Databases and Information Models for Semantic and Evolutionary Analysis in Fortified Cultural Heritage

Abstract

The paper explores two fortified systems, as case studies, to evaluate different modelling approaches, the parameterisation of architectural components, and semantic interpretation, to define a repeatable methodology for classifying, accessing, and sharing architectural information. The use of informational structures to categorise data and relations through semantic attributes addresses the need to make data queryable and interoperable. In architectural documentation, this practice opens up new perspectives for creating and sharing collaborative repositories, radically transforming the way of disseminating and accessing knowledge. Linking qualitative data information with 3D models enables the development of semantic structures that provide a deeper understanding of intrinsic characteristics, historical transformations of architectural complexes, and their cultural context. While databases effectively manage structured information and relationships between different signifiers, semantically enriched 3D modelling, through the management of surfaces or parameters, offers an accurate and flexible representation of each component. The evaluation of these case studies not only improves the ability to understand and appreciate built heritage but also provides new opportunities for researchers in architectural documentation and history. Semantic subdivision processes of built complexes into individual components facilitate the analysis and the digital preservation of heritage, enabling more precise interpretations and faithful reconstructions.

1. Introduction

The use of semantically enriched 3D models is increasingly recognised as an effective tool for cultural heritage documentation, aiding in the study of architectural and technological features, as well as in representing its transformations and the evolutionary processes undergone over time [1,2,3]. Today, 3D data are a crucial component in recording the shape of architectural objects and sites so that, at least digitally, they can be preserved and handed down to future generations. In this regard, strengthened digitisation technologies, through the acquisition of raw data, provide adequate support for structuring 3D databases and models, serving as an essential knowledge base for encoding information [4]. Although these technologies already sufficiently meet the requirements for accuracy and comprehensiveness, they still fall short of addressing the cognitive challenges associated with architectural representation [5,6,7,8].

The development of models, starting from documentation and survey practices, aims to define metrically reliable forms. Especially with digital technologies, measurement, even at an infinitesimal scale, becomes a subject of investigation. Yet, there is often an excessive emphasis on pursuing highly accurate metric surveys, overlooking the essential focus on the critical analysis of form, composition, and study of architectural works [9,10].

The critical analysis of the architectural model should be based on descriptive aspects and considerations on shapes, relationships, and signs [11,12]. Only later, at a more advanced cognitive stage, should confirmation of the various hypotheses be sought through measurement. In this sense, there is a desire to return to more efficient surveying methodologies that align with the timeframes required for interpreting, studying, and analysing the technological and compositional characteristics of the architectural artefact.

Another aspect to consider concerns the overabundance of information from digital surveys, which often exceeds the actual requirements for documentation and valorisation. This informational surplus derived from data acquisition may pose a challenge to data interpretation, potentially compromising the clarity of the message and the information intended to be conveyed through digital representation. To ensure effective communication, it is therefore essential to undertake a data refinement and discretisation process aligned with the predefined objectives. Moreover, analysing the qualitative information obtained from interpreting documentary sources is crucial. Consequently, innovative technologies are needed to support researchers in processing, organising, and analysing the data to achieve critical representations of architectural artefacts [13,14,15,16,17].

From these reflections emerges a broad spectrum of considerations, encompassing the relationship between individual elements and the architectural complex, the methodologies of geometric representation, the temporal evolution of buildings, and the management and visualisation of information [18,19,20,21]. These closely interrelated issues pose epistemological and operational challenges that must be addressed to maximise the effectiveness of new digital techniques (Figure 1).

Reality-based 3D modelling has shown considerable efficacy in capturing the morphological complexity of architectural heritage [22,23,24,25]. However, geometry alone is not sufficient to adequately represent the architectural complexity of a building, which also includes constructive and technological as well as functional and symbolic aspects. Geometric complexity represents only a part of the communicative element, which will have to integrate abstract and signifying realities, which in their wholeness constitute a fixed system of three-dimensional relationships, existing between significant “objects” [26].

