parametric objects. This has led to a technological refinement of the methods of acquiring three-dimensional data, as well as to an in-depth semantic analysis of the structure of the asset to be treated [59].
The characterization of the BIM process applied to archaeology, compared to a general existing heritage, is linked to the fact that the reality to be represented is not defined and concluded, but is incomplete and subject to interpretation and constructive hypotheses. Several researches to date have focused on the difficulty of parametric modeling The procedure adopted is based on the typical operations of HBIM for the existing heritage, wherein the fil rouge of the characterization on the specificity of the industrial heritage, highlighted in the scheme, is emphasized by the productive function and the centrality of the machines. The main operations after the constructive investigation therefore concern the digital survey, the construction of the informative model, its validation, and the applications of the ArchaeoBIM concept. What is detailed in the following sections focuses on the specificities of this approach to industrial heritage: application of ArchaeoBIM to industrial archaeology, in particular for the definition of a diachronic BIM model (Section 3.1); integrated survey to industrial sites (Section 3.2); HBIM reconstruction model (Section 3.3); and, finally, LOD definition for industrial heritage in the BIM process (Section 3.4).

Application of ArchaeoBIM to Industrial Archaeology
Within the acronym ArchaeoBIM, Garagnani [28] identifies BIM applications with archaeological contexts, a practice that in Italy began being tested in the early 2000s, on the basis of previous experiences in the Anglo-Saxon field. To date, there are numerous teams that are interested in informative modeling of archaeological evidence, more or less well preserved [37,59]. A key point in the application of the BIM process to the existing is certainly linked to the difficulty of transforming unique elements characterized by a high level of formal complexity into parametric objects. This has led to a technological refinement of the methods of acquiring three-dimensional data, as well as to an in-depth semantic analysis of the structure of the asset to be treated [59].
The characterization of the BIM process applied to archaeology, compared to a general existing heritage, is linked to the fact that the reality to be represented is not defined and concluded, but is incomplete and subject to interpretation and constructive hypotheses. Several researches to date have focused on the difficulty of parametric modeling applied to archaeological evidence, the creation of ad hoc families and libraries, accompanied by all the information necessary for the definition of the model and its correct use [37,59]. However, an equally fundamental theme is that of investigating the study of the evolutionary phases of the constructed object and therefore of its digital representation.
As defined by Garagnani [28], the BIM model of an archaeological site provides, digitally, the image of a context crossed by numerous historical phases, fragmentary and, often, characterized by completely hypothesized portions.
It is here that one of the great potentials of the BIM process can be applied, through the definition of the fourth dimension (4D), that is, the temporal one. The same argument is valid in the case of industrial archaeology, and in this paper, we highlight the potential of this use applied to the complex of the Segrè papermill. This is processed through the management of the phases within the BIM Authoring Autodesk Revit-normally distinct between existing, demolition, and new-declined instead in the more complex evolutionary phases of the industrial heritage of the case study in order to create a diachronic BIM. Each object, architectural volume, and building component is therefore prepared for the implementation of specific documents, and above all for the definition of construction time phases.
This reasoning will be analyzed in particular for an element that strongly characterizes industrial archaeology, that is, the machines of production. However, what was stated by Garagnani, in the definition of the ArchaeoBIM process, must be understood and reiterated: "We must be aware that the process is possible, but only to varying degrees of likelihood. Therefore, it must be stated right now that a fundamental point of this new method is not the mere evocation of the past, but the validation of the digital result through a replicable process based on the collection of all available information" [37].

From the Integrated Survey to Photogrammetric Reconstruction (Landscape to Building)
Deformations and irregularities of the case study have been reported with an integrated digital survey with multi-source and multi-resolution 3D imaging techniques [32]. For this reason, structure from motion (SfM) based on two image-based modelling (IBM) methods were used to recover 3D information in form of dense point cloud from images [60]. Notably, the close-range terrestrial photo acquisition with Sony A7R has been supplemented to low-altitude UAV (unmanned aerial vehicle) photos with DJI Phantom 4. The use of the UAV allowed for the obviation of the difficulty of reaching the parts of the Segrè Papermill, which are difficult to approach and detect-specifically, the indoor spaces not secured (e.g., the paper machine hall in Figure 3) due to decayed ceiling systems and the roofs themselves, visible only from above.
