A modular microbial–hydroponic system for circular urban plant cultivation

Urban environments need low-input cultivation systems that integrate plants into buildings while improving resource efficiency and circularity. This report presents the technical validation of the first operational Microbial–Hydroponic (Mi-Hy) prototype, known as SPIKA, which combines microbial fuel cells (MFCs), vertical hydroponics, and continuous sensing within an architectural-scale installation. The system was stress-tested for seven months (May–December 2025) in a demanding public setting at the Triennale Milano to evaluate its operational stability and robustness. A vertical hydroponic tower and an MFC stack were operated as parallel subsystems under a low-input, recirculating regime, supporting nine ornamental and aromatic plant species. Continuous sensing and imaging provided a longitudinal phenotypic record, enabling a semi-quantitative assessment of plant performance, including growth, flowering, and visual health. Results show stable physicochemical conditions (mean pH 7.1 ± 0.5; mean EC 510 ± 35 µS cm -1) and consistent plant growth without interruption. Over seven months, the system required 35 water refills (52.5 L) and 12 nutrient additions (190 mL). The 12-module MFC stack exhibited open-circuit voltages of 550–620 mV per module. These findings validate the operational reliability of the Mi-Hy framework and establish a baseline for future prototypes integrating nutrient recovery into hydroponic cultivation, advancing toward a fully coupled circular metabolism.

Cities increasingly face climate instability, resource constraints, and biodiversity decline, intensifying demand for cultivation approaches that can be embedded into the built environment while minimizing external inputs (Barker et al., 2024; Wang et al., 2025). Building-integrated planting can provide ecosystem services and support more resilient urban landscapes, but sustainable deployment requires systems that reduce water and fertilizer use, recover nutrients from waste streams, and maintain plant performance under constrained, variable conditions (Ali et al., 2024; Decardi-Nelson and You, 2024; Ranasingha et al., 2025). Circular cultivation frameworks have therefore gained attention as practical pathways to integrate plant production with resource recovery in urban settings (Lin et al., 2025). Microbial electrochemical technologies can convert organic substrates into electricity and transform nutrient-rich wastes into recoverable resource streams (Kong et al., 2024; Walter and Popuri, 2025). In particular, microbial fuel cells (MFCs) offer a route to couple waste processing with low-input plant cultivation when integrated with hydroponics (Nath and Ieropoulos, 2025). Within a circular bioeconomy framework, the MFC can be conceptualized as a “prosthetic rhizosphere”—a controlled microbial compartment that functions as a core nutrient recovery and transformation unit, enabling the circular metabolism of buildings by converting waste streams into horticultural resources. However, long-term demonstrations outside laboratory conditions remain limited, especially in settings where systems must operate reliably with minimal intervention while remaining accessible for public viewing. Additionally, most work emphasizes food crops, while the potential of circular microbial–hydroponic systems to support ornamental and aromatic biodiversity relevant to urban landscape design has been less explored.

To address this gap, the Mi-Hy (Microbial–Hydroponic) is develops a modular cultivation platform combining microbial electrochemistry, hydroponics, and continuous sensing to reduce water use, recover nutrients from waste, and stabilize plant growth under low-input operation. A core design principle is the development of a controlled microbial compartment (the “prosthetic rhizosphere”) intended to generate usable nutrient streams for plant growth while supporting system autonomy. This manuscript reports the technical validation of the first operational Mi-Hy prototype—a critical prerequisite before controlled experiments on crop performance can be undertaken. In urban horticulture, the transition from laboratory to public, real-world environments is a significant research gap. We argue that documenting uninterrupted seven-month operational stability in a non-controlled setting constitutes a form of validation for system integration and robustness, establishing the necessary baseline before subsequent crop-yield trials. The seven-month public exhibition at the Triennale Milano provided precisely this validation. Here we present the first operational Mi-Hy prototype, the Structural Protection for Interdependent holobIont Assembly (SPIKA), deployed as a living installation at the Triennale Milano from May to December 2025 within the exhibition (Triennale Milano, 2025; Armstrong, 2025; Colomina and Wigley, 2026). SPIKA integrates three functional elements: (i) an MFC stack operating on artificial urine to characterize bioelectrochemical performance, (ii) a recirculating vertical hydroponic tower for plant cultivation, and (iii) a sensing and imaging layer for continuous monitoring of solution conditions and plant development. In this first prototype, the MFC and hydroponic loops were operated in parallel rather than hydraulically coupled, enabling baseline assessment of long-term stability and operational requirements under exhibition conditions. This parallel configuration was a deliberate design choice to establish a validated baseline for the system’s fundamental feasibility and robustness before introducing the additional complexity of hydraulic integration in future iterations. The objectives of this study were to (1) document the design and operation of SPIKA as an architectural-scale demonstrator, (2) evaluate the stability of the hydroponic subsystem and the feasibility of sustaining multiple ornamental and aromatic species under low-input recirculation using semi-quantitative performance metrics, and (3) establish a validated baseline for subsequent prototypes that will introduce MFC-derived nutrient streams into the hydroponic loop. The data presented—continuous sensor readings, photographic records of plant development, and the uninterrupted operation of all subsystems over seven months—constitutes a form of validation of the system’s fundamental feasibility and robustness. This data establishes a critical baseline and indicates that the prototype demonstrates the necessary technical stability for the next stage of hypothesis-driven experimentation, where MFC-derived nutrients will be integrated into the hydroponic loop. By reporting system stability, monitoring performance, and cultivation outcomes over seven months, this work provides practical evidence for integrating circular microbial–hydroponic infrastructures into building-integrated planting and climate-adaptive urban landscape applications.

