Speaker
Description
Complexity in contemporary urban areas, profound uncertainty regarding future intervention outcomes, and the multi-decade social-ecological implications of strategic planning policies are increasingly central to spatial planning discourse and remain essential for transformative adaptation. Despite this urgent necessity, traditional urban planning systems predominantly rely on rigid, top-down, and linear decision-making hierarchies. These conventional governance frameworks create a significant structural mismatch between short-term political cycles, typically spanning five years, and the long-term ecological consequences of urban interventions, which often take thirty years to showcase. European policy frameworks currently ask for this extended institutional horizon, with biodiversity strategies and nature restoration regulations mandating no net loss rules and continuous increases in urban green spaces through 2050. In practice, every spatial decision, from altering a localized maintenance regime to unsealing a commercial plot of land, actively closes off certain futures while simultaneously opening others, ultimately committing cities and their fragile living ecosystems to multi-decade developmental trajectories. When past strategic failures go entirely unexamined and critical natural actors are denied a voice in the decision-making process, a documented erosion of civic trust in planning institutions naturally follows.
Everyday uncertainty in these complex environments is increasingly shaped and governed by algorithmic systems and data-intensive digital infrastructures that mediate resource access, influence spatial prioritization, and structure urban knowledge (Batty, 2024; Bonin & Frankhauser, 2025). While urban planning features a history of utilizing computational simulations to navigate human-nature futures, ranging from cellular automata to early agent-based models, persistent conceptual limitations severely restrict the efficacy of contemporary digital tools (Kogut & Van Der Heijden, 2025). Standard urban digital twins frequently remain confined to single-sector silos (Crooks et al., 2021), most notably transportation and mobility, and treat the natural environment merely as a static spatial layer rather than an interconnected participant. Furthermore, the simulated actors embedded within these conventional twin architectures are typically reduced to reactive entities governed by fixed behavioral rules or basic utility maximization, inherently failing to capture the proactive, heterogeneous desires of actual urban populations (from human to nature). To address these tensions between traditional institutional governance and modern model-mediated reasoning, this research introduces an advanced methodological framework: an exploratory bio-digital twin specifically designed to operate at the meso-urban scale, defined as a cohesive cluster of adjacent neighborhoods spanning approximately two to ten square kilometers.
This exploratory bio-digital twin moves beyond the limitations of standard digital replicas by functioning as a calibrated computational environment that actively models the complex relational dynamics between human populations and the living ecological systems that surround them. The structural core of this architecture is a dedicated, auditable translation layer. This critical mechanism directly resolves the challenge of integrating unstructured, qualitative social-ecological knowledge into quantitative variables and functions. The translation layer systematically converts qualitative inputs such as community perceptions, co-creation workshop outputs, stated citizen priorities, and non-human behavioral rules into formalized, quantified algorithmic parameters. To ensure methodological rigor and to definitively prevent the arbitrary quantification of subjective values, the framework enforces a strict traceability mandate, requiring every resulting computational parameter to remain explicitly linked to its original evidentiary source. By hybridizing computational social science with advanced landscape ecology, this structure applies comprehensive global sensitivity analyses across the translated parameter matrix to determine exactly which integrated social and environmental values drive the most significant variances in urban outcomes. Consequently, the twin establishes a unified formal ontology where dynamic ecological rules and qualitative human desires actively coexist, offering a powerful computational tool to address the environmental crisis without defaulting to purely technocratic or anthropocentric metrics.
Deep uncertainty is intrinsically embedded into the bio-digital twin through a highly specified, multi-species agent-based model (ABM) architecture (Chapman et al., 2023; Rosello et al., 2022). To reflect near real-world political and institutional decision-making processes under conditions of uncertainty, the framework abandons reactive agent profiling in favor of proactive, bounded-rational entities driven by a sophisticated belief-desire-intention (BDI)(Ghaffarzadegan et al., 2024; Marçal Russo et al., 2025) architecture. Within this specific ontology, agents process information based on context-dependent, localized environmental realities rather than relying on omniscient or perfectly optimized synthetic data. The precise categorization of these agents is deliberately designed to influence visibility, equity, and legitimacy in everyday planning practices. Human agent groups are systematically stratified across diverse socio-economic, age, and gender lines, ensuring that the differentiated impacts of urban planning on vulnerable demographics remain computationally visible rather than obscured by generalized, aggregated municipal averages. Emphasizing a vital more-than-human perspective (Ayatollahi et al., 2026; Lazurko et al., 2025), the model simultaneously treats selected flora, fauna, and soil profiles as active stakeholders possessing their own inherent spatial requirements. This unique non-human agency is operationalized using simplified ecological BDI proxies parameterized directly from biological literature, encompassing specific scientific rules for habitat suitability, phenology, and multi-species dispersal capabilities. To maintain computational tractability within the simulation, the overarching behavioral goals of these highly heterogeneous agents are explicitly bounded by the analytical dimensions of the Nature Futures Framework. The interconnected relationships and spatial dependencies between all human and non-human agents are subsequently mapped using detailed causal loop diagrams, identifying the complex balancing and reinforcing feedback mechanisms that dictate long-term ecosystem resilience.