Consequently, a first close-range terrestrial photographic acquisition campaign was conducted over all accessible indoor and outdoor spaces, both uncovered and covered (e.g., porches). In addition, two different UAV type of flights were carried out to complement the Sony A7R image set. The first flight was carried out in manual mode with GPS stabilization for the purpose of obtaining nadiral and oblique photographs of the exterior space, including the portion of the landscape between Tivoli, the Aniene river valley, and the remains of Pastificio Pantanella, including the temple of Hercules.
Notably, a video of the landscape was acquired to integrate its frames into the photogrammetric project in order to reduce grey areas [61]. Figure 5 shows the camera poses of close-range, low-altitude UAV photo acquisition through videogrammetry. Due to the high slope and spatial layout of this landscape, the flight altitude was not set constant. The UAV was either moved closer or further away from the acquired object, depending on its spatial complexity, trying to guarantee forward and side overlaps between two consecutive images of 80% and 60%, respectively [62]. This allowed for a better reconstruction of the photogrammetric point cloud and ensured an acquisition with a constant ground sample distance. Moreover, the second type of flight was also carried out in manual mode, this time without GPS stabilization, as it was conducted in the interior spaces. Indeed, for reasons related to the danger of collapse of parts of the roofs, the property explicitly requested that certain parts of the complex were surveyed avoiding any direct access. In these interior spaces, it was also necessary to ensure a proper forward and side overlap between two consecutive images. As the distance between the drone and the interior spaces was greater than the distance between the drone and the outer shell of the "Segrè Papermill", it was necessary to acquire much more photos.
A total of 1961 photographs with different points of view and scales were selected to carry out the photogrammetric project and obtain a dense point cloud of the case study. The SfM software used for the photogrammetric reconstruction was Agisoft Metashape version 1.5.1.

BIM Reconstruction
The richness of the data acquired was treated with attention, as it varied in reliability and accuracy [63]. The SfM outputs are unstructured points clouds, far from being informed models, a goal sought from the morphological and semantic point of view, to incorporate all meta-information [64]. Both methodological fields, therefore, require an important intervention of external editing, exemplified in the case of study, in which the operator must supervise automations both in the cleaning of the raw cloud data [65], to extract only the relevant architectural object, and in the reduction that allows passage from the generic points to the evidence of significant points for the three-dimensional metric description [66]. Then, the cloud was subsampled to have a uniform distribution of 3D points and was cleaned of outliers with the application of statistical outlier remover (SOR) in CloudCompare 2.11.3 [67]. In addition, automatic point cloud classification tools were applied [68] using trained classifiers to eliminate all those elements that could hinder the correct comprehension, e.g., vegetation and site works. Hence, the generated point cloud was imported to the modelling environment of Autodesk Revit, after the format conversion from .e57 to .rcs performed in Autodesk Recap. Thus, the point cloud was used Moreover, the second type of flight was also carried out in manual mode, this time without GPS stabilization, as it was conducted in the interior spaces. Indeed, for reasons related to the danger of collapse of parts of the roofs, the property explicitly requested that certain parts of the complex were surveyed avoiding any direct access. In these interior spaces, it was also necessary to ensure a proper forward and side overlap between two consecutive images. As the distance between the drone and the interior spaces was greater than the distance between the drone and the outer shell of the "Segrè Papermill", it was necessary to acquire much more photos.
A total of 1961 photographs with different points of view and scales were selected to carry out the photogrammetric project and obtain a dense point cloud of the case study. The SfM software used for the photogrammetric reconstruction was Agisoft Metashape version 1.5.1.

