Middle-out design for more-than-human cities: integrating human–animal relations in urban sustainability planning Open Access

Urban environments face escalating socio-environmental challenges, including rapid urbanisation, habitat degradation, climate change and extreme weather events. Sustainable urban development has emerged as a critical concept for addressing such contemporary urban challenges and creating pathways towards more resilient and liveable cities (Yigitcanlar et al., 2019). Urban sustainability has moved beyond theoretical discourse to become deeply embedded in practical policy-making, directly influencing how cities approach planning, development, and environmental management decisions (Alberti, 2024; Davidson and Arman, 2014; Polk, 2011). The integration of sustainability principles into urban planning represents a fundamental shift in how cities conceptualise and implement their development strategies (Wheeler, 2013). Urban sustainability planning is a multidisciplinary approach to developing and managing cities that balance environmental protection and climate change mitigation strategies with social equity and economic development (Almulhim et al., 2024; Birkeland, 2016; Hamida et al., 2024).

Human-centred design, a widely adopted methodology, focuses on addressing people's needs, behaviours and experiences, ensuring that solutions enhance human well-being and usability (Giacomin, 2014). More-than-human design expands this perspective by considering the needs and well-being of non-human entities, such as animals, plants, ecosystems and intelligent systems (Fieuw et al., 2022). This holistic methodology acknowledges the interconnectedness of all life forms, yet its implementation presents numerous challenges. Society's extensive history of policies, laws and social conventions has influenced design and decision-making, often reinforcing human-centred priorities (Nguyen et al., 2024). However, as environmental impacts, and political ideologies change, it is increasingly important to adopt approaches that incorporate diverse knowledge systems, such as Indigenous knowledge, scientific research and ecological perspectives, to promote multispecies viewpoints and an eco-centric mindset (Jebari and Sandberg, 2022; Kalashyan, 2025; Kortenkamp and Moore, 2001).

To respond to such impacts and changes, collaborative approaches and methodologies involving a diverse representation of actors are required for better design and decision-making outcomes within urban environments. These approaches should also foster new ways of building the capacity of urban planners to consider more-than-human perspectives and ecological interdependencies (Raymond et al., 2025). Drawing on the collective knowledge, skills and creativity of actors, such an approach would better reflect the needs, wants, and aspirations of both humans and non-humans. Building upon this foundation, we propose Middle-Out Design (MOD) as a structured framework for the collaborative design, development and deployment of design and decision-making outcomes. MOD demonstrates how collective knowledge, skills and creativity can be harnessed to reflect the needs, wants, and aspirations of both human and non-human actors. By leveraging the expertise of all actors, MOD fosters inclusivity and collaboration, particularly within the built environment, offering a pathway for addressing the gaps in existing approaches (Fredericks et al., 2023).

The overarching aim of this article is to advance Middle-Out Design (MOD) as both a conceptual and methodological framework for collaborative and multispecies urban sustainability planning. The article develops the theoretical foundations of MOD and illustrates its application through four vignettes (Leszczynski, 2020) spanning collaborations between humans and non-humans. These vignettes demonstrate how MOD supports the formation of coalitions between diverse representatives, which enables equitable decision-making that extends beyond conceptual discussion into real-world implementation, while also revealing the challenges and limitations encountered in practice.

The remainder of this article is structured as follows. First, we review existing approaches to participatory design, planning and coalition-building. Next, we introduce the MOD framework by outlining its four stages. We then present four vignettes that demonstrate how this framework facilitates collaboration among diverse human and non-human actors. Finally, we synthesise by reflecting on the broader implications of MOD for inclusive urban sustainability planning and considering its potential to support more adaptive, responsive and ecologically attuned decision-making practices.

Participatory approaches, including community engagement, are widely practised in urban planning to inform, consult and collaborate with diverse actors. Community engagement is commonly used to gather public feedback on policies, strategic planning and infrastructure projects, with the goal of integrating community perspectives into decision-making (Fredericks et al., 2019, 2020). However, while these processes can generate meaningful input, they are often constrained by the objectives of top-down decision-makers. In many instances, community engagement is reduced to “engagement theatre” (Foth and Burton, 2026; Kamols et al., 2021), where participation legitimises predetermined decisions rather than shaping planning outcomes.

Participatory governance approaches, such as citizen assemblies and participatory planning initiatives, aim to broaden democratic involvement by bringing together diverse representatives to contribute to public decision-making (Ferejohn, 2008). Citizen assemblies, widely used in European countries, were initially introduced to give voters a voice on sensitive issues and are now commonly employed to shape policies, including climate legislation (King and Wilson, 2023). Participatory planning ought to enable communities to influence urban development by contributing local knowledge and collaborating with government and institutional actors to shape outcomes (Foth, 2018; Healey, 1998). Other deliberative formats, such as Delphi panels and dialectical inquiry, adopt expert-led, structured methods to refine decisions through informed debate (Frauenberger et al., 2018; Meijering et al., 2018; Roosen et al., 2020). While these approaches can enhance inclusivity, they often remain constrained by institutional structures that determine who participates and whose voices are prioritised, and this then limits their transformative potential.