Additionally, the evolution of architectural structures over time provides an additional layer of complexity. Buildings have undergone transformations over the centuries, both due to deterioration processes and restoration or reuse interventions. Integrating these temporal transformations within a 3D model requires tools and methodologies capable of dynamically representing the changes a building complex has undergone during its history. Considering these aspects, a critical issue that remains to be addressed by architectural heritage researchers is the visualisation of all relevant information within a single representation. The integrated visualisation of data not only facilitates the documentation of a building’s current state but also enables a deeper understanding of its historical evolution [27,28,29].

2. Materials and Methods

This paper outlines the results of research studies conducted on fortified systems that addressed issues related to the integrated application of fast-survey methodologies, the development of information models, and the development of visualisation and management systems. Through the analysis of two case studies—the Montorio Castle in Verona (Italy) and the Almonecir Castle in the Valencian Community (Spain)—the present work focuses on methodologies for structuring geometric components, developing historical reconstructions and applications for their digital representation (Figure 2).

Starting from the case studies, the aim is to apply these approaches to gain an in-depth understanding of the relationships between the architectural object, its three-dimensional representation, significant features, and the evolutionary temporal path.

Tracing a fortress’ evolutionary process, by reconstructing its defensive elements’ characteristics, requires a holistic approach that includes understanding the various historical phases and the different defensive structures employed over time.

This section outlines the methods employed for data collection and how the morphological, historical, technological, and other types of information were used and structured into 3D systems concerning the complexities of the case studies, which are comparable in their state of preservation and historical richness, yet culturally distinct [30,31] (Figure 3). A relevant aspect is the possibility of creating a digital repository capable of organising, archiving, and analysing a heterogeneous information set [32,33,34].

2.1. Acquisition Methodologies Through Integrated Fast-Survey Techniques

Structuring three-dimensional representations related to the different historical and construction phases of the case studies required the establishment of a methodological protocol that would allow obtaining cognitive databases on which to build models of the architectural objects under examination [35,36]. Additionally, to keep the workflow towards the development of 3D models, the field operations were organised to streamline acquisition times by setting up efficient digital documentation strategies. The fast-survey solutions for the multi-scale documentation of historical and cultural heritage allowed us to address complex operational contexts, often marked by challenges such as limited accessibility, time constraints, or the large areas that need examination [37,38]. In this context, the now traditional data acquisition procedures, which involve using terrestrial laser scanners frequently integrated with other measurement systems, often require an extensively conducted on-site data collection phase.

However, these methods can reasonably be replaced by methodologies that foster rapid information collection and organisation, allowing for a broader cognitive phase conducted directly in the field [39,40,41].

In the specific context of Almonecir Castle, a methodological digitisation process was tested using mobile laser scanning (MLS), integrated with UAV and DSLR for Structure from Motion (SfM) photogrammetry processes. The applied methodology, which relies on tools with varying resolutions and reliability, was developed to represent the case study at different scales of detail, both architectural and territorial [42,43]. In the case of Montorio Castle, the application of Structure from Motion (SfM) photogrammetry methodologies was supported by acquiring direct measurements.

An aspect to consider when developing a fast-survey documentation campaign is the level of detail and reliability required for the 3D database. However, this aspect must be addressed in relation to the purpose of the research and the goals to be achieved. If the model’s overall goal is to convey the historical evolution of the building or the technology behind its construction, then morphological accuracy becomes more important than metric accuracy.

2.2. Three-Dimensional Modelling Processes and Structuring of Evolutionary Stages

Raw data obtained from the survey campaigns were processed and integrated for the subsequent 3D modelling phase, finalising operations not only for understanding architectural features but also for the temporal representation of elements within the entire architectural complex.

This allowed for a discussion of the relationships between acquisition times, the extent to which architectural forms can be modelled, how elements can be semantically subdivided, and what type of information can be linked.

A critical variation between the methodological approaches followed in the two case studies concerns the characteristics of the site and the type of modelling employed. Depending on the type and reliability of the sources used for the reconstructions, this resulted in a different granularity of semantic decomposition.