BIM Reconstruction
The richness of the data acquired was treated with attention, as it varied in reliability and accuracy [63]. The SfM outputs are unstructured points clouds, far from being informed models, a goal sought from the morphological and semantic point of view, to incorporate all meta-information [64]. Both methodological fields, therefore, require an important intervention of external editing, exemplified in the case of study, in which the operator must supervise automations both in the cleaning of the raw cloud data [65], to extract only the relevant architectural object, and in the reduction that allows passage from the generic points to the evidence of significant points for the three-dimensional metric description [66]. Then, the cloud was subsampled to have a uniform distribution of 3D points and was cleaned of outliers with the application of statistical outlier remover (SOR) in CloudCompare 2.11.3 [67]. In addition, automatic point cloud classification tools were applied [68] using trained classifiers to eliminate all those elements that could hinder the correct comprehension, e.g., vegetation and site works. Hence, the generated point cloud was imported to the modelling environment of Autodesk Revit, after the format conversion from .e57 to .rcs performed in Autodesk Recap. Thus, the point cloud was used as a scaffolding for the setting of the HBIM model of Segrè Papermill, integrating the architectural survey and the digitalization of the various evolutionary phases, where there were deficiencies due to the collapses and degrade of the case study.
One of the major challenges in using BIM for the documentation of architectural heritage in general, and notably industrial archaeological sites, is overcoming the propensity of BIM Authoring software towards standardization. In this context, HBIM is exploited for its high potential in the organization and management of data, whether historical, geometric, or related to the construction engineering. The uniqueness of the existing assets, along with the pre-industrialized logic behind them, required the use of BIM families structured to ensure that significant geometric details, construction characteristics (materials, construction technology), and the phenomena of degradation and deformation are embedded in the model.

LOD Definition for Industrial Heritage in BIM Process
The information modeling aspects of the BIM process are entrusted to the definition of the LOD (level of detail). The quantitative explanation of this concept has taken different forms to date depending on the territorial context of application [69].
In the United Kingdom, the definition of LOD was made explicit by the PAS 1192-2 directive, settling on a numerical scale ranging from 1 to 6. In the USA, the American Institute of Architects (AIA) has published the document G202-2013 Project Building Information Modeling Protocol, wherein the concept of LOD (in this case the acronym stands for Level of Development) is defined on a numerical scale expressed in hundreds, from 100 to 500. In Italy, UNI 11337 in parts 5 and 6 defines the concept of LOD on an alphabetical scale (from letter A to letter G). Furthermore, the most recent UNI EN ISO 19650-1 introduced the concept of LOIN (level of information need), which should gradually replace that of LOD, but it is still under development and debate [70]. Although discordant in formalization, the various definitions agree in the use of progressively more detailed stage scales, and in the subdivision of two concepts that converge in the LOD-the geometric component (in Italy defined LOG, level of geometry) and the information component (in Italy defined LOI, level of information). The peculiarity of the Italian scale of definition of the LOD lies in the last two levels, which are particularly sensitive to the intervention on the built heritage, recovery, and restoration [59]. Indeed, level F defines objects that correspond to realities actually measured during site surveys (as-built), while level G refers to the current state of the components, combining virtualization and updating of the actual state of the places. The complexity of the information transposition in level G for recovery projects lies above all in the definition of the degradation of the elements.
Physical, chemical, superficial, or structural alterations represent important information to be filed in the definition of the LOD.
To these issues, the issue of information reliability and the degree of fidelity of the digital model to reality must be added. These two concepts converge in the definition of the level of reliability, fundamental for HBIM applications and currently the subject of constant scientific debate [71][72][73]. The issue becomes even more complex when the applications are made on the archaeological heritage, where the models are the subject of reconstructive hypotheses of a partially or totally lost reality.
The applications of the BIM process to industrial heritage share the same principle with those of archaeological heritage, in particular for production machines that are often in a state of severe degradation or are even completely absent in architectures that were mainly designed to house them. We are thus faced with a task that must necessarily take into consideration the definition of the LOD for MEP (mechanical, electrical, and plumbing) applications, already complex for the new building [74], as well as the applications to a non-more existing heritage, such as the archaeological one. The methodological proposal presented here focuses on the definition of the LOD for level G as the current representation of the current state of production machines in industrial archaeology, as well as for level F as a reconstruction of the last phase of their useful life. In defining these aspects, the reference of the machine description blueprints as archival study material is of fundamental importance, as well as the possibility of linking the blueprint itself to the informative model.

Results and Discussion
The digital survey presented here allows for the obtaining of greater data density to document specific paper mill instances and covers the limits of different reality capture systems in ranging from landscape to building scale. Consequently, even the most inaccessible indoor space was reached. Hence, the most unsafe areas such as the paper machine hall and part of the rags cutting room have been inspected by UAV using either existing openings from the inside or external openings with no windows frame.