Prototype (SPIKA)

Architecture conceptualization

The SPIKA prototype is the second-generation, architectural-scale manifestation of the Mi-Hy project’s core principles. It builds upon the foundational research of the first Mi-Hy prototype, the XENO (Xenomorphic Energy Necromantic Orchid), exhibited in Brussels and Eindhoven in 2023. XENO was a cybernetic proof-of-concept that integrated a Winogradsky column with a robotic actuator, demonstrating the harvesting of bioenergy from microbial metabolism to drive kinetic motion (Armstrong et al., 2023). SPIKA evolves this premise from a singular, notionally powered device into an integrated, multi-metabolic system, spatially organized to enable the circular cultivation of multiple plant species at an architectural scale.

The architectural objective was to materialize a vision of a circular building core that directly responded to the 2025 Milan Triennale’s curatorial theme: “How do forms follow bacteria?” (Milan Triennale, 2025; Colomina and Wigley, 2025b). The design takes the bacteriophage as its formal and conceptual archetype. Critically, it reframes infection not merely as a pathogenic event, but as an evolutionary driver. In biology, viral infections have been a source of evolutionary transformation; for instance, the genetic instructions from an ancient retrovirus were co-opted to enable the evolution of the mammalian placenta, giving rise to an entire new class of animals (Chuong, 2013). In this spirit, SPIKA was conceived as a symbiotic architectural infection of the building’s central services. This formal intervention aims to inject a new set of circular, replicating instructions into the host structure, compelling it to transform from a static consumer of resources into a self-sustaining ecological producer. Unlike passive sustainability policies, this approach constitutes an active, operational reprogramming of building metabolism, ensuring Green Deal principles become the core prion-like logic—a persistent, transmissible code—of its functions.

Designed for public engagement, SPIKA integrates three key functions: it performs (cultivates plants), protects (the enclosed biota), and communicates (its insistent ecological ethos). Its primary morphology—a six-meter-tall vertical tower—derives from the structural logic of a bacteriophage imagined at the scale of a building core. The external “tower of thorns” (Figure 1) is inspired by microbial defense and connective structures, translating the protein nanowires (pili) of electroactive bacteria into a dramatic, protective exoskeleton. This form asserts the fragility and value of the enclosed biological system, positioning it as an entity requiring and deserving protection within the public sphere. It extends the narrative of a self-contained phage and ‘prion’—an indestructible, transmissible instruction set—into an architectural scale facility for broader uptake, where the protective structure safeguards a productive, circular metabolic core.

Internally, this exoskeleton serves as a functional scaffold, vertically zoned around the core Mi-Hy processes. This internal organization translates the microbial-hydroponic synergy—first demonstrated at a centimeter scale in the XENO prototype—to an architectural meter-scale structure:

1. The Microbial-Electrochemical Engine (Lower Zone): This foundational section houses arrays of microbial fuel cells (MFCs) and their associated nutrient management systems. Here, organic waste is transformed into electrical energy and metabolized nutrients, rendering the building core independent of fossil-fuel-based fertilizers.

2. The Cybernetic Mediation & Bioreceptive Interface (Middle Zone): The central section contains the electronic controllers for system regulation. Outward-facing, this zone is clad with the textured, pre-inoculated bioreceptive panels. Their surfaces are designed to provide microniches that support the growth and persistence of the inoculated microorganisms, while also promoting spontaneous ecological colonization by the gallery’s ambient microbiota, effectively realizing the curatorial concept of the space as an “architectural-scale Petri dish” (Armstrong, 2025). This design extends the system’s functionality beyond its engineered core, positioning natural succession as a tangible architectural strategy.

