The Concept

Brief and Concept

Concept

Common Ground Food Lab61 Montague Road, South BrisbaneSharon Robinson

Public laneway with seating beside a water channel

01 · Client

Common Ground Food Lab

A not-for-profit organisation strengthening urban food resilience through controlled-environment agriculture, applied research and public education.

  • CharacterPractical, progressive and curious. Welcoming, purposeful spaces, honest materials and visible working processes.
  • Operational NeedReliable year-round growing, with public and research activities alongside a working farm.
  • Practical ConstraintSystems must support the farm’s operation, not function solely as demonstration.
Client Profile

Common Ground Food Lab is a not-for-profit organisation focused on strengthening urban food resilience through controlled-environment agriculture, applied research and public education.

The organisation seeks to establish a productive Smart Farm precinct in South-East Queensland where commercial-scale growing can operate alongside research, training and public engagement. The facility would support growers, researchers and industry partners while allowing students, community groups and visitors to experience contemporary food-production processes.

Common Ground is practical, progressive and curious, with a strong culture of innovation, research and shared learning. It values technology where it improves productive outcomes or helps people understand the relationship between food production, water and environmental systems. The architecture should reflect this mindset through welcoming, purposeful spaces, honest materials and visible working processes, balancing the robustness of a productive facility with warmth and a strong sense of public invitation.

Operational NeedCommon Ground requires reliable year-round growing capacity supported by controlled growing environments, water and nutrient management, research spaces, produce handling and associated operational infrastructure. Public and research activities must coexist with the practical requirements of a working farm.

Practical ConstraintThe precinct must manage water and energy demands within realistic operational and financial limits. Systems adopted by the project must therefore support the farm’s operation rather than function solely as educational or demonstrative features.

The site remains provisional. Common Ground prefers previously disturbed, underutilised or industrial land where a new productive use could help transform an existing condition while limiting unnecessary ecological disturbance.

02 · Site

61 Montague Road

The former Pauls / Lactalis dairy processing site on the Brisbane River.

  • Adaptive ReuseLarge roof areas, extensive hardstand and robust built fabric, rather than a cleared site.
  • WaterSubstantial roof and paved surfaces for rainwater capture, movement and storage.
  • Urban LocationMakes food production and water systems accessible beyond the working farm.

03 · Precinct

A mixed-use precinct

Housing, the Smart Farm and community / public uses. This concept focuses on the Smart Farm.

Aerial view of the precinct on the Brisbane River: the kept industrial buildings with solar roofs and silos, two timber housing blocks, the timber farm block and the white ETFE canopy with cafes, set in a riverside park, with the surrounding city faded to grey
Concept masterplan, shown on Google Earth imagery: proposed buildings and landscape added with AI. Base image: Google Earth; Image © 2026 Vexcel Imaging US, Inc.

04 · Concept

Between Rain

Rainwater is usually treated as infrastructure to be collected, stored and moved out of sight. This project investigates what changes when its capture, movement and use become legible within architecture.

At 61 Montague Road, the existing industrial fabric provides large roof surfaces, robust structures and spaces capable of adaptation. Rather than applying a rainwater system to a resolved building, rainfall begins to influence roofs, surfaces, thresholds and occupied spaces.

Water moves from capture to transfer, storage and productive use, supporting the controlled-environment Smart Farm while remaining perceptible to the public.

But rainfall is intermittent. The architecture therefore cannot depend on water as spectacle. The same spaces must remain useful, inhabitable and meaningful between rain events.

Capture Transfer Storage Distribution Growing Reuse

Developed further in The Proposal →

The first study: The Architecture of Water
The Architecture of Water research board: the funnel image, the six-part sequence from rain to question, the essay and three close-up images, with references
The Architecture of Water, research board. Images generated with ChatGPT from written prompts.
The board, essay and prompts

05 · Rainwater System

Capture, transfer, store, grow, reuse

Sectional perspective showing the rainwater system from capture to reuse
Sectional perspective: capture, transfer, storage, distribution, growing, reuse

06 · Sequence

Four moments

Canopy directing rainfall into a column
01

Capture

Rainfall begins to shape roofs and surfaces, making capture part of the architectural form.

Glazed water column descending into a channel
02

Transfer

Captured rainwater moves through the retained industrial fabric, making transfer part of the spatial experience.

Public corridor alongside a glazed controlled-environment growing room
03

Productive Growing

Captured rainwater supports a controlled-environment Smart Farm for food production, research and education.

Public laneway with seating beside a water channel
04

Public Life: Occupation

The water system becomes part of an everyday public passage, while shade, seating and activity allow the space to remain useful between rain events.

Image Development

Each vignette began as a Midjourney generation. Capture and Productive Growing were edited in ChatGPT; Transfer and Public Life were first used as generated (checked pixel for pixel against the Midjourney files, 21 Oct 2026). Transfer and Productive Growing were then refined in Nano Banana (24 Oct 2026). The prompts and results for each edit are on the Process page of the original concept presentation, linked below.

Midjourney canopy study
Midjourney origin: Test 03, canopy study
Final Capture vignette
Final

01

Capture

6 ChatGPT prompts

Edited in ChatGPT with targeted, single-change prompts: the waterfall replaced by a column, a dry-weather correction, and the soffit material. One prompt was a critique: the AI had changed the timber direction, which was marked up against the Midjourney original and corrected.

Midjourney origin of the Transfer vignette
Midjourney origin: a second image from the Test 01 prompt
Final Transfer vignette
Final

02

Transfer

ChatGPT edit rejected; 8 Nano Banana passes

One ChatGPT edit was made and rejected, and the Midjourney original was kept. It was later refined in Nano Banana with single-change passes: the wall corner, roof trusses, base plate, water in the column, a shallow channel, full-height glass, the frame blended in and a seated person. Two corrections were drawn directly (the truss bottom chord and the glass frame), and the image was cropped at the top and sides.

Midjourney storage study
Midjourney origin: Test 01, storage study
Final Productive Growing vignette
Final

03

Productive Growing

3 ChatGPT prompts; 4 Nano Banana passes

The water basin behind the glass wall became ordered growing beds; workers were added, kept secondary; the growing area became a sealed, climate-controlled enclosure. Then refined in Nano Banana: the glazing head and edge aligned, the bench legs and pipe regularised, and the lettuce made natural.

Midjourney origin of the Public Life vignette
Midjourney origin
No edits

04

Public Life

0 edits

Used exactly as generated.

Concept Process

Presentation and process

The original concept presentation, with its Process page: the AI workflow, the iterations and rejected images, and how the concept developed.

Open the concept presentation