The results are presented below with respect to: (Section 4.1) application of Archaeo-BIM to industrial archaeology, (Section 4.2) configuration and representation of the current state from the integrated survey to BIM reconstruction ( Figure 6), (Section 4.3) current and lost machines, the use of LOD definition for industrial heritage in the BIM process. level F as a reconstruction of the last phase of their useful life. In defining these aspects, the reference of the machine description blueprints as archival study material is of fundamental importance, as well as the possibility of linking the blueprint itself to the informative model.

Results and Discussion
The digital survey presented here allows for the obtaining of greater data density to document specific paper mill instances and covers the limits of different reality capture systems in ranging from landscape to building scale. Consequently, even the most inaccessible indoor space was reached. Hence, the most unsafe areas such as the paper machine hall and part of the rags cutting room have been inspected by UAV using either existing openings from the inside or external openings with no windows frame.
The results are presented below with respect to: (Section 4.1) application of Archae-oBIM to industrial archaeology, (Section 4.2) configuration and representation of the current state from the integrated survey to BIM reconstruction ( Figure 6), (Section 4.3) current and lost machines, the use of LOD definition for industrial heritage in the BIM process.

Application of ArchaeoBIM to Industrial Archaeology
In the application of the BIM process to industrial archaeology, the fourth dimension of BIM, namely, the temporal one, plays a fundamental role. The numerous phases that followed one another in the life of the building organism are described through the use of the phases of the BIM authoring chosen for the modeling. Through the management of the project phases, eight phases were created and applied to all instances in the BIM model ( Figure 7). This made it possible to expand the ArchaeoBIM feature to the industrial heritage, and in particular to the case study, creating the diachronic BIM of the Segrè Papermill. In particular, three preliminary evolutionary phases were identified relating to previous uses-four main phases of evolution of the Cartiera and a final phase relating to

Application of ArchaeoBIM to Industrial Archaeology
In the application of the BIM process to industrial archaeology, the fourth dimension of BIM, namely, the temporal one, plays a fundamental role. The numerous phases that followed one another in the life of the building organism are described through the use of the phases of the BIM authoring chosen for the modeling. Through the management of the project phases, eight phases were created and applied to all instances in the BIM model (Figure 7). This made it possible to expand the ArchaeoBIM feature to the industrial heritage, and in particular to the case study, creating the diachronic BIM of the Segrè Papermill. In particular, three preliminary evolutionary phases were identified relating to previous uses-four main phases of evolution of the Cartiera and a final phase relating to its current use as a Museum of the Sanctuary of Hercules, for a total of eight phases described in the model (Figure 8).
arms factory and perhaps the first paper mills. They are represented in their documented state in 1840 by the gunpowder mill and the iron foundry of the Villa di Mecenate (indicated as the second phase). Between these two phases, there is a large period of time concerning the medieval and modern phases of use, for which the model is ready to collect the contributions of medieval and modern archaeology and topography scholars. The third phase indicates the one that starts the most intensive industrial development, with the installation of the hydroelectric plant and the start-up of the paper mill (1884).  Starting from the first historical phase that can be found, namely, that of the II-I century BC of the Sanctuary of Hercules, it moves directly to the early stages of the industrial age, which originated, as has been written, starting from 1600. There was a foundry, an arms factory and perhaps the first paper mills. They are represented in their documented state in 1840 by the gunpowder mill and the iron foundry of the Villa di Mecenate (indicated as the second phase). Between these two phases, there is a large period of time concerning the medieval and modern phases of use, for which the model is ready to collect the contributions of medieval and modern archaeology and topography scholars. The third phase indicates the one that starts the most intensive industrial development, with the installation of the hydroelectric plant and the start-up of the paper mill (1884).
The Società Forze Idrauliche per gli usi civili ed industriali (Hydraulic Forces Company for civil and industrial uses) acquires the gunpowder mill from Nerli and the iron foundry from Vannutelli.
From there began the construction of the canal and the jump inside the Canevari tower at around 1870. The power plant was built on the north front, and in 1889, it channeled the Mecenate waterfalls, which thus, after centuries, disappeared from the city panorama. The power plant was dismantled in 1899 due to the need for expansion.