Concerning the differences in the details of the reconstructions, two distinct modelling techniques were applied to the two case studies, summarised as follows:

-

The hybrid modelling of evolutionary phases (mesh-nurbs), in which a macro-semantic subdivision was structured.

-

The parametric modelling of evolutionary hypotheses (HBIM), where a micro-semantic breakdown was developed.

The semantic subdivision of elements was organised according to relational criteria by combining various levels, and consequent sub-levels, according to the complexity of each entity. The elements were categorised by technological/typological function and architectural styles associated with different historical periods.

To organise the information and connections resulting from this semantic structure, it is necessary to adopt systems, processes, and coding rules based on a consistently structured set of names and descriptions [44]. The coding procedure of the elements takes the following factors into account:

-

The shapes, geometries, and relative level of detail of the corresponding architectural component.

-

The building type to which the element belongs and its function.

-

The identification of each element’s position within the overall architectural complex.

-

The historical phase of construction and/or affiliation.

The alphanumeric code for each architectural element was defined by associating a character with the following parameters: (i) element name; (ii) type and function; (iii) position and sequential numbering for serial elements; and (iv) historical construction phase. This procedure ensures each element receives a unique code, considering form, position, and numerical differentiation for similar or serial elements [45].

Besides different modelling methodologies using two types of software, what differentiates the case studies is precisely how qualitative information is linked and visualised. Reconstruction methods and choices made in semantic breakdown modes can be summarised as follows:

-

In the first case study, reconstruction is addressed on a macroscopic level, divided by structures (towers, ramparts, ravelin, etc.) and starting from historical source analysis, which provides a certain level of reliability. The systematised information and documentary apparatus were associated only after the model was structured.

-

In the second case, reconstruction is treated more in detail, requiring the structuring of thematic layers to highlight, through colour scales, aspects related to reliability. In this instance, data were associated during the modelling phase.

2.3. Montorio Castle: Macro-Reconstruction of Evolutionary Phases Through Historical Sources

For the Montorio Castle case study, the aim was to create an interactive virtual tour that would facilitate an understanding of the defensive structure’s elements and their evolution over the centuries. The methodology followed for developing an interactive 3D model, therefore, entailed the need to optimise time and cost of the acquisition campaign as well as to obtain reliable databases—both in their current state and with reconstructions of historical phases—that could be easily managed on online platforms (Figure 4).

The employed approach integrates geometric/semantic modelling practices capable of dialoguing the complexities of the architectural object, combining raw data with the semantic information that emerges from the interpretative processes. The procedure consisted of the following steps:

-

Collecting and analysing documentary sources and iconographic material for extracting information on artefact consistency (geometric form, surface appearance, and physical characteristics) and the interpretation of remains;

-

Correlation between the data used in the reconstruction process and the level of uncertainty in each constituent element;

-

Automated 3D model reconstruction of the castle’s actual state;

-

Freeform modelling for historical phases, adapting basic geometric primitives to represent architectural features for the semantic enrichment of models;

-

The validation of evolutionary reconstructions and the subsequent development and design of a platform for its visualisation.

In the analysis process of the castle’s evolutionary phases, three main stages were considered: the current state, the 19th-century historical phase corresponding to Habsburg rule, and the 15th-century Scaliger phase (Figure 5). To represent the castle’s current image, a textured mesh model was used, generated from an automated reconstruction process through photogrammetric techniques. This approach aimed to speed up the modelling of the existing state, allowing us to focus on the reconstruction of historical phases.

The possibility of modelling the evolutionary phases on a realistic base also enables clear and direct visualisation of the differences between the existing structure and the reconstruction. Starting from the geometric model, which corresponds to the morphological representation of the architectural object in its current state, the idea was to structure a model, through theoretical interpretation, that goes beyond the visible form and can evoke a set of knowledge and information connected to the semantic construction of the architectural element. The transition from simple representation to theoretical interpretation requires human intervention to select, filter, and interpret the data collected, retaining only those elements that enrich the conceptual model. This synthesis process is essential for creating meaningful information aligned with the project’s analytical goals.

