1 Introduction

The material balance of the Earth is being challenged. The year 2020 was marked as the year when the total weight of human-made materials globally surpassed the weight of all life on Earth, while it is estimated that in the years to come the growth rate of mass added to the anthroposphere will increase exponentially (Elhacham et al., 2020). In this context of hypergrowth coupled with the climate emergency, the growing rate of urbanization and the increasing social and political awareness on the matters of the Anthropocene, the topics of resource depletion or insufficiency are being reframed. Operating with an abundance mindset rather than from a place of scarcity (Gausa et al., 2020) is a new paradigm, especially relevant in the practices related to the design and production of the built environment, since it expands the definition of “resources” and where resources, raw and non-raw materials can be found and “mined”. If, as agents involved in the design, planning and construction process, we could shift the attention to the Anthroposphere as a source of, rather than just a destination for, processed goods, then we might have the potential to disrupt linear patterns of design and enhance circularity in cities and the built environment.

First introduced in the 1980s to refer to the recovery of rare metals trapped in waste electrical and electronic equipment (Nanjyo, 1988), the term urban mining is becoming a useful concept to explore and exploit the material stock in urban systems and to address the particularities of extraction and the use of anthropogenic resources in new production cycles (Baccini & Brunner, 2012). By reframing waste products generated in urban environments, or through urban activity, as anthropogenic stock, the concept of urban mining helps to shift the perception of resources from a finite starting point in a linear use process toward a cyclical state of materials in a circular process. As a key enabler of the current urban systems, as well as a beneficiary of the constant urban change rate, the construction industry is no doubt contributing a large percentage of material stock to the urban mine. As buildings are both affected and effectors in the current linear material system, it is pertinent to ask the question of how architecture, engineering, construction and operations (AECO) can contribute to the circular economy today.

2 Review contents

2.1 From waste to building matter

Starting from the material scale, we can observe that different types of urban or agricultural waste have the potential to become a new class of sustainable building composites through different processes that enhance their properties to match them with the building industry’s needs. This is the scope of the Digital Matter research studio at the Institute for Advanced Architecture of Catalonia in Barcelona (IAAC) that explores waste streams and experiments with upcycling waste into appropriate building components for different architectural features.

One of the case studies researches the production of value-added building composites using agricultural residues and food-processing side streams, specifically barley straw. The use of barley straw in construction is not new, but the construction industry has neglected this resource since there are numerous limitations on achieving surface and geometry variations, as well as on scaling up without the need for secondary structures. The case study focuses on overcoming current limitations in surface and geometry manipulation by introducing new additives and digital manufacturing techniques for the creation of a double-curved self-standing building skin. The composites explored include a variety of performances: hydrophobicity, for instance, is achieved by integrating natural resins and oils into the material system, which makes it appropriate for outdoor applications; and numerically controlled cutting surface subtraction for different textured surfaces provides humidity control and aesthetic richness. Within a holistic approach of circular design, the waste material is sourced by identifying the agricultural areas that can provide barley straw waste within a maximum distance of 30 km from the center of Barcelona, the city selected for the case study on retrofitting building skins with up-cycled materials (Fig. 1).

Fig. 1
figure 1

(Left) Barley straw waste material and up-cycled barley building composites for outdoor, indoor, structural and non-structural applications. (Right) Robotically controlled fiber placement and vacuum-assisted injection molding for the production of composites. The fabrication processes include vacuum molding, compression molding, 6-axis CNC, multi-point forming and curved mold jigs for double-curved building skin components. Image by: IAAC Digital Matter 2022, M. Housen, A. Ferragu, T. Mian, M. Muller-de Ahna

Full size image

Another case study based on experimentation follows similar logics and material protocols for the creation of cork-based and mycelium composites using the waste stream of the active cork industry in Spain and Portugal. Approximately half of all cork products are estimated to end up in landfills at the end of their useful life, while the other half is downcycled or used in biomass processes for clean energy generation. Taking into consideration that cork is a lightweight material with exceptional acoustic and thermal insulation properties, it is considered as an important natural alternative for architectural applications (Knapic et al., 2016). As cork is an appropriate medium for mycelium growth, the prototypes combine cork waste – in the form of granular cork or cork stoppers – and mycelium to create acoustic insulation products for urban applications and construction (Fig. 2).