Parallel to those plants, Segrè Papermill takes shape, the four phases of which have been defined in detail in the diachronic BIM model. This is the last industrial reuse of the sanctuary, built in part on the previous manufacturing structures, mainly occupying the area between the podium and the northern portico with iron and reinforced concrete structures that, however invasive, have contributed to the preservation of part of the arcades. A second expansion was placed in the previous Porta Scura iron foundry where the cycle was enlarged with another paper machine. The fully automated industrial plant was first served by hydroelectric turbines and subsequently by an electrical substation.
The first detailed phase of the paper mill is that of the configuration assumed in 1920, wherein the construction of the paper machine room is highlighted (which took place in 1915), as well as the north and south expansion of the shed for the preparation of paper with English trusses (around 1915).  In the 1920s, the paper mill therefore presented the plant, which it maintained until the end with three continuous machines, located, respectively, in the sacred area, in the substructures at the level of Via Teca, and in the square in front of the Antiquarium. The second detailed phase of the paper mill (1930) concerns the raising and expansion of the room for the paper machine, which is equipped with a new roof supported by metal trusses, as well as the construction of new volumes containing the deposits for wastepaper. The third phase (1940) concerns several changes that occurred in the 1930s to the existing volumes for the introduction of the widespread use of reinforced concrete structures and prefabricated frames for large roofs. This phase includes, among the various volumetric changes, the construction of new volumes, the refurbishment of the roofs, and the expansion of the "stracceria" (cotton rags cutting room) (1936)(1937)(1938), new two-story warehouses for storage and the choice of paper (1939), and for the paper storage (1935), the construction of the new electrical substation (1936)(1937), and the construction of a retaining wall for the new storage area of the pulp manufacturing area (1937). The layout of the site is therefore significantly changed. The fourth and last phase of detail (1955) highlights the changes that occurred following the war damage, wherein the wooden truss roofs were replaced with a reinforced concrete structure and two-barrel vaults with SAP technology. Some existing structures were also raised.
The Segrè Papermill was decommissioned in 1956, and the complex was passed to the state property. Today, the sanctuary is partially reused as a museum (Museo del Santuario di Ercole Vincitore). Thanks to the demolitions carried out since the 1990s, evidently marked by the elimination of volumes of poor architectural value, the complex is now a unique combination of ancient and industrial archaeology.

Configuration and Representation of the Current State: From the Integrated Survey to BIM Reconstruction
The Segrè Papermill is modelled in BIM using an object-oriented approach. Data are represented in specific family instances and can have specific attributes. The stratification during the centuries is returned in the model, integrating the industrial heritage (e.g., related to the phases of use as a paper mill) with the classical heritage (e.g., related to the phase of the Sanctuary of Hercules). Notably, the modeling process aimed at deepening the restitution of the archaeological heritage, both classical and modern. The building components unfold over time, and each phase has its own specificities.
Furthermore, a different kind of detail is required to geometrically describe the classical archaeological heritage compared to the industrial one.
On the one hand, the classical heritage lacks industrialization process of components and standardization of the elements. Consequently, it was necessary to solve the problems deriving from the application of the modules used in antiquity with discrete, unique, and unrepeatable elements (e.g., variable thickness walls, different moldings depending on the architectural element). Moreover, the richness of the data coming from the point cloud has been critically interpreted through the detailed study of manuals and treatises on historical construction in order not to confuse geometric and constructive information with the component degradation. In this case, the objective of the study was not aimed at the restoration of the complex but at the comprehension of constructive, spatial, and temporal features. Therefore, a general geometric schematization with an association to a specific historic phase of every single element was preferred to documenting the degradation state (e.g., moldings and classical elements). Thus, the model leads to BIM use for time phasing simulation for the investigation of the complex stratification and interconnections between the various historical periods (Figure 8). A more careful study regarding the association of informative and documentary aspects regarding the deterioration state has been done for specific objects of interest, which have required in-depth research ( §4.3). This strategy was chosen on the basis of the objective of the study. The risk of element over modeling would have led to the creation of heavy models, greatly increasing the computational time.