3. The Hydroponic Cultivator (Upper Zone): Dedicated to plant cultivation, the upper part features LED grow lights and a central vertical hydroponic column. This enables the precision cultivation of plants within a recirculating nutrient loop.

Thus, SPIKA embodies an alternative vision: an organic building core that functions as a foundational, productive unit. It establishes a new approach for building-integrated cultivation, with potential for aggregation and scaling into larger urban farming systems, while making the imperative of microbial collaboration tangible in architectural form, function, and narrative.

Microbial fuel cell unit

The internal processes of a Microbial Fuel Cell (MFC) have been described in detail by Santoro et al. (2017). The MFC system exhibited at SPIKA utilized a compact stacking design in which the power of an individual MFC was scaled up by increasing its surface area to volume ratio. In a single modular tray, several cylindrical porous membranes are placed, each enclosing the air-breathing cathode half-cell. This cathode is composed of activated carbon blended with PTFE (80:20 wt%) pressed onto stainless-steel mesh and positioned tightly against the inner ceramic wall to support oxygen reduction. The anode half-cell surrounding the membranes consisted of a combination of carbon fiber veil (30 g m -2; surface area 945 cm²) and ceramic granules enclosed in a stainless steel mesh to facilitate microbial growth (Nath and Ieropoulos, 2025). Finally, artificial urine is supplied to the tray as the MFC feedstock creating a hydraulic connection between these cylinder units and effectively connecting them in parallel. In this way the tray can be viewed as an MFC collective with optimized electrolyte to electrode exposure. Multiple identical tray modules can then be stacked in all 3 dimensions and connected in series and/or parallel to both combine and stabilize the voltage and current of the entire assembly.

At SPIKA the MFC system comprised of 2 columns of 6 such stacked modules. At the top of the stacks, these columns shared a feeding tray where the artificial urine enters the system and then cascades through each module to a shared outlet at the bottom. The system was inoculated with electroactive bacteria using a mixture of wastewater from a local wastewater treatment plant, and CO 2 impacted soil.

Bioreceptive panels

The bioreceptive panels form the porous, outward-facing cladding for the middle section of the SPIKA tower. Their primary function is to operationalize bioreceptivity —the intrinsic capacity of a material or surface to be colonized by living organisms (Guillitte, 1995; Sanmartín et al., 2021). In an architectural design context, modulating this property through deliberate geometry becomes a form of spatial programming for microbial colonization. This enables a novel mode of ambient gardening and cultivation, designed to select for beneficial local microbiomes and reveal the latent biological character of a space through visible patterns of succession.

Acting as an experimental, open-system extension to the closed-loop hydroponic core, the panels were designed as textured substrates that provide micro- to meso-scale landscapes (Stohl et al., 2025) to attract, retain, and support the succession of ambient microbial life. The intended progression was to initiate the system with a diverse microbial inoculum, thereby promoting the subsequent establishment and colonization of photosynthetic organisms such as cyanobacteria and mosses, ultimately demonstrating the potential for low−energy, vertically oriented ecological development on built surfaces. Positioned at human height facing into the gallery, they were intended for close public scrutiny as living interfaces that directly interacted with the ambient environment—including light, humidity, and airborne particulates from other installations like Daphne’s Skin and DeepForest3 as well as spores and microorganisms present in the surrounding air.

Each panel, measuring approximately 1m by 0.5m, was fabricated using a layered approach. The primary substrate consisted of a porous, mineral-based hydraulic lime plaster, selected for its high capillary porosity and slightly alkaline pH to provide an ideal physicochemical medium for initial colonization. The surface of this plaster was manually scored and pitted with a specific pattern. This geometric intervention serves to dramatically increase surface area, create protective micro-shadows, and enhance localized water retention, embodying the complex topography of rocky natural substrates that support cryptogamic covers. As hygroscopic buffers, the panels were designed to absorb atmospheric moisture and release it slowly, contributing to a localized microclimate.

Prior to installation, to catalyze the intended ecological succession, the panels were inoculated with a biotic primer consisting of local pond water from Parco Sempione (May 2025)— a diverse assemblage of native microorganisms, including bacteria, cyanobacteria, microalgae, and fungi—combined with moss propagules from established local green walls. This “seeding” strategy initiated the panels as active, living interfaces. Symbolically, their living, evolving surface was designed to create a deliberate contrast with the sharp, defensive thorns above, visualizing the core concept of SPIKA: a defended, high-tech interior alongside a generous, outward-facing, fertile interface that engages with the world on its own terms.