Beyond deliberative governance models, participatory design and co-design practices have been employed in urban planning to foster participation and improve collaborative outcomes. Participatory design engages diverse actors in identifying challenges, co-developing solutions and iteratively refining interventions (Frauenberger et al., 2018; Moore and Elliott, 2016). Co-design builds on these foundations by structuring engagement through workshops and focus groups in which participants collectively shape planning decisions (Munthe-Kaas, 2015; Taboada et al., 2024). While these methods broaden participation, several studies have critiqued their limited capacity to sustain collaboration and transformation beyond the design and implementation stages. Participatory processes are often short-term and dependent on facilitation, which can make them vulnerable to institutional priorities once design activities conclude (Blomkamp, 2018; Light and Akama, 2012; Ten Holter, 2022). In social-ecological contexts, co-design has also been shown to underrepresent non-human or ecological actors, because it often focuses predominantly on human needs and perspectives (Cooper, 2022; O'Donnell et al., 2025). These critiques reveal both conceptual and practical limitations across participatory and co-design approaches, which have yet to embed mechanisms for continued intermediation during implementation and governance. Building on this work, MOD addresses these challenges by extending participation beyond design into deployment and post-deployment activities, ensuring that collaboration among diverse human and non-human actors continues throughout decision-making processes. This approach enables the ongoing negotiation of responsibilities between top-down, bottom-up and mediating actors, offering a structured pathway for more inclusive and enduring participation in urban sustainability planning.

Despite advances in participatory urban planning, these approaches remain shaped by institutional structures and human-centred priorities, which limit their responsiveness to ecological and non-human considerations. Existing participatory approaches often privilege human-centred concerns, overlooking the complex interdependencies between human and non-human actors in urban environments (Clarke et al., 2019; Foth, 2025; Heitlinger et al., 2024). This limitation undermines the potential for truly inclusive and ecologically responsive planning processes, particularly as cities face increasing environmental challenges, from biodiversity loss to climate change. As urban sustainability planning increasingly acknowledges the interdependence of human and ecological systems, there is a pressing need for frameworks that bridge governance structures with grassroots initiatives while embedding more-than-human considerations into planning processes. To address this gap, the following section introduces the MOD framework, which offers a structured approach for fostering collaboration across top-down governance, bottom-up grassroots action and mediating actors.

The concept of “middle-out”, introduced by Kinchla and Wolfe (1979), describes a visual processing sequence where information converges from top-down and bottom-up channels. Over time, it has been adapted in various fields, including engineering, biochemistry and design, to balance hierarchical and grassroots dynamics in decision-making. For example, Costa and Ferrão (2010) used middle-out to enhance sustainability in product design, integrating top-down management priorities with bottom-up worker insights to foster innovation. Janda and Parag (2013) applied middle-out strategies to optimise energy performance in building design, influencing policies and client preferences from a central position. However, these early interpretations primarily addressed coordination within organisational or technical systems rather than participatory and collaborative approaches. They focused on balancing human actors within institutional hierarchies, with little consideration for sustained intermediation or ecological participation.

Building on these foundations, middle-out approaches have also been used to address complex urban and community challenges. For example, Biedermann et al. (2023) applied a middle-out approach for their PosterTalk project, a public survey platform designed to shift decision-making influence towards community-led participation. Unlike conventional surveys, the platform allowed community members to co-create questions, ensuring that public engagement reflected bottom-up concerns rather than institutional priorities. In the context of community support services, Dow et al. (2019) applied a middle-out approach to mediate between grassroots organisations and government actors, using civic technologies to capture lived experiences and support equitable participation, while overcoming bureaucratic constraints and austerity-driven policies. Similarly, Caldwell et al. (2021) investigated middle-out as a mechanism for engaging diverse stakeholders in city-making processes through two case studies in Mexico and Thailand. Using a combination of digital and physical media, they analysed how top-down decision makers and bottom-up community groups could collaborate in the “middle” to foster more inclusive and culturally grounded city-making initiatives. Together, these cases illustrate the versatility of middle-out approaches in fostering collaboration across varied urban and social domains. Yet, across these applications, middle-out has largely been employed as a descriptive lens for collaboration rather than as a structured design methodology. Few have detailed the processes that sustain collaboration and reflexivity once initial projects or interventions conclude.

The MOD framework builds on these earlier applications by adapting the concept into a structured design framework that facilitates collaboration and inclusion among diverse human and non-human actors. Middle-out is an approach that can facilitate greater levels of collaboration and inclusion by bringing together a diverse range of actors. It draws on the collective knowledge of top-down and bottom-up perspectives, enabling them to meet in the middle and jointly work outwards as a representative coalition (Fredericks et al., 2023) (Figure 1). Unlike other participatory models that often operate in controlled environments or conclude once a proposal is developed, middle-out embeds collaboration in real-world settings and continues throughout planning, implementation, and post-deployment (Fredericks et al., 2016). While grounded in principles of participatory design and co-design, middle-out extends these practices by supporting ongoing engagement during deployment and evaluation. Rather than ending with a product or policy, it emphasises continuous iteration based on lived experience, enabling outcomes to adapt to evolving needs and local conditions. This makes middle-out a valuable pathway for addressing gaps in urban planning and the implementation of urban sustainability strategies. It supports inclusive and adaptive outcomes that integrate both human and more-than-human considerations in complex socio-environmental contexts. MOD advances participatory design, co-design and middle-out approaches by embedding sustained collaboration within real-world urban systems, ensuring that design, implementation and governance remain interconnected processes rather than isolated stages.