By systematically projecting detailed trajectories from 2028 to 2048, the bio-digital twin demonstrates how ABM models contribute to framing the boundaries of possible futures. As human-nature relationships continually evolve under deep uncertainty, the computational model makes no attempt at singular, deterministic point predictions. Instead, it bounds a robust space of plausible future scenarios and rigorously evaluates them utilizing the Dynamic Adaptive Policy Pathways (DAPP)(Haasnoot et al., 2013, 2024; Langendijk et al., 2025) methodology. DAPP provides a structured algorithmic approach for analyzing complex sequences of strategic planning decisions over multi-decade horizons, allowing urban authorities to evaluate localized interventions comprehensively before irreversible institutional capital is committed. Crucially, these adaptive policy pathways are not generated through isolated computational abstraction; they are explicitly co-produced alongside local planning actors, urban ecologists, and diverse citizen groups through localized forecasting and visioning sessions. The digital twin executes continuous simulations across these co-created pathways within an open-source NetLogo environment, projecting critical physical environmental outcomes such as habitat fragmentation rates, changing localized thermal conditions, canopy development milestones, and surface soil sealing. Through these multi-decade projections, the tool exposes the profound potential long-term costs of maladaptation, structurally influencing how everyday planning resources are prioritized, legitimized, and eventually deployed.
The theoretical architecture and operative forecasting capabilities of this exploratory bio-digital twin are empirically grounded and tested through its direct application to the GOCCIA (Green Opportunities to Clean-up Contaminants through an Interspecies Alliance) urban regeneration project in the Bovisa neighborhood, northeast of Milan, Italy, under European Urban Initiative funds. The area is expected to host more 20 thousands of university affiliated users, visitors, and locals with long-term phytoremediation plans for soil in certain limited areas. Characterized by its highly complex industrial heritage and a dense, spontaneous urban forest, the GOCCIA site represents an optimal meso-scale testing environment for modeling social-ecological conflicts and addressing deep planning uncertainty. Within this specific institutional and geographical context, the bio-digital twin computationally simulates the intricate, tense dynamics between acute urban development pressures, the mandated necessity of preserving spontaneous biodiversity, and the socio-economic requirements of the surrounding Milanese districts. Human, flora, and fauna agents within the GOCCIA simulation are parameterized considering projected real-world conditions and historical information to accurately reflect the diverse status of the local condition. Simultaneously, non-human agents accurately represent the established flora and fauna of the GOCCIA spontaneous forest, strictly driven by ecological rules dictating their survival and biological dispersal in the face of proposed infrastructural shifts. By projecting twenty-year developmental pathways specifically tailored for the GOCCIA site, the twin allows Milanese planning authorities to dynamically test how differing degrees of commercial vegetation removal or targeted soil sealing directly impact the equitable, long-term distribution of vital ecosystem services among stratified human populations and local habitats.
Ultimately, the broad integration of formal computational ontologies and decision-making tools into strategic planning carries profound implications for the legitimacy of contemporary planning practices. The bio-digital twin is methodologically constructed to function as an iterative, highly feedback-sensitive learning instrument. The critical transition from a theoretical spatial simulation to a legitimate, embedded governance tool is achieved through continuous "return and learn" validation cycles. Simulation outputs and detailed scenario pathway assessments are translated into accessible, non-technical formats, including strategic fact sheets, analysis charts, and targeted municipal policy briefs. These actionable insights are formally presented to planning decision-makers during biannual assessment sessions to facilitate the collaborative, multidisciplinary evaluation of future ecological tipping points. Concurrently, analytical results are shared directly with local residents and civil-society actors through open neighborhood information sessions, an approach that deliberately prioritizes the inclusion of socio-economic demographics routinely absent from conventional, top-down planning consultations. This structure ultimately ensures that adaptive urban planning, even when operating in deeply data-intensive environments, remains both structurally robust and fundamentally legitimate in the face of escalating everyday uncertainty.