Hydroponic system

The hydroponic component consisted of a vertical tower connected to a shared nutrient reservoir. The tower was 140 cm high (Figure 1a) and had 8 layers, each with 8 holes for hydroponic growth baskets. Thirty-two hydroponic growth baskets were inserted in alternate holes and the other 32 were sealed with opaque 3-D printed covers to prevent light entering the tower. At each time point, the 32 active hydroponic baskets were divided among the species, with approximately 3–4 baskets per species. These were distributed systematically across the top, middle, and bottom strata of the tower to account for potential variations in light and nutrient flow. Depending on the species of plant being grown, each hydroponic basket was filled with 1 cm diameter clay beads supporting a seedling or a clump of seedlings that filled the basket cavity. A submersible pump circulated the nutrient solution from the 30 L solution tank (Figure 1a) up to the top of the tower, from where it cascaded down the inside of the tower (Figure 1c) back into the solution tank providing continuous wetting of plant root in the baskets. The solution tank was topped up from a 20 L reservoir tank. Nutrient supplementation was performed at 3-week intervals using a commercial tripartite nutrient solution (T.A. Tripart) comprising distinct N–P–K formulations for vegetative growth, flowering, and micronutrient supply. Each dosing cycle included ~10 mL of the “Grow” solution (N–P–K: 3–1–6), ~5 mL of the “Bloom+” solution (N–P–K: 0–5–4), and ~1 mL of the “Micros” solution. The micronutrient formulation supplied essential trace elements at typical concentrations of approximately: iron (Fe) 0.10%, manganese (Mn) 0.05%, zinc (Zn) 0.015%, copper (Cu) 0.005%, boron (B) 0.01%, and molybdenum (Mo) 0.001%. Illumination was provided by integrated LED strips (LED Type 2835, brand CHICIRIS) with a full-spectrum light on a 12/12-hour photoperiod. The chemical properties of the solution were measured in the reservoir tank. The closed-loop configuration minimized water loss and recirculated nutrient solution for the nine ornamental and aromatic species cultivated during the exhibition. This design enabled consistent plant support while aligning with the Mi-Hy system’s circularity principles.

Operational role in the prototype

The core components of SPIKA—the MFC stack, hydroponic tower, and bioreceptive panels— were integrated into a single, interdependent operating system to materialize the Mi-Hy project’s circular principles. In the present prototype, these components were operated as distinct but parallel subsystems, enabling controlled baseline testing under exhibition conditions. This parallel configuration was a deliberate methodological choice: before introducing the complexity of hydraulic coupling between the MFC and hydroponic loops, we first needed to establish a validated baseline demonstrating that each subsystem could operate stably and reliably over an extended period. The MFC subsystem was therefore operated to characterize its bioelectrochemical performance and nutrient transformation processes, while the hydroponic subsystem was operated independently to establish stable cultivation conditions. This approach ensures that when we proceed to the next prototype—in which MFC-derived nutrient streams will be integrated into the hydroponic loop—we can attribute any changes in system performance to the hydraulic coupling itself, rather than to underlying instabilities in either subsystem. The MFC and hydroponic subsystems were operated in parallel rather than hydraulically coupled in this first iteration. This was a deliberate methodological control to establish a validated baseline for individual subsystem stability under exhibition conditions before introducing the complexity of nutrient-stream integration. The managed internal circuit forms a closed hydrological and nutrient cycle. A submersible pump circulates a nutrient solution from a shared reservoir up through the vertical hydroponic column, providing sustenance to the cultivated plants. In parallel, the microbial fuel cell (MFC) subsystem operated as a bioelectrochemical unit, processing artificial urine to generate electrical output and to characterize microbial metabolic performance. Simultaneously, the spontaneous external interface operates through the bioreceptive panels cladding the middle zone. Unlike the engineered internal loop, this subsystem is an open, porous boundary to the gallery environment. The panels’ material properties and geometry passively modulate the local microclimate (humidity, light, particle deposition) to catalyze and guide the colonization by both the microorganisms present in the initial inoculum of the panels and ambient airborne microbiota. This process of ecological succession, from attachment to visible cryptogamic growth, creates a living, evolving skin which works as a bioindicator and a low-energy, aesthetic complement to the precision cultivation happening internally.