We propose the MOD framework for the collaborative design, development and deployment of design and decision-making outcomes (Figure 2). Comprising four stages, the MOD framework brings together perspectives from different directions – top-down and bottom-up – into the middle of the process. Each stage builds upon the previous one, enabling the coalition to progress through the design process as follows:

1. Coalition – The first stage is to identify people who will form part of the coalition, which incorporates three distinct roles: (1) top-down, for example, representatives from government agencies, international organisations and private enterprises; (2) bottom-up, for example, representatives from community groups, grassroots entities, First Nations communities, culturally and linguistically diverse communities and local neighbourhoods and (3) mediators, those who act as facilitators or neutral parties without a direct stake in the design and decision-making processes.

2. Design – The second stage focuses on understanding the design context, including identifying the needs, wants and issues of both human and more-than-human actors. The coalition works together to define design objectives and develop a shared purpose.

3. Development – The third stage empowers the coalition to generate ideas, prototype concepts, initiatives and policies, and test these with each other. This also provides an opportunity to refine and reiterate the design requirements based on the data collected during the process.

4. Deployment – The fourth stage incorporates the deployment of the design intervention and facilitates decision-making within the coalition around the design topic. This continuous process involves the coalition in the post-deployment phase as new requirements, needs and issues emerge.

While the MOD framework establishes a structured process for mediating collaboration between top-down institutions, bottom-up communities and intermediary actors, its original formulation remains largely human-centred. Yet, as urban challenges increasingly intertwine ecological, technological and social systems, there is a pressing need to consider how non-human actors can be meaningfully integrated into participatory frameworks. The next section extends the MOD framework by incorporating concepts from multispecies justice and more-than-human design to address this gap and to explore how middle-out processes might engage a broader ecology of participants in urban planning and design.

This section outlines how the middle-out concept has been adapted to incorporate more-than-human and multispecies perspectives within participatory urban sustainability planning. Previous research on middle-out approaches has focused primarily on mediating relationships between institutional and community actors. The MOD framework extends this focus to recognise ecological, material and technological actors as participants in collaborative design and decision-making. This adaptation highlights MOD's capacity to bridge fragmented practices and provide a coherent scaffold for inclusive, multispecies governance.

There has been a growing shift in urban planning and design towards extending principles of equity and inclusion beyond human communities, recognising the value and agency of non-human actors in governance, decision-making and infrastructure development (Heitlinger et al., 2024; Luusua et al., 2017, 2023; Sheikh et al., 2021). This includes animals, plants, ecosystems and artificial entities such as robots or artificial intelligence (AI) systems. While the concept of multispecies justice (Celermajer et al., 2022) continues to inform some of these efforts, our focus is on integrating more-than-human considerations into participatory processes that can inform design and planning practice (Light, 2024; Metzger, 2016).

A growing body of research explores methods for integrating non-human voices into participatory frameworks. One such approach involves the development of non-human personas as design artefacts that represent the interests, needs and experiences of non-human entities within participatory design workshops (Tomitsch et al., 2021). These personas help human participants engage empathetically with non-human perspectives and inform design choices that better accommodate ecological needs. Another emergent practice is ecological monitoring, which leverages sensor networks and AI-driven analytics to track environmental variables, enabling urban planners to incorporate real-time ecological data into decision-making processes (Sheikh et al., 2023a). This allows for adaptive management strategies that respond dynamically to changes in ecosystems, fostering urban environments that are resilient and responsive to non-human concerns.

Participatory environmental decision-making approaches also increasingly recognise the importance of collaborative governance structures that include governments, communities, scientists and First Nations peoples (Graham et al., 2024; Page and Memmott, 2021; Sheikh et al., 2023a). These collaborative approaches draw upon diverse knowledge systems to inform policies that align human development goals with ecological preservation. For instance, Indigenous land management practices, often based on a deep, place-based understanding of ecosystems, offer valuable insights into sustainable urban governance and land-use planning. Together, these examples illustrate a growing yet fragmented landscape of more-than-human engagement methods. MOD responds by offering a unifying process framework capable of connecting such efforts across institutional, community and ecological domains.

Artificial non-human artefacts, such as robots and autonomous systems, present new opportunities and challenges for integrating more-than-human considerations in urban spaces (Foth and Caldwell, 2018; Sheikh et al., 2021). When designed with both ecological and human needs in mind, these systems can mediate human–environment interactions by adjusting indoor conditions such as lighting, air quality or thermal comfort. For example, in office environments, robotic furniture can support the wellbeing of both human occupants and indoor ecosystems (Loh et al., 2023, 2024). However, such systems also raise questions about how non-human interests are represented in design and governance processes (Metzger, 2016).

Despite these promising developments, significant challenges remain in representing non-human perspectives within participatory design approaches. A persistent limitation is the human-centred bias embedded in most decision-making processes, which often marginalises non-human needs unless they directly align with human priorities (Clarke et al., 2019; Loh et al., 2020). Even methods such as non-human personas and ecological monitoring, while valuable in highlighting non-human concerns, ultimately rely on human interpretation, which risks oversimplifying or misrepresenting non-human experiences.