Beyond analytical considerations, it is crucial to reflect on the gap between direct representation and the theoretical model and how these can be integrated into a unified representation system [46]. Various types of information were used to create the historical reconstructions, ranging from considering the castle’s remains to utilising documentary and iconographic material related to the two historical periods under examination. Thus, the construction of the past states’ geometric models was based on comparing the two historical states represented in iconographies (2D data) and the 3D representation of the current state.

Three main modelling operations were carried out to structure the geometric entities based on the polygonal model of the existing state. Through the operations of definition, editing, and removal, the geometry of the architectural entities was reconstructed according to their temporal distribution and with the transformation events [47,48]. The use of the specified modelling operations can be illustrated as follows:

-

Definition: modelling of the entities that exist in the examined temporal state but are absent in the current or past states;

-

Editing: the geometric entities already present within an evolutionary state were adapted to the state under examination through splitting, merging, or deforming the geometric entities based on the visual appearance in the analysed source;

-

Removal: in the transition from modelling one phase to another, all entities not present in the examined temporal state were deleted.

For structuring the historical phases’ geometric entities, the mesh model of the existing state was imported into the freeform surface modelling software McNeel Rhinoceros 6.0. Before modelling a single phase, structured layers were established for each historical period, along with sub-layers categorised by element type (to which an alphanumeric code was assigned). This created an organised structure facilitating semantic enrichment and information association.

2.3.1. Macro-Reconstruction of the Scaliger Phases

The first modelled reconstruction phase was the chronologically least recent, corresponding to the castle’s original state. Information obtained from field analyses of archaeological remains, supplemented by a 17th-century iconographic representation, were used to model the elements. The historical source, a pseudo-axonometric view, allowed for identifying some typological elements that are no longer present, such as the curtain towers and the buildings within the fortified perimeter [49]. It was also possible to identify a moat on the southeastern side of the walls, with a drawbridge leading to the castle entrance and a gate with a ravelin system, whose foundations remains are still visible. Similarly, for the northeastern wall section, now missing (demolished and replaced by earthworks used as defensive positions in the 19th century), archaeological evidence and iconographic representation enabled the reconstruction of the original appearance (Figure 6).

In the historical reconstruction, these elements and the original terrain configuration were represented by removing and reconstructing the photogrammetric model’s surrounding terrain portion by creating a surface using control points. The volumetric reconstruction of the curtain towers also drew upon typological elements from same-age towers found in other castles in the Verona Region. Elements within the courtyard, including the church and houses illustrated in the iconographic representation, were modelled to enrich the overall image of the reconstruction.

2.3.2. Macro-Reconstruction of the Habsburg Phases

For the philological reconstruction of the Habsburg period, the iconographic sources used are from two surveys conducted on the castle by the Austrian military corps of engineers, dated 1859 and 1860. The documents include various two-dimensional drawings depicting modifications made by the Habsburg Empire to adapt the castle to the defensive systems of the time.

Starting from the reconstruction of the Scaliger period, removal and editing operations were performed on existing geometric entities. Then, based on a comparison with the iconography, all the entities not present in the previous historical phase were defined. In this phase, gun platforms were constructed within the boundary walls and in the northeastern portion of the fortress. The layout plan from archival sources was used as a baseline to integrate the new volumes, by scaling and placing them within the model (Figure 7).

During the modelling process, the position and orientation of the elements in the 2D representation were cross-checked and validated against interpretative analyses of the structure’s remains (Figure 8).

2.4. The Castle of Almonecir: Micro-Level Reconstruction Hypotheses Based on Evidence

The methodological approach employed for the Almonecir Castle involved applying scan-to-BIM processes to create a model in which all objects are interconnected with the entire complex through parameterising its architectural elements [50,51,52,53,54] (Figure 9).