Fig. 2
figure 2

(Top) Cork waste streams (granular cork or stoppers) and mycelium composites for acoustic insulation in buildings’ secondary skins. (Bottom) Cork and mycelium growth and composite assembly tests. Image by: IAAC Digital Matter 2022, M. Groth, F. Magaraggia, M. Calmanovici

Full size image

Paper is another waste material with the potential to be upcycled and used in specific architectural applications. Cellulose-based composites, for instance, can be used for the creation of temporary building skin components with the goal of enhancing biodiversity (specifically, protecting pollinators) and creating habitats for other species to inhabit the built environment (Fig. 3).

Fig. 3
figure 3

(Top) Secondary building skin of cellulose-based composites produced using up-cycled paper. The project aims to enhance biodiversity in cities and uses the material system to create urban beehives and pollination stations in buildings’ blind façades. (Bottom) Material explorations of cast and 3D-printed cellulose composites. Image by: IAAC Digital Matter 2022, A. Gultekin, A. Kalra, N.J. Pattanshetti and P.V. Patilkulkarni

Full size image

From barley straw to cork, paper, or even polyester and the leftovers of a highly contaminating textile industry, these approaches explore the benefits and potentials of applying waste or “pure” residue material in the built environment. Given that the construction industry is responsible for the largest amounts of material consumption annually, and it occupies the largest volume in the anthroposphere, integrating circular and up-cycled materials into architectural applications can have an exponential impact, extending into the other layers of our ecosystem.

Mining and processing non-traditional materials for construction, however, requires complex levels of certification or even standardization that are not easy to achieve in academic and research environments. At the same time, fundamental changes in the construction norms are required for extending the application of such materials. Moreover, it is significant to underline the aesthetic affordances of such “waste-to-matter” approaches. Buildings are more than the sum of their parts, and apart from serving the physical purpose of housing and protection they also represent cultural stock, which can reach deep into centuries of history as well as projecting visions for the future of places and communities. Responding to all these needs with a new aesthetic that is also true to the materials’ origins and specificities is a difficult but relevant task for our contemporary sustainable building practices. Aside from innovating with waste-driven building materials and exploring new construction systems in which these nontraditional materials can safely be used as building components, a new territory related to the aesthetics debate will be more relevant than ever for reflecting the source of the materials and promoting the culture of the circular economy.

2.2 Reframe the current building stock

“We build in the express knowledge that all buildings will disappear.”

(Reiner de Graaf, 2017 )

“Cities are the mines of the future” suggested Jane Jacobs, recognizing that the postwar construction boom was causing a temporary displacement of materials (Jacobs, 1961). Decades later, urban mining adds validation to this recurrent statement and provides a framework for new scoping and extraction processes centered on non-raw materials in the urban stocks. While this concept is effective at creating an opportunity for part of the waste to be reprocessed (increasing its residual value by adding it into a new cycle), sourcing and extraction from landfills or unstructured extraction from demolition sites poses problematic concerns both in terms of the loss of material value and the safety and wellbeing of those involved. For an efficient integration of urban mining in the mainstream production processes, a holistic concept of efficient resource extraction, production, use and reuse is necessary, tracking relevant material data through the cycles of dormant, transformative, active use and re-harvesting states of the material throughout its lifecycle.

A rather daunting observation for the architectural profession, the realization that the imminent fate of all buildings is their disappearance (largely through demolition) can be outweighed by reframing current building stock as the largest container of future resources in relatively stable compounds. In this context, we observe a significant shift in the location of raw materials from their original geologic compounds, now bound up in existing anthropogenic structures – especially buildings (Hillebrandt et al., 2019). A particularity of buildings as anthropogenic compounds, in contrast to their geological counterparts, is their volatile stability. Buildings are created as a response to multiple factors, including physical or social need as well as cultural and economic pressures. Similarly, their active-use state, renewal and demolition periods are affected by the same external agents. Materials in structural or integrity components of a building have a life expectancy reaching several hundreds of years, meaning that the buildings’ lifespan states often do not correspond to the lifespan of the materials used in construction. It is this desynchronization between the timespan during which a building is necessary and valued and the timespan during which its constituent materials are sound for use that reveals a great potential for material reuse, allowing the original raw materials to increase in value through reprocessing and reuse in various building cycles before their final reintegration into the natural environment.