On the other hand, the modern industrial archaeological heritage modeling is better suited to BIM logic than classical heritage modeling, as exemplified in Figure 9, where the BIM families are used to represent the different kind of roofing systems (i.e., Howe truss, Polonceau truss, reinforced concrete slab, SAP vaults, and extractor hood).
tween the various historical periods (Figure 8). A more careful study regarding the association of informative and documentary aspects regarding the deterioration state has been done for specific objects of interest, which have required in-depth research ( §4.3). This strategy was chosen on the basis of the objective of the study. The risk of element over modeling would have led to the creation of heavy models, greatly increasing the computational time.
On the other hand, the modern industrial archaeological heritage modeling is better suited to BIM logic than classical heritage modeling, as exemplified in Figure 9, where the BIM families are used to represent the different kind of roofing systems (i.e., Howe truss, Polonceau truss, reinforced concrete slab, SAP vaults, and extractor hood).  The sub-systems of the building have been detailed, returning tie rods, struts, beams, and frames that the integrated survey has allowed to detail. The integration of these industrialized components led to the observation that the benefits experienced during the construction of the paper mill (e.g., reduced construction time) are parallel to those related to the optimization of the model delivery time. In fact, BIM employment allowed for the processing of a model by assembling parameterized BIM components in a much more efficient and less time-consuming way. Notably, C1 and C2 in Figure 9 show how the BIM model allows for the elaboration, study, and description of the engineering solution of the modern and classical archaeology intersection exemplified by the Howe and Polonceau truss and the Roman portico. Consequently, the model exemplifies the potential of BIM to handle multiple and complex phases. In addition, information from historical and archival documents has increased the potential of BIM as an information repository for material properties and stratigraphy ( Figure 10). efficient and less time-consuming way. Notably, C1 and C2 in Figure 9 show how the BIM model allows for the elaboration, study, and description of the engineering solution of the modern and classical archaeology intersection exemplified by the Howe and Polonceau truss and the Roman portico. Consequently, the model exemplifies the potential of BIM to handle multiple and complex phases. In addition, information from historical and archival documents has increased the potential of BIM as an information repository for material properties and stratigraphy ( Figure 10). Figure 10. Axonometry of case study with a focus on the reinforced concrete slab with the information retrieved by historical and archival documents (A.D.P. and R.R.). Figure 10. Axonometry of case study with a focus on the reinforced concrete slab with the information retrieved by historical and archival documents (A.D.P. and R.R.). In addition, it was possible to integrate into the BIM model the particular topographic conformation of the portion of the landscape among Tivoli, the Aniene river valley, and the remains of Pastificio Pantanella, including the temple of Hercules. Actually, the ground part of the point cloud has been converted into a topographic surface for the present phase of the former Segrè Papermill. This has allowed for the elaboration and the study the model through volumetric and landscape considerations as a whole. Moreover, synthetic axonometric elaborations extracted from the BIM model allow for the communication of the current state of the complex and wondering about its refurbishment.
The results and strategies actually fostered the promotion of the HBIM model of the Cartiera to a landscape scale [75] and could allow for binding to the concept of HLIM (heritage landscape information model) [76]. As previously mentioned, the geometric schematic solution has been preferred for the description of the classical archaeological heritage, and necessary simplifications have been made for the representation of industrial archaeology components. Consequently, the need to describe the reliability of the BIM model arises. Specifically, the HBIM model generated has to be validated by measuring the distances between the parametric objects and the point cloud. Hence, Autodesk Point Layout was used to compute the distance between the model and the point cloud directly in the Revit environment. Although there is no criterion specifically shared in the scientific literature regarding the acceptable distance, the choice of an average value oscillating between −0.05 m and +0.05 m, with a standard deviation less than 0.10 m, is being consolidated [62]. Figure 11 shows the analysis performed on the roman vaults part, and there is a good correspondence between the model and the cloud. Medium-high values of deviation are located in specific structural parts, probably corresponding to an advanced strain pattern.