Sensor networks (detailed in 2.3.2) monitor both subsystems, tracking physicochemical parameters of the hydroponic solution, electrical output from the MFCs, and capturing visual data on plant and panel growth. This integrated design demonstrates a model for building cores where tight, circular resource management coexists with generous, porous interfaces that enable spontaneous ecological engagement with their surroundings.

Exhibition and construction setting

Deploying the SPIKA prototype within a major public exhibition was a core methodological strategy, deliberately extending its development beyond laboratory validation into a real-world, socio-cultural context. This setting provided a robust arena to probe the system’s operational stability and, more critically, to interrogate its social and cultural acceptability as a model for future building cores. The gallery became a living testbed where technical performance intersected with public narrative.

The prototype was installed in a central gallery space at the Triennale Milano with overhead ambient light and indoors climate. The seven-month operational timeline from May to December 2025 was chosen to allow for the observation of long-term system performance and visible ecological succession on the bio panels. The system was commissioned one week prior to the public opening to establish initial stability.

Construction and Integration

The physical realization of SPIKA in the gallery followed a modular, five-step sequence that integrated its architectural, biological, and technological subsystems (Figure 2):

Assembly of the primary structure: The main steel frame and crown spikes were erected, forming the central support.

Integration of core systems: The hydroponic tower, MFC stacks, and LED lighting were mounted within the central structure, aligning all recirculation, energy generation, and illumination components.

Addition of vertical extensions: Thin extensions were added to heighten the visual profile while maintaining structural porosity for light and air.

Cladding with bioreceptive panels: The pre-fabricated and pre-inoculated bioreceptive panels, along with ancillary spikes, were installed to form the living, secondary skin.

Final elevation and stabilization: The structure was raised onto its spiked supporting legs, elevating the technical components and completing the installation including sensors.

Public interaction considerations

The design intentionally mediated public interaction. The imposing, thorn-like structure prevented physical touch, directing engagement toward visual observation and contemplation. Interpretive signage explained the system’s circular logic and the protective mythology of SPIKA, including the “Tower of Thorns” narrative and the conceptual role of ‘the cleaner’ as an immune agent. Through their cleaning rituals, this figure performed a curatorial function, selectively permitting environmental particulates like microbial spores, or plant seeds, to colonize the bioreceptive panels (Armstrong, 2025). Dynamic, non-verbal points of engagement were provided by the faint glow of MFC-powered LED lights and the evolving biological life on the bio panels. This setup encouraged a mode of interaction based on respect and observation from a distance, aligning with the exhibition’s theme of protecting fragile ecologies within shared public spaces.

Seven-month operational timeline

The system operated continuously for the duration of the exhibition. This extended period was crucial for several evaluations: assessing the stability of the coupled MFC-hydroponic system; monitoring plant growth and health across multiple species; documenting the gradual colonization and succession on the bioreceptive panels; and evaluating the robustness of all technical components, including pumps, sensors, and wiring, in a semi-public, real-world setting outside a laboratory.

Plant species selection and maintenance

Plant species were selected based on three primary criteria: ornamental value, ecological contribution, and functional traits relevant to urban environments. The goal was to integrate species that provide visual appeal while also contributing to biodiversity, fragrance, or pollinator support. Functional traits such as drought tolerance, compact growth habit, and adaptability to hydroponic conditions were also considered to ensure stable performance within the closed-loop Mi-Hy system.

Plant maintenance occurred once per week, during which the nutrient solution in the hydroponic reservoir was monitored and adjusted to maintain appropriate pH and electrical conductivity levels. Water levels were replenished regularly to account for plant uptake and evaporation. Lighting remained constant throughout the exhibition, providing adequate photoperiod and intensity for all species. To ensure continuous visual quality and system performance, plants were replaced every two months primarily to ensure aesthetic continuity for the public exhibition; biological constraints (such as root volume of biofilm) did not require removal, always maintaining a minimum of three species in the tower at any given time. Routine care also included the removal of senescent leaves and cleaning of structural components to maintain optimal growing conditions.

Sensors and data collection

Picture-based plant growth

High-resolution images from the hydroponic tower were captured every hour to observe leaf expansion, canopy architecture, coloration, and overall growth trends. Each captured image had a file size of 1.2 MB, providing a rich phenotypic record from which pixel-level health data could be extracted, including colour parameters derived directly from pixel values. Pictures were taken using a picture-based evaluation system (Raspberry Pi ® Zero 2 and Camera Module 3), with the camera installed to provide an overall view of the tower. In addition, the plants were visually monitored once per week, and supplementary pictures were taken to capture specific developmental events or changes not covered by the automated imaging cycle. These picture indicators served as the primary source of plant performance assessment, enabling a semi-quantitative evaluation of growth rates, flowering success, and visual health indicators. The qualitative growth rates (Fast/Moderate/Slow) presented in Table 1 were derived from a 1–5 visual scoring scale (where 5 = optimal canopy expansion/no stress; 1 = significant necrosis) recorded during weekly manual inspections, cross-referenced with hourly image data. Manual inspections occurred during routine maintenance but were not systematically recorded. The combined imaging approaches enabled consistent monitoring of plant responses to nutrient dynamics and environmental variations over time.