HBIM methodologies facilitate an organised structuring and a detailed deconstruction of elements and sub-elements [55,56]. This approach enhances technological and constructional understanding and traces their evolution across different historical phases [57,58,59]. For the Almonecir Castle, the process relied on archaeological evidence as well as deductive and inductive operations regarding possible configurations to perform historical reconstructions since no iconographic material was available. The process consisted of the following steps:

-

Analysis of archaeological evidence and interpretation processes;

-

Analysing contemporary case studies and interpretative processes to define the main characteristics of the structure (geometric shape, surface appearance, and physical characteristics);

-

Correlation between the data used in the reconstruction process and the level of uncertainty characterising each constitutive element;

-

3D reconstruction of the phases (including the current state) through scan-to-BIM;

-

Validation of evolutionary hypotheses and subsequent design of a platform for visualisation.

The 3D database, which provides a starting base for the modelling phases, was optimised and exported into a compatible format with the BIM platform Autodesk Revit. This optimisation process was conducted within the Autodesk Recap Pro 2023 software, where the database underwent a cleaning process to enhance the readability of structures for the subsequent modelling phase. The point cloud, which was segmented and divided into specific regions (categorised by area of interest), was then exported in a format (.rcp) directly compatible and viewable within the BIM platform. The segmentation and independent reading of key areas improves data visualisation and management effectiveness (Figure 10).

The modelling of the castle was structured according to two criteria: on the one hand, a temporal division of construction phases, and on the other, a typological breakdown of individual elements at both the architectural and territorial levels. To this end, pre-set parametric families available within the software were used wherever possible; however, it was necessary to extend the default abacus with new project-specific elements for certain components.

Before proceeding with the typological modelling of the elements, based on historical considerations, work phases were set up within the project file. These phases correspond to the stages in which the evolutionary model of the castle was developed. Each phase was preliminarily configured using the software’s dedicated functions. There are a total of five phases defined through document analysis and interpretative processes (three of which are historical), and they were structured as follows:

-

The current state of the castle;

-

The castle before restoration and consolidation interventions;

-

The expansions of the castle and the Albacar (14th–15th century);

-

The castle during the time of Jaime I (12th–13th century);

-

The first fortified structure: the Arab ribat (9th–11th century).

Once the working environment was set up, the model components were developed based on the current image. Starting with modelling the components common to all phases, individual architectural elements were then reconstructed for the chronological representation.

2.4.1. Micro-Reconstruction of the Actual State of the Castle

As the initial step in modelling the current state, a topographic surface was structured, onto which the blocks of architectural structures were subsequently added. The terrain morphology was created by exporting the DTM (Digital Terrain Model) derived from photogrammetric processing, which involved the removal of anthropogenic elements. The optimised DTM was imported into the BIM software. Starting from the imported elevation data, the software automatically structured the terrain’s topographic surface. This surface retains the same coordinate reference system as the point cloud within the model space, which served as the basis for the scan-to-BIM of the castle’s current state.

Structuring the model through a BIM platform allowed and facilitated the semantic breakdown of the castle’s elements. The parametric modelling logic, using predefined families and components, enables an organised coding of individual elements used to represent the historical phases. From reading the point cloud data, the main elements were identified and isolated: walls, floors, vaults, roofs, openings, beam systems, stairs, railings, and complementary elements. Each one was associated with the following informational components: creation phase, demolition phase, geometrical parameters, and material characteristics. All model elements were renamed within the BIM software according to an established alphanumeric code, which, as mentioned, includes the element name, type and function of the corresponding element, position and sequential numbering for serial elements, and the historical construction phase.

The architectural components’ models were structured by modifying the parameters of the families available in the Revit library’s model, creating new families tailored explicitly to the case study. The in-place modelling methodology was considered appropriate to represent the irregular shape of the walls with variable cross-sections and non-linear floor plans, using a subtraction between model solids. For this purpose, “void” model components were generated through an in-place profile drawing, which allowed the extruded walls to be cut and shaped.

The workflows for modelling the main architectural components in the current state followed the specific stages listed below.