Current and Lost Machines: Use of LOD Definition for Industrial Heritage in BIM Process
Following the LOD scheme introduced in UNI 11,300 and explained in the methodological part (Section 3.4), we drew up a standard form for the definition of the LOD increase for production machines. The example shown ( Figure 12) is related to the Hollander beaters and is divided from LOD A to LOD G. LOD A is characterized by a two-dimensional element of the overall dimensions, defining a rough positioning of the machinery. LOD B constructs a simple extrusion of the volume, assumed on the previous level of detail. The LOD C begins to characterize the

Current and Lost Machines: Use of LOD Definition for Industrial Heritage in BIM Process
Following the LOD scheme introduced in UNI 11,300 and explained in the methodological part (Section 3.4), we drew up a standard form for the definition of the LOD increase for production machines. The example shown ( Figure 12) is related to the Hollander beaters and is divided from LOD A to LOD G.
use of the fourth dimension on the elements of production machines. The understanding and study of their placement must necessarily start from archival documents and archaeological evidence, that is, from the remains present within the building. In the case of Segrè Papermill, a complete mapping of the possible original locations of the machines was elaborated ( Figure 13). The figure shows the location of the machines in their latest configuration. As we have already written [77], the mechanization and industrialization of paper production saw the appearance of some machines that required the predisposition of special building configurations.  LOD A is characterized by a two-dimensional element of the overall dimensions, defining a rough positioning of the machinery. LOD B constructs a simple extrusion of the volume, assumed on the previous level of detail. The LOD C begins to characterize the shapes of the machine, describing the essential elements: the cylinder and the hollow tank. The LOD D defines the Hollander beater in the right conformation and describes in more detail the beater wheel, blades, and sinkers. The LOD E then goes on to detail the exact type theoretically used, in this case, XI A, which inserts further and is defined by mechanical elements, as well as the informative properties of production capacity and necessary driving force. The most significant LODs for recovery interventions are represented by LOD F and LOD G. The information aspects of the single instance actually installed will be defined in the LOD F, wherein the information is recoverable, in this case, for example, the installation date of 16 September 1934, and the manufacturer, i.e., Officine Carcano Maslianico, with an evident parallelism to what defined for an as-built. In the end, the LOD G provides detailed information on the state of degradation of the machine, on the missing components, and on its state of decommissioning, also preserving the informative, but not geometric, aspects of the previous LOD. A specific discussion should also be made for the use of the fourth dimension on the elements of production machines. The understanding and study of their placement must necessarily start from archival documents and archaeological evidence, that is, from the remains present within the building. In the case of Segrè Papermill, a complete mapping of the possible original locations of the machines was elaborated ( Figure 13). The figure shows the location of the machines in their latest configuration. As we have already written [77], the mechanization and industrialization of paper production saw the appearance of some machines that required the predisposition of special building configurations. It is clear that in the spaces of the Sanctuary's substructures and in those specially made in the following phases, the machines have a direct dimensional and functional relationship. In the oldest paper mills mechanization took place through the pestle mallet, introduced by the masters of Fabriano in the 15th century and which had to be set up parallel to the power water duct, but we no longer find them in modern paper mills [8].
On the other hand, Hollander beaters are still present in production, in Tivoli as well as in the working paper mills. They have been fundamental in the production of paper since the 17th century and are organized according to the positioning of the drive shaft and the transmission of motive power. The paper machines, in use in Italy since their first installation in 1826 in the Lefevbre paper mill [78], require a room with a large longitudinal development, the cardinal room of every modern paper mill. Boilers and chimneys are then the basic elements of the configuration of every modern factory.
The virtual reconstruction of the production plant is intrinsically linked to the volumetric evolution of the building organism. In the case study presented, a reconstruction of the hypothetical configuration of the last operating phase of the paper mill was therefore carried out (the one defined in 1955 in Figure 8). The instances of the machines have been modeled according to the LOD scheme shown in Figure 12. It is possible, by phase filters, to create representations of the last phase of production by highlighting in red the aspect that the machines they could have had, such as a hypothetical reconstruction of the archaeological heritage ( Figure 14). It is clear that in the spaces of the Sanctuary's substructures and in those specially made in the following phases, the machines have a direct dimensional and functional relationship. In the oldest paper mills mechanization took place through the pestle mallet, introduced by the masters of Fabriano in the 15th century and which had to be set up parallel to the power water duct, but we no longer find them in modern paper mills [8].