Metzger (2016, p. 594) ponders, “How can planning methods be developed that force those involved to face up to how their practices which inevitably aim at enabling the flourishing of some by necessity also imply the radical exclusion (even eradication) of others, and the consequences that come with such an insight?” His question corroborates the need for frameworks and approaches that not only acknowledge these exclusions but actively seek to mitigate them by creating more equitable conditions for multispecies participation.

The MOD framework directly addresses this gap. It extends the original model, which was designed to connect top-down and bottom-up processes, by incorporating non-human and multispecies considerations as integral to participatory practice. This adaptation enables more continuous forms of intermediation, allowing ecological, technological, and material actors to shape decision-making rather than be represented symbolically. By embedding more-than-human concerns within its coalitions, MOD creates negotiation spaces where the interests of both humans and non-humans can influence design outcomes. In doing so, it provides a timely and generative framework that aligns with global efforts to integrate multispecies justice into urban governance and sustainability practices.

The following section presents four vignettes that illustrate how the MOD framework operates in practice across contrasting urban contexts. Each vignette is structured around the MOD process to show how coalitions worked together to negotiate and intermediate both human and non-human concerns (Frauenberger et al., 2018; Teli et al., 2022). Through these examples, the article demonstrates the framework's versatility and its potential to support more equitable and multispecies urban futures.

The selection of the vignettes in this section was based on their ability to demonstrate the different stages of the MOD framework across diverse urban contexts. These vignettes are drawn from some of our broader body of work examining urban projects that engage both human and non-human actors in sustainability and planning processes (Foth et al., 2017; Gemperle et al., 2023; Nguyen and Vande Moere, 2024; Tomitsch et al., 2025). Each vignette is retrospectively structured within the MOD framework to highlight how coalitions formed, operated and adapted throughout the design and decision-making process.

The vignettes were intentionally selected to represent diverse and contrasting contexts rather than to demonstrate internal coherence or direct comparison. This approach, similar to a maximum variation sampling strategy (Suri, 2011), was used to illustrate the versatility and adaptability of the MOD framework across different domains of urban sustainability and more-than-human design. While the framework diagram depicts dynamic relationships and practices, it is not intended as a linear or prescriptive process. Collectively, the vignettes provide an exploratory account of how MOD principles are applied in situated contexts, demonstrating the framework's flexibility and relevance across multiple scales of urban practice. While the specific findings and evaluations of these projects are detailed in their respective publications, this section offers a comparative perspective to examine shared principles, opportunities and challenges of applying the MOD framework in practice.

The first vignette presents an urban community garden project located at The University of Sydney campus in Sydney, Australia (Gemperle et al., 2023). The project investigated the role of urban community gardens as co-existing spaces for human and non-human occupants. Conducted across three workshops, the study implemented participatory methods for (1) understanding how gardening communities interact with their environment, (2) identifying non-human entities and incorporating their perspectives into the design process, (3) developing and prototyping an urban interface – an installation incorporating digital and physical elements for human and non-human occupants and (4) testing and deploying the interface to foster co-existence within the garden.

The coalition comprised representatives from multiple levels. Top-down actors included university officials who contributed towards operational policies and provided permissions for the garden's use. Documents provided by the local municipality outlined regulatory frameworks and self-management standards for community gardens. Bottom-up actors included university students seeking outdoor social connections and local community members with gardening expertise. Their knowledge and ongoing participation helped shape the project's direction.

Non-human entities were considered through persona-based design methods (Pace et al., 2025) developed using publicly available ecological data. These personas enabled participants to engage with the garden from a non-human perspective. Facilitating the coalition as the mediator was an undergraduate student investigating the use of participatory design methods to enhance user experiences in community gardens through digital and physical interactions. For this study, the student remained impartial to ensure that the coalition operated smoothly and free from direct bias and vested interests.

The first workshop introduced participatory mapping and photography exercises to help participants document and interpret human and non-human interactions within the garden. These activities identified key areas of interaction and highlighted spaces frequently used by non-human entities. Participants were then tasked with re-exploring the garden through assigned non-human personas, prompting reflections on ecosystem dynamics and species-specific needs (Figure 3).

Insights from these exercises informed the second workshop, where the coalition co-created a conceptual model for an urban interface that would enhance human and non-human interactions. Prototyping involved physical and digital engagement elements, including an interactive display, a non-digital discussion board, a wet space sink for processing and measuring produce, hanging planters for native seedlings and a nesting area for blue-tongued lizards (Figure 3). The coalition worked iteratively, testing and refining the prototype based on observations and feedback.

The final workshop involved the deployment of the urban interface in a central and accessible location within the garden. The coalition continued to meet weekly, adjusting elements based on emerging needs and evolving dynamics. This iterative adaptation ensured that the urban interface remained aligned with both human and non-human users' needs, fostering a continuously evolving engagement process rather than a static intervention.

This vignette demonstrates how MOD can bridge top-down urban sustainability planning with bottom-up community participation, ensuring that non-human perspectives are embedded into decision-making processes (Salama et al., 2024). By integrating participatory design with ecological considerations, the project highlights both the opportunities and challenges of designing for multispecies co-existence in urban spaces. One key challenge was ensuring that insights from non-human personas meaningfully informed design decisions. While persona-based methods encouraged participants to engage with non-human perspectives, the design process remained iterative, requiring ongoing reflection and adaptation to integrate these insights into actionable interventions. This reflects broader complexities in urban sustainability planning, where balancing human and non-human needs requires adaptive, multi-actor approaches rather than “one size fits all” design solutions.