Management and Placement of Wall Families (Figure 11):

-

Definition and positioning of walls using the tracing tools available in the software;

-

Using the “Modify Wall” command, the irregular profile characteristics of the castle walls were traced in place;

-

The walls are equipped with accessory elements, such as putlog holes or openings, which are added by creating and positioning specially parameterised components.

Management and Placement of Wooden Floor Families (Figure 12):

-

Definition of properties and creation of in-place profiles for the slabs using the available tracing tools in the software;

-

Definition of properties and creation of in-place profiles for areas with continuity solutions in the floors meant to host vertical connections, given that they had diverse and challenging shapes to standardise;

-

Definition of the supporting structure of the timber floor through a beam system.

Management and Placement of Vaulted Floor Families:

-

Using the same method as wooden floors, properties were defined for each slab, followed by in-place modification of the extrusion profile using available tracing tools;

-

Creation of vaults as components through the definition of a local model. After studying the vault type, its shape and profiles were parameterised to allow model adaptation;

-

Placement of components within the slab. The model modification parameters were adjusted by controlling adherence to the point cloud data in place.

Management and Placement of Vertical Connections:

-

Placement process follows the same approach used for walls and floors. The definition and positioning were carried out from a parametric system family, adapting the object to the point cloud by controlling the ramp’s direction and length parameters;

-

Editing and defining accessory parameters such as the number and height of the treads, adapting the modelled stairs to those present on site;

-

Addition of the parametric components of the railing and handrail already available in the Revit library. Based on the point cloud data, parameters were set to define the overall dimensions of the balustrade components, such as the number and spacing of vertical elements and total height.

2.4.2. The Reconstruction of the Castle Phase Before the Consolidation Works

Starting from the model of the actual state, all reconstruction phases were defined. The first reconstruction modelled was the representation of the castle before the recovery and consolidation work. During the archival research activities, it was possible to retrieve the digital drawings made before the interventions. CAD vector files were then linked to the Revit model, and each was oriented and aligned with the side view. The in-place profiles of the existing walls were traced based on the linked drawings (Figure 13).

For regular walls without any complex shapes and with a vertical or oblique cross-section, the profiles were designed in place, using the linked drawings as a reference, by tracing and extruding the model components.

For walls with a variable cross-section and non-linear plan section, like the southern wall portion of the tower, the profile was created by subtracting solids using components modelled as in-place voids (Figure 14). By making the necessary approximations, it was possible to characterise and represent the ruined walls of the castle accurately. Defining this pre-restoration phase, which represents the castle’s unaltered form, enables preliminary comparative analyses with the current condition and its state of conservation (Figure 15).

2.4.3. The Reconstruction of Historical Phases and the Development of Reliability Scales

Once the parameters and association with the corresponding historical (or work) phase were defined, each element, linked to its respective parametric family, was defined, modified, or removed within its specific work phase using the point cloud as a morphometric reference and the software’s tracing tools. Through the processes of analysis and interpretation, all elements were modelled and reconstructed based on archaeological evidence and documentation. One challenge in the reconstruction process is establishing a system capable of representing the reliability of each reconstruction. To depict probability, ambiguity, reliability, or uncertainty in 3D reconstructions, the method employed involved the use of layers characterised by a colour scale [45,60]. This system allows for an immediate understanding of the uncertainty of the hypothetical reconstruction for each artefact. The colour associated defines the degree of uncertainty according to pre-established parameters, considering whether or not the reconstruction is supported by archaeological evidence. Therefore, each element of the reconstructed model is identified by its corresponding level of uncertainty, which visually assesses the appropriate level of knowledge related to the reconstruction process.