On the other hand, Hollander beaters are still present in production, in Tivoli as well as in the working paper mills. They have been fundamental in the production of paper since the 17th century and are organized according to the positioning of the drive shaft and the transmission of motive power. The paper machines, in use in Italy since their first installation in 1826 in the Lefevbre paper mill [78], require a room with a large longitudinal development, the cardinal room of every modern paper mill. Boilers and chimneys are then the basic elements of the configuration of every modern factory.
The virtual reconstruction of the production plant is intrinsically linked to the volumetric evolution of the building organism. In the case study presented, a reconstruction of the hypothetical configuration of the last operating phase of the paper mill was therefore carried out (the one defined in 1955 in Figure 8). The instances of the machines have been modeled according to the LOD scheme shown in Figure 12. It is possible, by phase filters, to create representations of the last phase of production by highlighting in red the aspect that the machines they could have had, such as a hypothetical reconstruction of the archaeological heritage ( Figure 14). BIM modelling in the heritage context enables automatic reconnaissance of the spatial arrangements of various historical phases. This is accompanied by the possibility, thanks to the spatial and temporal referencing of the machines, as we have seen, to use the results for coherent fruition bending between the real museum and the virtual one.
Virtual fruition is a fact [79] that has been spreading for a long time for many sites, and nowadays the possibilities are being explored for their impact on the whole territory [80]. Furthermore, with respect to the fourth dimension, the HBIM so structured is constitutively a model that includes the 4D and therefore it is possible to extract, at the desired LOD and with the desired mesh, an environment within which to make the virtual experience. The tool is interesting in this sense because it is possible to automatically activate the evidence and the superposition of the phases through key parameters. BIM modelling in the heritage context enables automatic reconnaissance of the spatial arrangements of various historical phases. This is accompanied by the possibility, thanks to the spatial and temporal referencing of the machines, as we have seen, to use the results for coherent fruition bending between the real museum and the virtual one.
Virtual fruition is a fact [79] that has been spreading for a long time for many sites, and nowadays the possibilities are being explored for their impact on the whole territory [80]. Furthermore, with respect to the fourth dimension, the HBIM so structured is constitutively a model that includes the 4D and therefore it is possible to extract, at the desired LOD and with the desired mesh, an environment within which to make the virtual experience. The tool is interesting in this sense because it is possible to automatically activate the evidence and the superposition of the phases through key parameters.

Conclusions
It is believed that the paper offers a specific direction for the study and informed representation of industrial antiquities heritage. It has been seen to be a complex field, and it is believed that from what has been developed, it should be approached by grasping both the directions drawn by the proponents of Heritage BIM as well as the developers of ArchaeoBIM. The problems of modeling, both informative and geometric, of the industrial heritage have been addressed, on the basis of archival sources, techniques on direct survey, responding to two main aspects.
On the one hand, there is the diachronic representation of the evolution of production processes, and on the other hand, the representation and evolution of the building object in a context of continuous superimpositions and replacements, a intricated example of lost heritage. The logic of standardization on which BIM was born may represent an obstacle and not well interpreted by the operator. However, from the point of view of Industrial Heritage BIM, one can see the extraordinary potential, through widespread mapping and the definition of successive cases over time, to be able to build a shared library of the experiments of building solutions, the preparation of machines, and the definition of processes that have occurred over time. In this way, it is possible to create an exceptional repository for the study of industrial heritage, both to capture the exceptional elements of each case and to read the areas of diffusion of machines, procedures, and knowledge.
From the point of view of communication, moreover, the industrial-BIM/VR application would allow for an incredible increase of virtual fruition, both in the archaeological area, for the parts that are not accessible, and at a distance, in order to allow a wide tourist and cultural attendance of the industrial heritage also from remote. In this sense, VR seems to be the new "Atlas" like the one used by Julius Verne in his adventurous travels around the world, from the confines of his study as a writer of great adventures where he conceived the famous Voyages extraordinaires.
As in the case of Verne, what is decisive is the thirst for knowledge, an interest in data that comes from many sources and a point of view with which acquired knowledge is proposed, and, in the case of a AR based on a BIM, there is the sensible possibility of a flexibility of the point of view and of the degree of deepening that makes it a tool that could bring results both in research and in the construction of a shared cultural conscience of industrial heritage.