Furthermore, the role of the mediator proved critical in balancing the diverse priorities of regulatory bodies, community members and ecological considerations. Their ability to facilitate dialogue and maintain engagement throughout the deployment phase highlights the importance of ongoing governance structures in sustaining multispecies-inclusive urban planning initiatives. By embedding human and non-human considerations into participatory processes, this vignette reinforces the potential for MOD to reshape urban spaces as sites of co-existence rather than human-dominated landscapes. This highlights the potential of MOD to support urban planning practices that respond to both human and non-human needs, fostering cities that prioritise biodiversity, ecological health and co-existence.

The second vignette focuses on a flooding event, which involved a coalition to decide whether to rebuild or relocate. The vignette is based on a workshop held in conjunction with an academic conference. The workshop implemented user enactment (Seland, 2009) to role-play the coalition and decision-making process of MOD. Based on a real flooding event, we created a fictional scenario involving two suburbs: Casuarina Waters – the area affected by the floods, and Casuarina Heights – offering land for relocating residents and businesses. The vignette was part of a larger research project investigating the use of AI to support collective decision-making and to represent non-human perspectives (Tomitsch et al., 2025).

The 12 workshop participants were assigned roles representing a balance of top-down and bottom-up perspectives: local artist, state member of parliament, mayor, water engineer, emergency services coordinator, city planner, environmental activist, president of “Save Casuarina Waters”, flood-affected resident and cafe owner. The roles were prepared by the research team beforehand based on their involvement in community engagement processes and captured in the form of proto personas (Jain et al., 2019). Two researchers role-played community engagement workshop facilitators.

To integrate a non-human perspective in the coalition, we created an AI agent, referred to as “Sage”. Sage was embodied through a physical artifact housing a speaker and featuring an LED light ring to indicate when Sage spoke (Figure 4). We used a wizard-of-Oz (Dow et al., 2005) approach to simulate the AI functionality, which involved a researcher controlling the speech interaction from another room (invisible to the participants). We prepared a series of statements and generated on-the-spot responses using ChatGPT. The researcher controlling Sage followed a predefined protocol developed from prior pilot workshops. Prompts were based on background research about the local river ecosystem and pre-tested through example statements generated with ChatGPT. During the workshop, spontaneous participant questions were relayed to ChatGPT and paraphrased by the researcher to maintain conversational coherence. All interactions were audio-recorded, and researcher interventions were logged in a shared worksheet. To reduce facilitator bias, the human facilitators were instructed not to steer participant discussion and to let Sage respond only when prompted or contextually appropriate. This setup ensured transparency and consistency in how the non-human voice was simulated.

The MOD workshop consisted of three activities. During the first activity, participants (while role-playing their assigned characters) recorded the benefits and disadvantages of rebuilding Casuarina Waters using sticky notes. To facilitate this activity, the research team created a paper prototype for an A0-sized interactive map (Figure 4). Participants were able to “swipe” to change between the two suburb maps and “click” on an area in the map to add a sticky note. The second activity focused on recording the benefits and disadvantages of relocating to Casuarina Heights.

In the third activity, participants discussed their recorded notes. Sage participated in this conversation, adding the perspective of the river ecosystem. Sage's contribution was initially prompted by one of the workshop facilitators. Subsequently, participants could ask Sage directly and Sage, at some points, interjected to add the wetland's perspective. Throughout this activity, Sage was positioned at the table near the interactive paper prototype (Figure 4).

At the conclusion of the workshop activities, the workshop facilitators asked participants to express their position on rebuilding versus relocating by placing themselves along a line across the room. This allowed participants on the fence to place themselves somewhere along the spectrum. Given the speculative nature of the workshop, the vignette does not include development and deployment phases.

The vignette investigated role-playing to facilitate a MOD process for a complex scenario, where it might be otherwise difficult to bring actors with the lived experience together. At the same time, the vignette evaluated the use of AI to include a non-human voice in the coalition. To collect feedback on both aspects, the researchers conducted a focus group discussion after the workshop activities. Participants found the enactment approach effective, a finding supported by our observations of participants augmenting the stories of their characters and staying in character throughout the activities. Sage was considered to help keep nature's perspective in mind; as one participant stated: “The way it sort of, every now and then speaks to us … it reminded us of this sort of agency of nature.” The use of AI to represent a non-human actor further encouraged MOD coalition participants to reflect “whether [the wetland] has a vote” and contemplating the “different species [living] in the wetlands.” While there are benefits of this approach, there are also potential risks that need to be considered, such as biases, associated with the use of personas and AI.

Our third vignette examines a design-led research initiative addressing sarcoptic mange in bare-nosed wombats (Vombatus ursinus), a disease threatening the survival of this iconic Australian marsupial (Foth et al., 2017, 2026). Mange leads to severe skin lesions, infections, and, if untreated, death (Ringwaldt et al., 2023). While treatment methods exist, they are often ineffective due to logistical constraints, treatment resistance and difficulties in ensuring consistent application in the wild (Driessen et al., 2021). This project focuses on designing and deploying hospital quarantine burrow enclosures – purpose-built structures improving treatment outcomes for rehabilitating wombats.

