img(height="1" width="1" style="display:none" src="https://www.facebook.com/tr?id=2939831959404383&ev=PageView&noscript=1")

Slime and the city: can a mysterious mould aid urban planning?

Words:
Pamela Buxton

London architect ecoLogicStudio is harnessing slime mould's ability to establish the most efficient networks – as shown at a new Pompidou Centre exhibition

Biopainting from GAN-Physarum: la dérive numérique by ecoLogicStudio. ©NAARO. The project applies learnings from the behaviour of slime mould to create an AI tool for city planning
Biopainting from GAN-Physarum: la dérive numérique by ecoLogicStudio. ©NAARO. The project applies learnings from the behaviour of slime mould to create an AI tool for city planning

Many practices have studio dogs. ecoLogicStudio, a London-based architecture and design innovation practice, keeps slime mould. Unsurprisingly, this is not for its pet appeal. Instead, this mysterious organism is the driver for the practice’s research into a greener approach to urban design, informed by a deeper understanding of biological systems and models.

Slime mould (physarum polycephalum) has long fascinated computer scientists, who have been intrigued by its abilities to use collective resource distribution to learn from its surroundings and swiftly establish networks with the most efficient routes in its pursuit of nutrients.

For a decade, ecoLogicStudio has been harnessing the organism’s methods in its work on how to plan cities so as to prioritise blue-green networks (waterways, planting and parks) to encourage greater resilience against climate change. Its latest project GAN-Physarum: la dérive numérique, is exhibited as part of the Réseaux-Mondes (Worlds of Networks) exhibition at the Pompidou in Paris.

This offers an alternative to city planning, approached from the point of view of the built environment, with its ensuing emphasis on zone, boundary, scale, and typology, according to ecoLogicStudio co-founder Marco Poletto.

‘Slime mould is a living infrastructure,' he says. 'Through this model, we can plan a city, starting from the landscape and the resources a city needs to survive and grow.’ This, he adds, shifts attention towards the ’living, soft and wet’ networks.

For the Pompidou installation, nutrients were introduced in a laboratory test that represented areas of current green densities in a study area of 10 x 10km around the building. Over a period of three weeks, the slime mould first grows to explore the entire area before regrouping to establish the optimum routes, with thicker networks between the food sources. The research then zooms its frame of study to focus on a 100m area around the Centre Pompidou.

These patterns were then translated into planning strategy using a GAN (generative adversarial network) AI algorithm. Trained by ecoLogicStudio’s bio-computational team, it reads the patterns of a slime mould’s behaviour and learns to act in a similar way. This is then applied to the streets of Paris to reinterpret the intricate urban fabric through the eyes of the slime mould. In doing so, it maps the city’s biotic resources to create a distributed network of blue-green path systems.

The installation shows a largescale ‘bio painting’ featuring a slime mould as it stretches to feed on the nutrients (shown in a dehydrated state). A computational video shows the time-lapse training of the GAN algorithm as it strives to ‘read’ Paris.

  • AI video and biopainting from GAN-Physarum:la dérive numérique by ecoLogicStudio ©NAARO
    AI video and biopainting from GAN-Physarum:la dérive numérique by ecoLogicStudio ©NAARO
  • AI video and biopainting and DeepGreen, Urbansphere video from GAN-Physarum:la dérive numérique by ecoLogicStudio. ©NAARO
    AI video and biopainting and DeepGreen, Urbansphere video from GAN-Physarum:la dérive numérique by ecoLogicStudio. ©NAARO
  • Video frames of GAN-Physarum’s evolution in Paris on the scale of the city map, 3x3km from GAN-Physarum: la dérive numérique, by ©ecoLogicStudio
    Video frames of GAN-Physarum’s evolution in Paris on the scale of the city map, 3x3km from GAN-Physarum: la dérive numérique, by ©ecoLogicStudio
  • Rain water collection in Guatemala City from DeepGreen by ©ecoLogicStudio
    Rain water collection in Guatemala City from DeepGreen by ©ecoLogicStudio
  • Biotic network analysis of path systems in Guatemala City from DeepGreen by ©ecoLogicStudio.
    Biotic network analysis of path systems in Guatemala City from DeepGreen by ©ecoLogicStudio.
  • Biotic network analysis for Guatemala City from DeepGreen by ecoLogicStudio.  Drawing of the proximity network of existing biotic systems, highlighting areas lacking connectivity and requiring re-greening, overlapped with waste collection network.
    Biotic network analysis for Guatemala City from DeepGreen by ecoLogicStudio. Drawing of the proximity network of existing biotic systems, highlighting areas lacking connectivity and requiring re-greening, overlapped with waste collection network. Credit: ©ecoLogicStudio
123456

The installation also includes another video, DeepGreen: Urbansphere, showing the application of an AI algorithm for blue-green city masterplanning projects as part of a project with the United Nations Development Programme. This combines AI and big data analysis to test the potential of a green planning interface by simulating scenarios of sustainable urban development.

EcoLogicStudio is confident that the biocomputational systems derived from the slime mould will have applicability, and change approaches to city planning based on zones and boundaries. Its aim is to develop new planning tools that, like the slime mould, will adapt to changing conditions and embody a bottom-up rather than top-down approach to planning. It hopes this can inform a more sustainable and responsive approach to city planning that takes account of the flow of water, energy and waste through cities.

Practice co-founder Claudia Pasquero, who is also professor of biodigital architecture at UCL, hopes it may lead to a shift in the concept of masterplanning, away from a single vision towards a more collaborative process, and one that extends to non-human agency through such algorithms.

‘It’s a crucial shift for the profession and our role as designers and planners,’ she says.

According to Poletto, as time runs out for the introduction of a more radical approach to blue-green planning, this slime-mould-inspired method is ready to be adopted.

‘The more it [is] deployed, the more advanced its training will be,' he says. 'It is a co-evolutionary process that we need to start at scale as soon as possible with the involvement of a vast network of bio-designers globally!’

Both GAN-Physarum: la dérive numérique and DeepGreen: Urbansphere have been acquired by the Centre Pompidou as part of its permanent collection.

GAN-Physarum: la dérive numérique is part of the exhibition Réseaux-Mondes (Worlds of Networks), until April 25 2022, Galerie 4, Centre Pompidou, Paris, France.