Hydroponic and MFC sensors

The hydroponic loop and water quality conditions were continuously measured with digital probes for pH, electrical conductivity (EC), dissolved oxygen (DO), temperature, oxidation–reduction potential (ORP) and CO 2 (Atlas Scientific ®). To assess the performance of the MFC, an electrical monitoring sensor recorded voltage output over time. Both the hydroponic sensors and the MFC energy sensor were transmitted data every 15 minutes, allowing detailed evaluation of system stability, nutrient dynamics, and electrochemical performance. Sensor readings were logged through the internet to an IoT cloud service (ThingsBoard) using the MQTT protocol for real-time visualization and long-term storage1. (Arduino ®). Resource-use metrics—including water refilling events, nutrient adjustments, and routine maintenance actions—were tracked to contextualize sensor data and understand the operational demands of the system during the seven-month exhibition.

System performance during public exhibition

The full Mi-Hy system—integrating the hydroponic tower, microbial fuel cell (MFC) stack, bio panels, and monitoring sensors—operated continuously and without major interruption during its installation in the “ We the Bacteria ” exhibition at the Triennale Milano from May to December 2025 onward. All components functioned as intended, with stable water recirculation, consistent electrical output from the MFC unit, and successful interaction between the biological and structural elements of the SPIKA architecture (Figure 3). The system remained accessible for weekly maintenance while staying fully exposed to public viewing. The installed sensors supported routine monitoring by providing reliable information for solution adjustments and general upkeep. Over the seven-month period, the system required a total of 35 water refill events, totaling 52.5 litres (based on one weekly refill of 1.5 L), and 12 nutrient dosing events were performed, totaling 190 mL of nutrient solution. This corresponded to approximately 120 mL of “Grow” solution, 60 mL of “Bloom” solution, and 12 mL of “Micros” solution. Based on these volumes, the cumulative macronutrient input was ~3.96 g N, 1.26 g P, and 9.24 g K, while micronutrient contributions over the experimental period were: Fe 0.19 g, Mn 0.095 g, Zn 0.0285 g, Cu 0.0095 g, B 0.019 g, and Mo 0.0019 g. The MFC stack consisted of 12 modules, each exhibiting an open-circuit voltage ranging from 550 to 620 mV. Modules were connected in pairs in parallel, and the six resulting pairs were connected in series, yielding a total stack voltage of 3.9 V (6.8 V for the entire unpaired stack).

Plant cultivation in the hydroponic tower

The hydroponic tower supported the successful long-term cultivation of nine plant species representing aromatic herbs, ornamentals, and medicinal plants: Ocimum basilicum, Mentha spp., Thymus vulgaris, Lavandula angustifolia, Salvia officinalis, Petroselinum crispum, Impatiens hawkeri (New Guinea impatiens), Origanum majorana, and Kalanchoe. The diverse mixture of leaf textures, growth forms, and functional traits highlighted the system’s capacity to sustain ornamental biodiversity within a circular microbial–hydroponic framework (Figure 4).

To provide a more structured, semi-quantitative assessment of plant performance, we analyzed our high-resolution image database and weekly maintenance logs to compile a performance summary for each species. Table 1 presents, for each of the nine species: (i) qualitative growth rate, (ii) duration of sustained visual quality, (iii) flowering status, and (iv) observed stress responses. This analysis transforms the visual data into a structured performance record, demonstrating that all species maintained good health throughout their cultivation periods under the low-nutrient, recirculating regime. For Impatiens hawkeri, an average of 3 branches plant were maintained throughout the observation cycle.

Plant monitoring system

The automated imaging system provided a longitudinal phenotypic record, with each captured image having a file size of 1.2 MB (Figures 4c, d). Pixel-level analysis of these datasets allowed for the objective quantification of canopy expansion and the maintenance of chlorophyll-related hue values. These indicators served as the primary source for the semi-quantitative assessment of nutritional and physiological status across all nine species. Progressive canopy expansion, stable leaf coloration, and consistent morphological patterns were observed throughout, with no significant deviations requiring intervention beyond routine maintenance. The image sequences also enabled precise timing of plant replacement to maintain installation visual quality, with the data in Table 1 derived from this systematic image analysis combined with weekly maintenance observations.