Based at one of Australia's largest wombat sanctuaries near Canberra, this initiative responds to the need for more effective, humane interventions. The sanctuary provides critical care for wombats affected by mange, habitat destruction, vehicle collisions and human-wildlife conflict. Its approach is distinct in its cohabitation model, where human caregivers form close relationships with wombats, generating deep insights into their behaviour and welfare (Fieuw et al., 2022). The project builds on this more-than-human participatory design to develop burrow enclosures that prioritise wombat wellbeing while enhancing treatment efficacy.

Comparable challenges of human–animal coexistence are evident globally. Examples include interactions with macaques in South and Southeast Asian cities, hyenas and baboons in parts of Africa, and urban-adapted species such as foxes, raccoons and wild boar in Europe and North America. As a result, planning, governance and everyday practices must continually negotiate shared spaces between humans and non-humans.

Traditional wildlife disease management follows a top-down model where government agencies implement programs based on epidemiological models. These interventions often exclude the experiential knowledge of wildlife carers and volunteers, who interact with affected animals daily (Englefield et al., 2018, 2019). These carers provide vital insights into wombat behaviour, treatment responses and stress indicators – perspectives often overlooked in scientific and policy-driven decision-making.

This project employs MOD, integrating participatory action research and participatory design (Foth and Axup, 2006; Foth and Brynskov, 2016; Frauenberger et al., 2018). A coalition of design researchers, wildlife carers, veterinarians and conservation advocates collaboratively develops solutions that balance scientific expertise with hands-on, situated knowledge (Foth et al., 2007). Wildlife carers play a pivotal role in this coalition. Unlike field researchers who observe from a distance or policymakers who work with aggregated data, carers engage in daily, embodied interactions with individual wombats. Their ability to recognise subtle behavioural cues (e.g. stress, discomfort or recovery) directly informs the design of quarantine burrow enclosures. Their involvement ensures that enclosures are adaptable, functional and aligned with rehabilitation needs.

The MOD coalition also integrates ecological and infrastructural considerations, embedding burrow enclosures within the sanctuary's landscape design. Factoring in elements such as drainage, vegetation, and microclimatic conditions, the project aims to support both individual wombats and the broader environmental systems sustaining them.

The quarantine burrows were iteratively prototyped and tested on-site with the objective to develop enclosures that facilitate mange treatment while minimising stress and maximising wombat autonomy. Unlike conventional rehabilitation enclosures, these designs draw on more-than-human design methodologies (Sheikh et al., 2023).

The enclosures feature a subterranean burrow component, constructed using large drainage trenches that mimic natural wombat burrows. Above-ground shelters enable caregivers to monitor wombats discreetly while providing additional protection. In some cases, natural burrows created by healthy wombats were enhanced with complementary above-ground structures (Figures 5 and 6). A key innovation includes hinged roof sections and removable panels, allowing caregivers to administer treatment without disturbing the animal's refuge. This permits topical medications or injections without stressful physical handling or capture. The iterative design process involved multiple prototype refinements, with wildlife carers and veterinarians providing continuous feedback.

This vignette highlights the value of a MOD coalition in bridging top-down scientific knowledge and bottom-up experiential expertise by combining wildlife carers' intimate knowledge of wombat behaviour with veterinary expertise. The project demonstrates how non-human needs can be meaningfully incorporated into urban sustainability planning through place-based collaboration. Ecological considerations, such as burrow placement and habitat integration, further illustrate MOD's capacity to embed more-than-human concerns in decision-making. By centring wombat wellbeing, the project challenges anthropocentric assumptions in conservation and provides a practical model for recognising non-human agency within design and landscape planning.

Finally, this vignette affirms the importance of iterative, participatory design in supporting long-term care. Ongoing collaboration between carers, veterinarians and researchers enabled adaptive responses that reflected the evolving needs of both wombats and their surrounding ecosystems.

The fourth vignette investigates how robotic furniture can dynamically reconfigure open-plan office environments to mediate between human, technological and ecological needs (Nguyen and Vande Moere, 2024) (Figure 7). Expanding on research into adaptive façade control (Le et al., 2022), this vignette examines how robotic systems respond to acoustic, visual and privacy-related disruptions, while also considering indoor environmental conditions and broader sustainability goals.

Beyond human comfort, office environments must also accommodate non-human concerns, from sustainability targets aimed at reducing environmental impact to local factors such as plant health and microclimatic conditions, including temperature, humidity and air quality (Poncelet et al., 2025). These overlapping demands create spatial and functional tensions. For example, an automated sunshade might be activated to reduce solar heat gain for energy efficiency, yet office workers may prefer access to natural light and outdoor views (Tabadkani et al., 2022). At the same time, indoor plants contribute to wellbeing through biophilic effects and require consistent sunlight exposure (Yin et al., 2018). These scenarios illustrate the need to negotiate between ecological needs, occupant preferences, and environmental performance. Aligned with the MOD framework, this vignette explores how robotic furniture can act as a mediator within a typical office environment, responding to the complex interplay between these diverse factors.