2.3 Data matters

Taking into account current information policies, practices and building standards, it has been widely suggested in global literature and reports on the industry that the maximum level of recovery and maximum performance in building demolition processes are already being achieved (Koutamanis et al., 2018). Such claims, however pertinent, consider only the present state through the lens of past policy and effectively demonstrate the need for further reconsideration of the political, technological and informational framework within which construction and demolition waste (CDW) recovery is to be framed. Baccini and Brunner identify key considerations for the efficient recycling of building material waste to be connected to the availability of accurate geolocated information and far-reaching urban-scale decision-making and stock management (Baccini & Brunner, 2012).

Often packaged under the term CDW, the material released at the end of a building’s lifecycle, is a specific type of stock resource found in active compounds within urban built structures, and it experiences significant value loss when scoped and extracted according to current practices of linear use. The urban mining concept regarding building products addresses a stock of engaged material with multiple use cycles. In this sense, the urban mine stockpile is active, so it must be scoped in such a way that time becomes a significant factor for its assessment. Thus, the construction and demolition stock in urban mining must be understood and treated as active stock. As opposed to raw extraction sources, where the material is in a “dormant” state, in the active state a material has a short window of availability: mainly during the demolition process. Significant value loss can be reduced if the material is mined before the end of this ideal timeframe.

It is in this regard that data about the specifications of the material – including dimensions, state, properties or even geolocation, among others – become highly relevant for the potential of reuse. Establishing methods for scoping, extracting, reprocessing, re-commercializing and reusing existing building stock can have a positive impact both by offsetting embodied carbon and by avoiding environmental costs. The urgency of the creation of a digital layer to enable the mining and reuse of otherwise purely physical construction and demolition materials is of utmost importance in this context. Additional data related to the market demand and the criteria for setting prices for the active stock is also necessary. Strategic planning, coupled with transparent and structured public information, can bridge the process and support market demand and fair pricing while the still-active material reaches the end of one use cycle. With a territorial framework for material data and local policy support, the urban mining of active building stock could be achieved for both existing and newly constructed buildings.

2.4 Material data on building scale stock

Scientific sources looking at CDW recovery often cite information deficit as the main culprit in ineffective CDW recovery and recirculation, and policies related to adding a digital layer to future urban stock are being adopted by key global players. Given the present level of urbanization, apart from looking at the construction industry’s ebb and flow cycles, and creating a framework for public information and circularity standards for future buildings, a more pressing and far less resolved issue involves addressing the existing building stock from the perspective of identifying and recovering trapped resources in structurally integral state.

One of the research lines of the Master’s in Advanced Robotics and Construction at IAAC explores the possible use of aerial and ground robots coupled with computer vision for scanning, sorting and digitally archiving existing material stock in buildings. The research aims to employ an automatic digitization method for the near-end-of-life stage of a building and considers it as a source of high-value assets. In order to encourage the use of valuable secondary sources of materials and to better inform early design choices when reusing construction waste, the resulting dataset is then distributed to designers and builders (Batalle Garcia et al., 2021).

The first part of the research entails using drone photos and computer vision to automatically identify building materials and components at pre-demolition sites and therefore contribute to the creation of an analytical and logistic support system that can be used to define optimum deconstruction and reuse strategies in terms of environmental and economic sustainability. The project contributes to the creation of a toolbox for quantifying, geolocating, indexing and assessing the state of materials and their potential for reuse in existing buildings, a significant territory that requires further development for the applications of circular construction (Fig. 4).

Fig. 4
figure 4

(Top): MatterSite uses scanning technologies to identify and sort the material stock of existing buildings for possible reuse. (Bottom): Raw imagery captured from the site is run through a material localization algorithm to pick out relevant recoverable materials from these views. Image: IAAC MRAC Matter Site, A. Batalle Garcia, Cebeci, Y. Irem, R. Vargas Calvo, M. Gordon

Full size image

The second section of the study aims to transfer the previous database to as many stakeholders as possible related to the construction sector. In order to impact design and construction, it is imperative to match offer and demand; therefore, a computational design tool is incorporated into existing construction software in order to offer recommendations for design adjustments in the early design stages. The suggested system additionally offers recommendations for matching specified design components with the materials that have previously been identified as available resources for reuse (Batalle Garcia et al., 2021).

The Barcelona-based start-up Scaled Robotics is another case study that uses robots and terrestrial scanners for conducting detailed site surveys at on-demand or at regular intervals in construction or demolition sites. While the technology for scanning and sorting is mature, it is important to highlight the lack of policies that could contribute to creating a larger market demand for reused components and materials (Fig. 5).