Hydroponic monitoring system

The hydroponic sensors operated continuously throughout the exhibition and recorded stable physicochemical conditions within the nutrient solution. Temperature values remained within 22–26 °C, providing a suitable range for both plant cultivation and microbial activity in the MFC subsystem. The pH fluctuated between 6.0 and 8.0, remaining within operational limits for the nine cultivated species and consistent with the expected variability of a circular nutrient solution. Mean pH was 7.1 ± 0.5 and mean EC was 510 ± 35 µS cm -1 throughout the deployment, demonstrating high operational stability. Dissolved oxygen concentrations ranged from 6–10 mg L -1, indicating adequate aeration and confirming that root-zone oxygen availability remained sufficient over time. CO 2 concentrations in the hydroponic compartment remained between 630–720 ppm, values typical of indoor environments and not indicative of excessive accumulation or depletion. Electrical conductivity (EC) measured 450–600 µS cm -1, reflecting a relatively low nutrient concentration that nevertheless supported plant development under the Mi-Hy system’s low-input conditions. Together, these measurements demonstrated that the hydroponic environment remained stable and physiologically acceptable for plant cultivation throughout the exhibition.

Integrating microbes into the built environment is part of a broader movement encompassing biodesign and planetary health, which reimagines buildings as dynamic, living ecosystems rather than sterile containers (Armstrong, 2024; Colomina and Wigley, 2025a). The SPIKA prototype advances this discourse by proposing that microbial logic can bestow metabolic agency on architecture. This represents a fundamental shift: by introducing microbial communities into a building’s core services—communities that feed on occupant waste —the closed, consumptive loop of conventional infrastructure can be broken. The building is transformed from a passive, fossil-fuel-dependent enclosure into an active, metabolic partner. This partnership internalizes the transformational loops of exchange that microbes manage in natural environments, bringing them into the heart of occupied space. Consequently, the building’s inhabitants are no longer mere consumers but are enrolled in a circular resource economy; their organic waste becomes the feedstock for microbial processes that, in turn, generate nutrients for plants and recover energy.

Framed as a symbiotic phage-like infection, SPIKA materializes this logic, demonstrating how circular, replicating instructions can reprogram a building core. This active reprogramming empowers occupants by converting their waste streams into a basis for cultivation, offering not only system autonomy but also the potential for direct access to plant produce. Thus, the project moves beyond mere technological integration to propose a new social and metabolic contract for urban living, where architecture performs the vital work of life, promoting resilience and direct engagement with the cycles of sustenance.

This metabolic agency is evidenced in SPIKA’s operational results. The stable cultivation of nine diverse ornamental species—including basil, mint, lavender, and impatiens—within a low-nutrient hydroponic regime demonstrates that plant performance can be reliably maintained under reduced input conditions in an architectural setting. The semi-quantitative performance data presented in Table 1 provides a structured record of species-specific growth rates, flowering success, and health indicators, transforming visual observations into a validated baseline for future work. While the MFC subsystem was not hydraulically connected to the hydroponic loop in this prototype, its parallel operation enabled detailed characterization of microbial nutrient transformation processes intended for future integration. Together, these results validate the feasibility of the Mi-Hy concept and establish a robust baseline for the next prototype, in which MFC-derived nutrients will be introduced as a functional input to plant cultivation. This suggests that the ‘prosthetic rhizosphere’ within the MFCs will not merely treat waste but effectively reconstitute it into a horticultural resource—a key prerequisite for moving circular cultivation from concept to reliable practice. Traditional hydroponics rely on a linear resource model (external fertilizer input vs. waste output). In contrast, SPIKA tests a circular ‘metabolic’ architecture. The advantage of this approach lies in its potential for architectural autonomy: by using the MFC as a ‘prosthetic rhizosphere’ to transform waste into nutrients, we decouple urban greening from synthetic fertilizer supply chains. This prototype validates the mechanical and biological stability required to host such a circular metabolism.