This project established a coalition to guide the design and deployment of robotic furniture systems within an adaptive office environment. Top-down actors include a research consortium focused on workplace automation and management. Bottom-up actors included office occupants, who contributed first-hand insights into diverse comfort needs, shaping design considerations based on their lived experiences. Mediators included architectural designers, occupational health specialists, comfort researchers, and economists. Their role was to evaluate how spatial, social and environmental conditions affected human wellbeing, energy use and ecological outcomes.

The project began with an observational study to identify needs, constraints and resources in the open-plan office environment. This stage examined how occupants navigated the space, how disturbances arose and how natural resources such as daylight, outdoor views and indoor greenery could be leveraged to mitigate discomfort. This was important, as multi-domain human comfort research (Bavaresco et al., 2024) indicated that discomfort in one area, such as noise, could be perceived as less severe if other conditions, such as visual comfort, were optimised.

Insights from this study informed a series of iterative co-design sessions. Participants engaged with robotic adaptation strategies through virtual reality simulations and physical prototypes. This process ensured that both human and ecological factors were explicitly represented in the design. During the final stage, a pilot deployment of the robotic furniture was undertaken to evaluate how well the system managed trade-offs between occupant preferences, technological responses and ecological parameters. A custom interface was also developed to make robotic decision-making visible and adaptable to changing conditions, supporting post-deployment governance.

This vignette demonstrated how MOD could be adapted to architectural contexts while incorporating ecological and more-than-human considerations. The coalition played a critical role in balancing tensions between occupant preferences, sustainability imperatives and economic constraints. As the design process progressed, the robotic furniture evolved from a tool guided by human governance to an autonomous system capable of negotiating the needs of humans, technologies and ecosystems.

One key challenge involved robotic interventions that served one group while disadvantaging another. For example, robotic partitions could increase human privacy but reduce light exposure for indoor plants. This illustrated the need for transparent systems that could mediate such trade-offs in real time. The custom interface functioned as a governance mechanism, visually displaying these decisions to foster trust and engagement.

Ultimately, this vignette showed how MOD could enable more inclusive infrastructural responses to sustainability challenges by embedding more-than-human considerations into the everyday operations of indoor environments. It demonstrated the potential of adaptive robotic systems to act not just as responsive technologies but as participants in multispecies urban governance.

We have proposed the MOD framework as a methodological response to the persistent challenges of embedding more-than-human considerations into participatory urban sustainability planning. The four vignettes highlight that while MOD holds significant promise for facilitating equitable collaborations across human and non-human divides, this approach remains work-in-progress. Several open questions persist regarding how to operationalise principles of multispecies justice, integrate more-than-human perspectives meaningfully and navigate the socio-political complexities of urban governance. Building on the analytical insights presented in Section 5, this section synthesises observations across the four vignettes to identify their broader theoretical, methodological and practical implications. Here, we discuss how MOD can inform the design of participatory governance processes, the integration of non-human and AI-supported actors, and the ongoing adaptation of urban sustainability practices. In this section, we reflect on three key themes that have emerged: (1) the need for intermediaries; (2) the importance of institutioning and (3) the necessity of advancing governance mechanisms that extend from legislation to obligation.

The vignettes reveal that coalitions within MOD are dynamic and often imperfect. Different actors operate from top-down, bottom-up or middle-out positions, bringing diverse scopes of power, interests, motivations and values. Human participants such as government officials, community activists and corporate representatives often hold distinct priorities that must be negotiated within the coalition. At the same time, non-human actors such as ecosystems, species and robotic systems possess their own forms of agency. This agency is typically mediated through human proxies such as researchers, advocates or AI trained on human-generated datasets.

This complexity requires ongoing intermediation to facilitate negotiation and understanding across diverse actors. As Teli et al. (2022) suggest, intermediaries play a critical role in institutioning and commoning, serving as bridges between divergent interests and fostering shared understanding. In the urban gardening vignette, intermediaries helped balance the desires of human participants with ecological concerns, ensuring that non-human perspectives were not merely tokenistic but actively shaped design outcomes. Similarly, in the wombat rehabilitation vignette, wildlife carers served as essential intermediaries, translating the embodied experiences of wombats into actionable design insights for researchers and policymakers.

However, intermediaries themselves are not neutral actors. They carry biases, personal beliefs and institutional constraints that shape their mediation (McQueenie et al., 2024). Recognising these biases is crucial for fostering transparency and reflexivity within MOD processes. By embracing this complexity, MOD enables coalitions to navigate tensions and limitations collectively, fostering a more holistic, transdisciplinary understanding of urban systems.

Coalitions formed through MOD cannot operate in isolation from larger institutional frameworks. While grassroots and community-driven initiatives are vital for fostering inclusive participation, they often lack the resources, legitimacy or reach to enact systemic change on their own. As Lodato and DiSalvo (2018) argue, institutional constraints can limit the scope of participatory design practices, particularly when public institutions maintain control over decision-making processes.

The flood response vignette illustrates this tension clearly. Despite the coalition's efforts to integrate non-human perspectives, which were embodied through the AI agent Sage, the final decision-making authority rested with institutional actors such as government officials and planners. This reveals the importance of engaging with and leveraging institutional processes to achieve broader impact.

