Buildings
= Buildings =
The built environment of Venice reflects centuries of architectural evolution, adaptation, and shifting patterns of residential and economic activity. This page provides an overview of how Venetian buildings are classified, how they are used today, and what recent analyses reveal about their structure, height, residential capacity, and vacancy. It incorporates extensive measurements, geospatial analysis, and modeling to improve building-scale understanding across the historic center.
Building Types & Classification
=== Overview of Venetian Classification Systems ===
Venice classifies its buildings through a structured system that identifies architectural typology, construction period, and predominant functional characteristics. This classification helps describe when, how, and for what purpose each building was constructed, enabling a clearer understanding of the city’s urban fabric and its evolution.
For more details on individual building types, uses, and classification logic, see this page (https://wiki.cityknowledge.org/index.php/Building). This page provides comprehensive descriptions of Venetian building categories, examples of each type, and further technical documentation that complements the classification system summarized here.
=== Architectural Typology (Tipologia Edilizia) ===
Architectural typology identifies building form, organization of interior space, and structural hierarchy. Typology categories include long-standing Venetian forms such as monocellular, bicellular, tricellular, polycellular, unitary, modular, and shed structures.
These typologies were essential to our models because different types exhibit distinct relationships among measured height, floor height, and the expected number of floors.
This map illustrates a block of buildings on the island of FANT, with each building footprint color-coded by its architectural type. The different colors represent variations in construction period, structural form, and intended use, allowing for easy visual comparison of how architectural types are distributed within a single urban block.
Building Use Analysis and Spatial Visualization
Understanding how Venice’s buildings are used—residential, tourist-oriented, mixed-use, or vacant—is essential for analyzing depopulation, housing availability, and long-term urban sustainability. Our analysis combines architectural classifications, building-level fieldwork, and ArcGIS-based spatial modeling to create the first citywide, building-specific visualization of residential, non-residential, and vacant space across Venice. This unified dataset and accompanying maps form the basis for a continuous monitoring system for urban change in the historic city center.
=== Estimating Residential Units and Building Use ===
Residential unit estimates were derived using a unit & residential estimation model that merges:
Water meter data (number of units per meter, tariff type, and consumption)
Livable volume estimates (footprint × estimated residential floors)
Census-unit totals (to ensure tract-level consistency)
Observed proxies: doorbells, shutter status, ground-floor use
Water data also allowed us to distinguish between:
Primary residences (residential tariff, active usage)
Secondary residences / STRs (higher tariff, intermittent usage)
Vacant units (≤0.5 m³ annual consumption, validated through shutter data)
Visualization: Model Accuracy Checks
Figure 4.4: This bar chart compares the predicted number of floors from our linear regression model to the actual number of floors observed in the field. The distribution shows that the model performs very accurately: most buildings fall at zero difference, meaning the estimate matches the absolute floor count. Smaller groups appear at ±1 floor, which is expected given variations in building type and parapet height. Only a few buildings deviate by 2–3 floors, indicating that significant errors are rare and limited to atypical structures. Overall, the chart demonstrates that the regression model reliably estimates floors across Venice’s diverse architectural styles.
Figure 4.5: This chart compares two independent proxies for estimating building units: the number of water meters and the number of doorbells recorded in the field. The differences cluster tightly around 0, showing that meter-based unit estimates generally match what is visible on the building exterior. Some buildings deviate slightly in either direction (±1–5 units), reflecting cases where doorbells represent shared entries or meters serve multiple units. A few outliers appear at larger differences, typically in buildings with commercial ground floors or unusual internal layouts. The overall pattern confirms that combining water meter data with observational indicators provides a robust method for estimating residential units.
Mapping Residential vs. Tourist vs. Vacant Space
The unified building dataset allowed us to classify every building in Venice into residential, mixed, non-residential, STR/hotel, or vacant categories. These classifications enabled the production of several city-scale maps illustrating how building use varies spatially.
3D Residential Volume Mapping
By extruding footprints based on estimated residential floors, we visualized the concentration of long-term housing across islands.
Together, the fieldwork and data analysis support Venice’s first building-level model for estimating population, residential units, and vacancy. This system enables the city to identify underused or vacant structures that could be reallocated for long-term housing, and to pinpoint neighborhoods where non-residential or tourist-oriented functions dominate.
The 3D visualizations below illustrate the distribution of residential and non-residential space within each building. By comparing total building height to the proportion dedicated to residential use (green) versus non-residential use (pink), the model highlights patterns of mixed use, tourist pressure, and areas with limited housing availability.
Tourist & Non-Residential Concentrations
The following map visualizes our vacant unit estimation model, highlighting buildings with presumed unoccupied residential space. Each extruded shape represents the portion of a building’s total height that is estimated to be vacant, allowing us to identify where unused residential volume is concentrated across the city. This visualization helps reveal patterns of underutilization and areas where long-term housing capacity may be recoverable.