To further strengthen the scientific framing of the system’s potential advantages, we explicitly compare the SPIKA architecture to conventional hydroponic systems. Traditional hydroponic operations typically follow a linear resource model: synthetic fertilizers are manufactured (typically from fossil fuels), introduced, taken up by plants, and the nutrient-rich effluent is discharged as waste (Yuan et al., 2025). This linear flow, while efficient for production, is inherently resource-intensive and creates external dependencies—on fertilizer supply chains, energy inputs, and waste treatment infrastructure—that undermine the sustainability goals of urban agriculture. In contrast, the Mi-Hy concept inverts this logic by proposing a circular metabolism in which the MFC functions as a nutrient recovery and transformation unit—a “prosthetic rhizosphere”—that converts organic waste streams (such as urine) into plant-available nutrients, positioning waste as the input and plant nutrients as the output. This circular approach offers several potential advantages: (i) reduced reliance on synthetic fertilizers derived from fossil fuels, (ii) recovery of nutrients from waste streams that would otherwise require treatment, (iii) enhanced system autonomy through on-site nutrient generation, and (iv) potential for integration with building-scale waste management systems. These advantages are not yet realized in the current parallel prototype but are framed as three future research directions that our ongoing research—building on the validated baseline established here—is designed to investigate: (1) MFC-derived nutrients can support plant growth equivalent to synthetic fertilizers; (2) the system can achieve net resource recovery, producing more usable nutrients from waste than are consumed in external inputs; and (3) the coupled system can operate autonomously over extended periods without external nutrient supplementation. The baseline established in this study—stable hydroponic operation, characterized MFC performance, and documented plant responses—provides the foundation against which these hypotheses will be tested in the next prototype generation, where the MFC and hydroponic loops will be hydraulically coupled.

Central to this proposition is SPIKA’s dual metabolic strategy, which synthesizes a managed, closed-loop hydroponic-MFC circuit with a spontaneous, open bioreceptive interface. This design resolves a key challenge in ecological architecture: it couples the precision and efficiency required for resource circularity with the porosity and adaptability needed for contextual ecological engagement. The bioreceptive panels are thus more than a technical innovation in surface receptivity (Guillitte, 1995; Stohl et al., 2025); they are dialogic surfaces that perform the negotiation between a curated technological system and the uncontrolled public environment. Their colonization, subtly mediated by the narrative of ‘the cleaner’ as an immune agent, completes the conceptual arc from infection to immunity. If the phage-like SPIKA infects the building with a circular logic, the ‘cleaner’ represents the acquired immune response—a curatorial practice that decides which subsequent environmental ‘exposures’ (spores, particulates) are allowed to persist, shaping the long-term ecology of the architectural host (Armstrong, 2025).

Within the framework of the circular bioeconomy, the MFC’s role extends beyond nutrient recovery to encompass the broader principle of metabolic integration. As highlighted by Kong et al. (2024) and Walter and Popuri (2025), microbial electrochemical technologies can serve as core units for transforming waste streams into recoverable resources within integrated systems. By positioning the MFC as a “prosthetic rhizosphere”—a controlled microbial compartment that functions analogously to the natural rhizosphere in soil ecosystems—we ground the architectural and engineering work in a clear theoretical framework. This framing explicitly links the Mi-Hy system to the growing literature on circular bioeconomy and positions the prototype as a foundational step toward buildings that function as metabolic participants in urban resource cycles, rather than as passive consumers.

By materializing these principles at architectural scale within the demanding context of a public exhibition, SPIKA provides a foundational case study. It tests both the technical feasibility of a microbial-hydroponic core and also its cultural viability, offering a tangible model for how future buildings might embody the prion-like logic of the Green Deal—not as a superficial policy layer, but as a fundamental, transmissible code for a regenerative urban metabolism and horticultural production (Young et al., 2023; Qu et al., 2024). For the fields of horticulture and landscape architecture, the SPIKA prototype offers both a practical toolkit and a conceptual transition in the role and understanding of the plant within the designed urban ecosystem. Practically, it demonstrates the viability of cultivating diverse ornamental and aromatic species within a closed-loop, low-input metabolic system, with the semi-quantitative performance data providing a basis for identifying traits of resilience and adaptability crucial for future plant selection in urban applications. Conceptually, it proposes a new typology of ‘metabolic landscape infrastructure’ —exemplified by its dual-loop strategy—that can be adapted to indoor cultivation, green roofs, living walls, and botanic gardens. These structures can be designed to not only conserve resources but also to perform as public interfaces, making the vital relationships between plants, microbes, and the urban environment visible, legible and engaging. Ultimately, advancing such circular, cybernetic systems requires a collaborative effort: plant scientists selecting for novel functional traits, architectural designers and landscape architects designing for metabolic integration, and engineers creating robust resource feedback loops. By embracing this integrative approach, the vision of truly sustainable, productive, and resilient urban landscapes becomes materially achievable and implementable.