The colour scale defined for the case study ranges in percentages from 0 to 100, where the first value indicates total uncertainty and the second indicates total certainty. The scale is divided into 6 gradients that correspond to a percentage of reliability (100, 80, 60, 40, 20, 0) to which the following categories are associated, in order: (i) presence of remains; (ii) reconstruction based on existing remains; (iii) reconstruction based on pre-existence and typological study; (iv) reconstruction based on adjacent remains and deductive processes; (v) reconstruction based on typological study and deductive process; and (vi) total absence of information (Figure 16). Examining and representing the evolution of an artefact or a hypothetical reconstruction linked to a specific era requires the consideration of both temporal and documentary dimensions [2,21,28]. The temporal dimension enables the placement of specific information in space and time, highlighting its context within the study and its relation to the historical reference period. On the other hand, the documentary dimension clarifies and evaluates the methodology used to analyse the architectural transformations of the artefact based on comparisons among various types of heterogeneous information [43]. The 3D representation, viewed as a metaphor for a cognitive system applied to an architectural complex, does not depict the object but our understanding of it (Figure 17). At the core of this system lies a cognitive structure that functions as a graphical interpretation of the case study. The geometric representations of the different architectural components could serve as query tools, facilitating interaction between data and users and allowing them to explore the information element by element. Through this framework, supported by a semantic graphic code, the representation can achieve the following:

-

Highlight inconsistencies in the documentation or its interpretation;

-

Display the level of incompleteness in the study;

-

Providing insights based on knowledge gained about a predefined object.

3. Results

Through reverse modelling technologies, it was possible to synthesise data and create models capable of exploring the complexities of the analysed case studies. These models allow a three-dimensional view of artefacts undergoing significant stylistic and structural transformations and evolutions over the centuries. The building’s descriptive model serves not only as a 3D representation but can also become a common denominator for establishing bilateral relationships between the object’s morphology and heterogeneous information.

Based on the development of these models and in line with the research objectives, systems for access and visualisation were set up to make the models’ semantic and evolutionary characteristics visible explicitly. The definition of interoperable platforms addresses the need to make the developed models accessible. In this regard, the aim is to define information systems capable of displaying not only the geometric information associated with objects but also all the correlated materials and information related to the reconstructions. Two types of information systems were developed for the case studies analysed, tailored to the distinct research requirements and to facilitate comparisons and potential enhancements.

3.1. An Interrogable Platform for Online Information Retrieval

The development of a platform for the Castle of Montorio served a dual purpose: to make the material accessible and to create a database to associate three-dimensional shapes with information (Figure 18). While using freeform modelling software simplifies the procedure by not being too bound to rigid parametric constraints, this approach lacks direct information association, requiring additional software or programming languages [61,62]. To increase its communicative potential, the 3D model, previously semantically subdivided into macro-components, was enriched with qualitative information.

This process led to the creation of a web-based visualisation system of the castle. Users can interact with the geometric elements, access informative pop-ups detailing the defensive systems’ history, characteristics, and functionality, and visualise the castle’s temporal development.

Thanks to the one-to-one correspondence between information and morphological components, the platform offers two ways of interaction: selecting entities directly in the 3D scene or querying entities through a database.

The decision to structure a platform accessible online allows for easier dissemination and access, including hyperlinks and associated QR codes.

3.2. A Territorial-Scale Information Platform Integrating BIM-GIS Systems

In the case of the Castle of Almonecir, a georeferenced platform was structured to relate the castle to its territorial context. For this reason, the model was integrated and validated within GIS, allowing the architectural object to be linked with the fortified territorial system and its geo-morphological context, ensuring a deeper understanding of the adopted construction criteria and opening up the possibility of digitally connecting the castle with other fortified case studies in the area [63,64]. The information model, which was previously georeferenced using topographic coordinates, has been integrated into the three-dimensional GIS platform ESRI ArcGIS Pro.

By overlaying and designing multiple layers, it becomes possible to visualise its historical development within the territory, digitally and ideally interconnecting the site within the platform [65,66]. Phase management within GIS is ensured by structuring layers to which individual parametric objects are assigned and selecting and organising categories based on temporal development. Individual categories were exported as “features” and grouped in the corresponding layer, enabling the visualisation of specific phases and all their associated categories of elements. This methodological development enabled the migration of predefined information and parameters of BIM models into the native software format.

