Data-driven urban governance: three implementations for sustainable urban land management through spatial data science

The environmental consequences of intense resource consumption from the planet´s accelerated urbanization process demand a better-performing built environment in the context of ongoing climate change. Recent research concludes that almost forty percent of global carbon emissions are linked to the Architecture, Engineering, Construction, and Operation (AECO) industry. The global construction sector alone consumes almost half of the world's total material production, implying it has the highest single-category material footprint across the global economy. Furthermore, the production of traditional building materials like concrete and steel requires a large amount of energy and accounts for high amounts of greenhouse gas emissions.
Data-based governance is a key premise for achieving significant circularization of urban metabolism, both in terms of construction materials and operational energy. A data management platform capable of handling both the geometric and semantic components of an urban digital twin represents a valuable tool for data-driven policy-making and a pragmatic shift from global warming mitigation to adaptation. This research describes the design process and prototype implementation of such an analytical tool as an alternative methodology for City Information Modelling (CIM) through the analytical curation of a “Bottom-up” data model. The technical development of this platform relies on the articulation of Building Information Modelling (BIM) and Geographic Information Systems (GIS) data, aggregated, managed, and presented through a Business Intelligence (BI) cloud-based platform. The proposed workflow doesn´t create a single data format or standard (such as the existing CityGML), but instead, diversifies data sourcing through the flexibility of BI interoperability, and allows for the creation of multi-scale, multi-variable three-dimensional geoportals.
The first pilot implementation of this methodology took place in the city of San José, Costa Rica, as an urban analytics platform, whilst further experimental iterations changed their location and main topic. The second of these experiments focused on the digitization of nature-based value chains as described hereunder.
One of the most promising alternatives in the pursuit of more sustainable construction and circular material management trends is the massification of wood use in the building sector, however, the mainstream praxis of timber-based construction is still connected to a rather manual and analog value creation chain. To make the process more time- and energy-efficient, the previously mentioned (data-driven) methodology has been adapted and implemented on a value-creation system of urban timber construction from the forest to the city in the region of Berlin-Brandenburg. The three-level framework, based on the principles of Industry 4.0, connects urban development planning with the vertical networking of digital technologies and the horizontal networking of phases along the value chain. The created Value Chain Digital Model (VCDM) has materialized in the form of online, interactive dashboards using a BI platform to explore the close relationships between the stages of the wood products´ life cycle.
Finally, in the search for nature-based alternatives, bamboo grass emerges as an interesting and feasible option in the global tropical belt (and the entire global south region) due to significantly shorter rotation times and high potential for mixed agroforestry when compared to softwood or hardwood species. To determine the potential for mass-scale bamboo harvesting within a value-chain work frame, the first step is to technically evaluate the viable land for cultivation.
This research concludes its operative proposal with the creation of a remote-sensing-based tool for surveying and exploring bamboo agroforestry potential over the entire territory of the Global Tropical Belt based on climatic conditions, altitude, topography, existing land cover, and soil characteristics (texture and acidity). This allows for technically driven soil selection, with optimal climatic conditions in the pursuit of faster plant development, higher number of culms, larger diameters, cheaper production, and shorter times until plant maturity.
Three different, yet connected, implementations of Spatial Data Science (SDS) come together in this document to demonstrate the potential of data-driven analysis in contemporary urban management (including planning and design). At the core of these implementations, an alternative methodology for data modeling and hypothetical scenario exploration (BIM/GIS/BI) constitutes the main contribution to the field and an open door for further research. The main benefits of the proposed methodology include widespread format interoperability, simultaneous multi-variable/multi-scale analysis, plausible articulation for mixed private-public data modeling, parametric scenario benchmarking, online widespread accessibility, low-latency (daily) data update capability, low implementation costs, and a high potential for AI-based sophistication in the near future. After five years of research and experimentation, this study hopes to prove the holistic compatibility between SDS and the global urbanization process, illustrating its potential as a new opportunity for architectural and urban career development.