Fig. 5
figure 5

Robotic scanning of the interior of a construction site, Image copyright @ScaledRobotics

Full size image

The above case study portrays the potential for constructing big data sets of reusable materials, digitally available for sharing and organizing material harvesting and facilitating the incorporation of those materials into new designs. Thanks to this anticipatory approach and digital classification, the materials identified (for example: windows, metal beams, doors, etc.) at deconstruction or renovation sites are directed to reuse actors and thus oriented towards new building construction programs, promoting the circular economy in a short loop.

While some questions related to the cost of reuse versus the cost of using new resources in construction are still to be answered, such processes and techniques facilitate the implementation of the concept of material passportsFootnote 1 for buildings, which encapsulates many of the necessary criteria to support reuse. The significance of material passports for the implementation of the circular economy in the construction sector calls for new policies that regulate the documentation submitted for new constructions. For new buildings, for instance, design should go beyond the project handover stage and include strategic planning for the material sourcing, assembly, dismantling, reprocessing and storing before a new use cycle. At the same time, reality scanning and computer vision (CV) provide the opportunity to include existing building stock in such policies and to compile its material passport retroactively.

2.5 Material data on urban scale stock

A coherent strategy for urban mining in the context of the built environment offers the possibility for bridging the gap between the circular production of materials and the urban system as a complex multidimensional locus both for the sourcing and the destination of its products. The key obstacle now lies in the difficulties of generating a global, data-rich, and open model of this multidimensional system. The concept of digital twins – introduced by NASA in 2010 and used to describe an alternative to the physical modeling of complex simulation systems – involves virtual representations of physical systems that can be used to contain time-sensitive data, perform simulations and make decisions with predictive or analytical AI modeling. Digital twins are currently used extensively to model complex systems in various disciplines and are gaining popularity in the AECO as well as in urban administration, especially in relation to smart buildings and smart cities. By adding a digital layer to an existing urban system, the digital twin creates an opportunity to hold geolocated and historical data and provides a necessary means of multiuser communication within a single simulation framework (Apte & Spanos, 2022).

We use digital twins daily without much forethought to observe, design and navigate our cities, and open-source satellite imagery as well as vector-based geographic databases have become deeply integrated into our professional practice. Starting, for instance, in 2007, an invaluable initiative was launched by Google, adding to the complexity of openly available geographic data: street view images were added to the company’s maps, providing an almost real-time digital access to even the most remote urban locations.

Digital technology affects most all present-day activity and, as has been pointed out in recent years, the digital bank contains unimaginable amounts of data, trivially called “big data”. While this data is available, it largely takes the form of unstructured data. Its vast extent makes it difficult to analyze using common tools. However, another branch of technology gaining popularity in tandem, artificial intelligence (AI), provides frameworks for making use of unstructured data through predictive means (Anderson, 2008).

In the intent to create an open repository to map the material availability in the existing building stock of a city, IAAC’s Master in City & Technology used AI and CV to develop a framework starting from the premise that relevant information pertinent to the material composition of buildings can be identified through the analysis of the building façades obtained through the Google Street View API (application programming interface). By applying predictive modeling at the city scale, the algorithm can identify, geolocate and quantify façade materials with a present accuracy of 87%. While, in general, we expect machine learning models to achieve much higher levels of accuracy, it is important to adapt this expectation to the specific application and to establish clarity on how much uncertainty can be allowed given the data quality (Raghu et al., 2022). The value of this process lies not in achieving near 100% accuracy but in the use of open unstructured data as a basis, making it a highly repeatable and expandable application, and attempting to propose a solution that crosses the boundaries of the original case study and could democratize access to data for local decision makers in countries where open public information is less accessible (Figs. 6 and 7).