References
Ayatollahi, A., Frantzeskaki, N., & Morello, E. (2026). Co-evaluating the social benefits of urban biodiversity: A case study of the Botanical Gardens at Utrecht University, the Netherlands. Urban Forestry & Urban Greening, 116, 129236. https://doi.org/10.1016/j.ufug.2025.129236
Batty, M. (2024). Digital twins in city planning. Nature Computational Science, 4(3), 192–199. https://doi.org/10.1038/s43588-024-00606-7
Bonin, O., & Frankhauser, P. (2025). Digital twins versus simulation models and planning concepts: Increasing the resemblance between a digital twin and the modeled territory to facilitate the construction and evaluation of urban scenarios. Environment and Planning B: Urban Analytics and City Science, 23998083251318474. https://doi.org/10.1177/23998083251318474
Chapman, M., Xu, L., Lapeyrolerie, M., & Boettiger, C. (2023). Bridging adaptive management and reinforcement learning for more robust decisions. Philosophical Transactions of the Royal Society B: Biological Sciences, 378(1881). https://doi.org/10.1098/rstb.2022.0195
Crooks, A., Heppenstall, A., Malleson, N., & Manley, E. (2021). Agent-Based Modeling and the City: A Gallery of Applications. In W. Shi, M. F. Goodchild, M. Batty, M.-P. Kwan, & A. Zhang (Eds.), Urban Informatics (pp. 885–910). Springer Singapore. https://doi.org/10.1007/978-981-15-8983-6_46
Ghaffarzadegan, N., Majumdar, A., Williams, R., & Hosseinichimeh, N. (2024). Generative agent‐based modeling: An introduction and tutorial. System Dynamics Review, 40(1), e1761. https://doi.org/10.1002/sdr.1761
Haasnoot, M., Fant, V. D., Kwakkel, J., & Lawrence, J. (2024). Lessons from a decade of adaptive pathways studies for climate adaptation. Global Environmental Change, 88, 102907. https://doi.org/10.1016/j.gloenvcha.2024.102907
Haasnoot, M., Kwakkel, J. H., Walker, W. E., & Maat, J. ter. (2013). Dynamic adaptive policy pathways: A method for crafting robust decisions for a deeply uncertain world. Global Environmental Change, 23(2), 485–498. https://doi.org/10.1016/j.gloenvcha.2012.12.006
Kogut, P., & Van Der Heijden, R. (2025). Digital Twins for Urban Governance: General Desires, Expectations, Challenges. In L. Raes, S. Ruston McAleer, I. Croket, P. Kogut, M. Brynskov, & S. Lefever (Eds.), Decide Better (pp. 9–31). Springer Nature Switzerland. https://doi.org/10.1007/978-3-031-81451-8_2
Langendijk, G. S., McEvoy, S., McCullagh, D., & Haasnoot, M. (2025). Supporting climate resilient development planning − a dynamic adaptive pathways based approach and an illustrative case from Cork City, Ireland. Global Environmental Change, 95, 103070. https://doi.org/10.1016/j.gloenvcha.2025.103070
Lazurko, A., Pater, M. de, Kim, H., Hebinck, A., Biesbroek, R., DeClerck, F., Krupnik, S., Okruszko, T., Pereira, L. M., Proka, A., Sakellari, E., & Harrison, P. A. (2025). Envisioning nature-positive futures for Europe: Inspiring transformative change at the biodiversity nexus. Ecosystems and People, 21(1). https://doi.org/10.1080/26395916.2025.2561107
Marçal Russo, L., Dane, G., Helbich, M., Ligtenberg, A., Filomena, G., Janssen, C. P., Koeva, M., Nourian, P., Patuano, A., Raposo, P., Thompson, K., Yang, S., & Verstegen, J. A. (2025). Do urban digital twins need agents? Environment and Planning B: Urban Analytics and City Science, 23998083251317666. https://doi.org/10.1177/23998083251317666
Rosello, C., Guillaume, J. H. A., Taylor, P., Cuddy, S., Pollino, C., & Jakeman, A. J. (2022). Identifying Minimum Information Requirements to Improve Integrated Modeling Capabilities: Lessons Learned From Dynamic Adaptive Policy Pathways. Frontiers in Water, 4. https://doi.org/10.3389/frwa.2022.768898