Structuring the model into temporally distinct reliability levels and elements indeed facilitated the re-association of previously defined information within the modelling software and the possibility of direct querying within the GIS system. For each element, it was possible to associate parameters such as the identification code, type, parametric family, and the information of relative position in relation to the complex (Figure 19).

This methodological approach creates digital data storage that can be configured as an implementable system over time. The platform allows for updating information about the building’s current conservation status and documenting its transformations over time, tracking the site’s historical development (Figure 20).

4. Discussion

The illustrated case studies have led to several considerations on the results obtained and the potential future research implementations. The first point to emphasise is the awareness and need to think about interoperability, focusing on developing dynamic communication and relational interfaces. In recent years, several research studies have focused on the compatibility of modelling systems, particularly BIM, with other management systems. [65,67]. Creating descriptive architectural models necessitates systems that can manage, utilise, and relate a vast amount of heterogeneous data. Consequently, the challenge lies in developing user-friendly systems that enable effective visualisation and use of this information.

The integrated HBIM-GIS system applied to Almonecir Castle represents a feasible solution for a better technological understanding of the architectural components, facilitating the efficient management of historical phases and interpretation on a territorial scale [64,68,69].

The use of BIM systems in structuring historical reconstructions enhances the management of technological components by providing a wealth of information that aids in understanding their characteristics. However, this approach faces challenges in compatibility with other utilisation systems, leaving many investigative issues open. The proposed solution of linking the information model to a GIS platform only partially addresses the issues of accessibility and data interrogation [70,71].

Another important aspect is related to the accessibility of the information system, which should be easy to access and use by scholars, researchers, and the general public interested in these topics. With the use of the ArcGIS Pro software, it is possible to publish a package of scene layers as a web-based scene layer, effectively structuring a web-based implementable platform. Reasoning in terms of organised and interconnected information systems remains a challenging approach to heritage valorisation, providing a solid foundation for understanding the architectural object and its cultural context.

In the case of Montorio Castle, using a free-form modelling approach to create a model accelerates the reconstruction process but complicates the development of an information management system. The implementation of an immersive web-based platform to visualise the Castle allows us to isolate individual elements and link them to descriptive documents. This system is particularly valuable for educational and dissemination purposes. In this context, incorporating semantic structures into a digital model helps clarify the relationships between the architectural object, its digital reconstruction, the relevant documentary sources, and its broader significance. However, the implemented system is still challenging, limited to this single case study, and lacks sufficient information and interrogation capabilities.

The developed approaches have also proven to be useful, especially in the case of Almonecir, for highlighting all limitations due to the lack of material, visualising the adopted yet undocumented reconstructive conjectures, and emphasising the most likely reconstructive solutions, thus faithfully reflecting the study and analysis process.

5. Conclusions

Integrating morphometric, temporal, technological, and historical data into a single representation requires advanced tools capable of combining these different types of information intelligibly. For this reason, it is necessary to develop new methodologies to enrich the semantics of three-dimensional representations, making them more accessible and usable across different fields of study. Semantically enriched 3D models facilitate the analysis and understanding of the complexities related to architectural objects (Figure 21). The multidimensional approach not only preserves the historical memory of buildings but also opens new avenues for research and access to cultural heritage, promoting informed usage and an immersive experience for all users. One of this research’s main contributions is highlighting how informative models can promote an accurate interpretation of fortified structures at various scales, from the architectural to territorial levels [72]. Through the structuring of information models enriched by a semantic structure, information can be organised at different temporal and spatial levels, providing a system for visualizing and analysing historical transformations. The implementation of these digital tools aligns with current research trends in this evolving field. Moreover, developing new approaches such as automated practices for segmenting raw data, recognition of geometries, and semi-automatic processes for parametric modelling could facilitate the efficient identification and management of formal and structural built heritage characteristics [73,74,75,76]. This approach would open new possibilities for large-scale study and classification of information, significantly expanding knowledge and analytical capabilities and presenting new challenges related to the quality and accuracy of outputs and margins of error in data interpretation.