Fig. 6
figure 6

(Top) A street view image and its corresponding labeled copy, identifying façade materials and components. (Bottom) The predicted materials on building façades that were not part of the training dataset. Image by: IAAC MaCT 2022, H. Shawqy A. Markopoulou, O. Taut

Full size image
Fig. 7
figure 7

The results of the predictive AI model are displayed in a web interface showing the material availability overall and the predicted material distribution on the street-facing façade of any building. Image by: IAAC MaCT 2022, H. Shawqy A. Markopoulou, O. Taut

Full size image

Working with this interface, the Internet of Buildings research team from the Master in City and Technology at IAAC identified further opportunities to overcome the limitations of the AI model and extrapolate common urban typologies that can improve the accuracy and expand the scope from just the façade to other building elements. The result of this research is proof that the task of quantifying different materials in the city can be achieved by combining the data from statistical models (from multiple open-source urban databases) with the predictions of façade materials based on computer vision and Open Street View to identify patterns that are relevant for making estimations of the quantity, state and projected availability cycle of: concrete, brick, stone (granite), metal (steel), timber and glass (Figs. 8, 9 and 10).

Fig. 8
figure 8

Decision tree showing the connection between combinations of data filters and the probability of finding wood in the skin or the structure of buildings. Image by: IAAC MaCT Internet of Buildings 2022, D. Lampriadis, J. Veiga, M.A. Kroetz, Y. Wadia

Full size image
Fig. 9
figure 9

(Top): A Sankey diagram shows the process and methods currently used in linear and circular concrete productions. This informed the decision to create distinct mapping for precast and in-situ concrete and the map filters used to identify both. (Bottom) The mapping of cast in situ concrete considers buildings constructed after 1960 with a proportion of a minimum 40% of the façade predicted as plaster and a façade area in the 90th percentile. The result is presented as a geolocated gradient map as well as a bar chart computing neighborhood totals. Image by: IAAC MaCT Internet of Buildings 2022, J. Lee, J.B. Saleh, K. Buhari

Full size image
Fig. 10
figure 10

Combined mapping of wood in existing buildings both in the form of structures and enclosures. The results are predicted based on the diagram in image 8. The bivariate graph on the left shows strong directed correlation between probability of structural wood and enclosure wood. Image by: IAAC MaCT Internet of Buildings 2022, D. Lampriadis, J. Veiga, M. A. Kroetz, Y. Wadia

Full size image

Apart from the currently available indicators used above (such as building age, height and use), more complex factors can be used in a comprehensive methodology tool for the computation of building typology clusters that can further improve the detail of the predictive modeling of urban mining potential. For example, a parallel study in the field of urban data science, “Global Building Morphology Indicators” (Biljecki & Chow, 2002) details globally relevant methods for extracting building- and context-specific topological information that can be used as a basis for a typological analysis comprising geometric, geographic and material indicators.

3 Conclusions

3.1 Notes on scale, temporality and accuracy

At different levels of geographic and temporal scales, the minimum viable accuracy levels vary proportionally. While the estimation based on unstructured data presented above can be improved by aggregating layers of building typology data, it remains less accurate than on-site observational study. However, there is an invaluable benefit achieved by providing a high-level overview at the territorial scale to support urban decision makers with enough anticipation and localization of future material stock as needed for a locally viable strategy regarding material circularity. As expressed above, in the context of the circular economy in AECO, construction professionals are simultaneously the suppliers and the buyers of recovered materials – only, the two states are separated by a long-spanning temporal axis. The simulation of material stock flows as supported by the urban scale model based on predicted data can successfully link the two endpoints, revealing the full scale of the problem/solution continuum of building-material circulation.

Supported by this global layer, non-invasive digital imagery and building information modeling (BIM) reconstruction methods, like the ones presented above, can be employed in a just-in-time (JIT) fashion as part of a demolition project to make a more accurate estimation of the materials and components that are available for extraction and to design a building-specific mining strategy. When considered from the perspective of urban mining, the process of demolition requires forethought and design, ensuring that a palette of elements is extracted following concepts of minimum effort and maximum gain. At the same time, the extracted material is only economically viable when it is required as part of a new building process. This duality requires the creation of a feedback loop between new building design and demolition design that can be supported by mass customization computational methods.

However, for both processes to become standard practice in the AEC, aside from the support of an urban-scale framework and policy, it is crucial for AEC professionals to climb the innovation adoption curve, creating demand for technological advances in material extraction for reuse in order to achieve the necessary economy of scale. Just as the current linear production methods are informed and innovate in response to current performative and aesthetic needs emerging from AEC professions, circular production materials and methods will also require the motor of market demand. Research transcending disciplines, proposing performative systems that consider the material at different scales, can achieve consistency and consequence between the chemical level and the aesthetic expression in the built environment. Projects like the material explorations shown here can become the key drivers of the scale adoption of new architectural systems that rely at their core on circular economy principles.