Drawing on Huybrechts et al. (2017), we argue that institutioning, defined as the process of embedding collaborative design practices within formal institutions, is essential for ensuring that the outcomes of MOD coalitions extend beyond isolated interventions. Institutions, whether governmental, corporate or non-profit, possess the capacity to enact large-scale change through policy-making, regulation and resource allocation. Therefore, effective MOD coalitions must strategically connect with these institutional entities to amplify their impact, so as to ensure that more-than-human perspectives are integrated into mainstream governance and policy frameworks (Foth and Turner, 2019).

The robotic furniture vignette offers a promising example of how MOD coalitions can engage with institutional processes to support more-than-human considerations. By involving workplace management consultants, building automation specialists and researchers in human-robot interaction, the project positioned itself to influence corporate policies around office design and sustainability practices.

Finally, the question of governance extends beyond merely enacting legislation – it involves fostering an ethical obligation to consider the rights, needs and agency of non-human actors.

Metzger prompts us to continually confront the ethical trade-offs inherent in the work of a MOD coalition. The MOD framework contributes to this discourse by providing a participatory structure that brings these tensions into focus and supports dialectical deliberation across human and non-human divides (Frauenberger et al., 2018). We recognise that merely assembling a coalition is not sufficient for fostering more-than-human perspectives. True equity requires advocacy, activism and systemic reforms that address the deep-seated anthropocentric biases embedded in existing governance structures (Sheikh et al., 2023a, b).

The wombat project vignette illustrates this point. While the design of quarantine burrows reflects a significant step toward recognising non-human agency, broader legislative and institutional reforms are needed to ensure the long-term protection of wombats in the wild. Similarly, the urban gardening vignette highlights how grassroots initiatives must connect with municipal governance frameworks to scale up their impact and embed more-than-human considerations into urban sustainability planning.

Critically, governance must be approached as an ongoing, adaptive process. Policies that support more-than-human considerations need to be flexible enough to respond to evolving ecological conditions, technological innovations and shifting social dynamics. By fostering continuous negotiation between human and non-human interests, the MOD framework offers a pathway for embedding this adaptive governance into urban planning practices.

This article has presented MOD as a conceptual and methodological approach for integrating human and more-than-human perspectives in urban sustainability planning. Through four vignettes, we have illustrated how MOD can facilitate collaboration across top-down, bottom-up and middle-out actors, fostering inclusive and adaptive design processes that extend beyond traditional participatory models. The vignettes were intentionally selected to demonstrate the framework's versatility across diverse and contrasting settings, ranging from designing community-based interventions to influencing governance processes. Collectively, they show how MOD operates as a transferable framework capable of adapting to varied institutional, social and ecological settings.

For designers, planners, and all levels of government, MOD offers a structured yet flexible approach for embedding community engagement and multispecies considerations throughout the design, development and deployment of urban projects and policies. The framework can be applied across multiple scales: (1) at the neighbourhood scale to support collaborative interventions and shared infrastructure; (2) at the city scale to mediate between agencies and communities in sustainability transitions and (3) at the national scale to inform legislative, procedural and institutional reforms. Its emphasis on coalition building and sustained intermediation provides a foundation for bridging governance silos and strengthening accountability across decision-making processes.

Integrating MOD with existing urban planning systems presents opportunities for governments to extend current frameworks such as strategic planning, environmental assessment and nature-based solutions towards more-than-human and ecologically responsive outcomes. At a local level, MOD can inform community engagement guidelines that strengthen partnerships between local municipalities and communities. At state and national levels, it offers pathways to embed inclusive and adaptive principles within policy development, infrastructure programs and sustainability legislation.

For researchers, MOD presents both methodological and theoretical opportunities. Future work should investigate how the framework can be applied in ongoing governance processes, tested longitudinally across different cultural and ecological contexts, and evaluated for its capacity to sustain collaboration over time. Methodological challenges remain around representing non-human agency, balancing power within coalitions and assessing long-term institutional change. These questions point to future opportunities for strengthening the framework's conceptual and practical contributions.

As urban environments become increasingly shaped by digital and data-driven systems, emerging technologies such as artificial intelligence will play growing roles in mediating participation and decision-making. MOD highlights how such technologies can act not only as analytical tools but as mediating participants within collaborative processes that support more inclusive and accountable urban governance.

By positioning collaboration as an ongoing, adaptive process that extends from design to governance, MOD reframes how urban sustainability planning can integrate human and more-than-human perspectives. Its continued development offers a practical foundation for governments, designers and communities seeking to shape urban futures that are more ethical, resilient and ecologically attuned across species, systems and scales. MOD therefore offers a scalable and actionable framework for inclusive, multispecies urban governance.

The studies reported in Vignettes 1 and 2 were approved by the University of Sydney Human Research Ethics Committee (#HE000675). The study reported in Vignette 3 was approved by the Queensland University of Technology Human Research Ethics Committee (#1913) and the University Animal Ethics Committee (#2006). The study reported in Vignette 4 was approved by the KU Leuven Human Research Ethics Committee (#IDN/22/003).

In line with Emerald's publishing ethics and principles of AI usage, the authors disclose that OpenAI's ChatGPT (GPT-5) was used under author supervision to refine the manuscript's structure, clarity, and language. The tool was also employed in Vignette 2 to simulate the AI agent “Sage” as part of the research process. All intellectual content, analysis and interpretation are the work and full responsibility of the authors.