Video: Cellular Forms, Andy Lomas 2014
Music: Max Cooper

ARTIST ANDY LOMAS TALKS TO US ABOUT COMPLEXITY, ITERATIVE GROWTH, AND REALLY ENJOYING HIS WORK

I meet Andy Lomas for a late afternoon demo at the Wellcome Collection in London, a museum and library exploring connections between science and art. It seems an appropriate meeting place for an artist who bases his creations on cellular growth systems, but, as I discover, his work is about more than meets the eye.

Bas Relief Cellular Form (detail), Andy Lomas 2017

Andy’s innovations first hit my radar at an interactive manifestation of his process during the V&A Digital Design Weekend 2019, which led me to his website – a minimalist, systematic index of his works to date. There’s something sparse and untelling about the site and it serves as a reminder of the heart of a mathematician’s approach: to find the simplest, most efficient solution. Indeed, Andy’s inspiration for his work on cellular growth systems began with a question sparked whilst observing coral on a Hawaiian dive trip: how simple can a rule be to generate complex forms?

And his prints and videos exemplify this apparent contradiction. The pulse of organic shapes populating his videos seems at first to follow a strict, linear pattern, but, as the forms throb and implode, they divide and multiply at rates, and in directions and intensities, greater than my mind can process. Other works remind me of coloured ink drops spreading in gently moving water, dancing and morphing into jellyfish forms which then envelop themselves into shelled cocoons. When these iterations are stopped in time, as they are in his beautifully intricate monochrome prints, our brains want to categorise them as things we recognise, but our perception keeps shifting, as the forms really belong to just that sliver of ambiguity which belies classification.

In his ‘Cellular Forms’ series, Andy used digital simulation of a simplified biological model of morphogenesis, with three-dimensional structures generated out of interconnected particles to represent cells.

Video: CellularForms HD (X-Ray version), Andy Lomas 2014
Music: Max Cooper

He describes how each cluster of cells in this series of videos is ‘incrementally developed by adding iterative layers of complexity to the structure,’ the aim being to ‘create forms emergently: exploring generic similarities between many different shapes in nature rather than emulating any particular organism, revealing universal archetypal forms that can come from growth-like processes rather than top-down externally engineered design.’

Andy goes on to explain that the ‘cell division is controlled by accumulated nutrient levels. When the level in a cell exceeds a given threshold, the cell divides, and various parameters control how both the parent and daughter cells re-connect to their immediate neighbours.’ Any ‘new nutrients can be created by photons in cells hit by incident light rays,’ and these ‘nutrients can also be allowed to flow to adjacent cells. The simulation process is [then] repeated over thousands of iterations and millions of particles, with each of the final structures comprising over fifty million cells [1].’

Video: Cellular Forms HD, Andy Lomas 2014
Music: Max Cooper

Unsurprisingly, these mathematically-derived, computer-generated simulations of growth processes are reminiscent of scientific cell growth representations and time-lapses of flowers in bloom. And yet, they hold intrigue in their layers of complexity. Andy tells me that he’s ‘generally trying to push to the high limits of detail and complexity,’ the rule of thumb he uses: ‘you need about a million of something before it starts becoming visually intricate and detailed.’

A background as a mathematician paved his way into an extensive set of CG Supervisor credits in Hollywood films. However, as we sit together in the Wellcome café for my introduction to a sophisticated system of algorithms and parameters, he impresses upon me that mathematical simplicity is the underlying key to his intricate images. And, as he shows me his Species Explorer interface, a software he developed and a crucial element to his work, I notice it looks similar to other editing software in its layout and functionality. According to Andy himself, ‘most of the software is just how to make [his] work fast, rather than actually make it do something really complex.’ This is one of his key aims, to explore ‘how ridiculously simple the core system can be,’ and he believes, ‘the simpler it is, the richer the results.’ It’s a simplicity of software and generative system which has taken five years to develop, in addition to eight years of knowledge and skills accumulated from his time working in films.

“Generative systems are typically based on algorithmic processes that are controlled by a number of parameters. Given a set of parameter values the process is run to create an output. Classic examples include Conway’s Game of Life (Conway 1970) and reaction diffusion equations (Turing 1952)”

— (Lomas, 2018) [2]

In the early stages of choosing parameters to be used with his generative systems, Andy was only able to loosely predict whether he would get the results he wanted using perhaps a maximum of three parameters on up to two cells at a time - with Species Explorer to assist in generating parameter values, his work begins with a minimum of forty two cells. The software incorporates both evolutionary methods and machine learning techniques to allow Andy more flexibility in how he uses the system, depending on what he feels like on a given day, the combination of these two ways of working allows for finding authentic and novel results.

Although it sounds as though there is a heavy reliance on Species Explorer to make choices that control the outcomes, and perhaps even raises the question of where the artist is in this seemingly computer-led process, Andy is certainly not removed from it. His subjective and instinctual response is built into the process by way of a system of manual categorisation that allows him to direct and teach the software what he finds interesting. Compared to the traditional painting process I’m more familiar with personally, I can definitely see the computational speed and processing advantages of an assistant like Species Explorer.

However, Andy’s foray into 3D printing, in collaboration with The Bartlett School of Architecture (UCL), demonstrates some of the practical limitations which currently exist when producing computer-generated art. Due to the nature of 3D printing, overhangs in a three-dimensional form - particularly for the sort of intricate details inherent in Andy’s work - can become problematic in production. Normally, to deal with this sort of production limitation, the solution would be to reduce the range of produceable options to only include the consistently produceable results. However, this method also greatly limits the system from finding interesting and novel results.

To solve this dilemma, Andy set up parameters to account for materiality. He increased growth in some areas and reduced growth in others to help push the forms towards being 3D printable. By twisting the data he is able to take into account materiality, meaning that most of the time he would get a lot of intricate variety in the results that would still mostly be printable. One example of this careful balance is the ‘Mycelium Vase Form,’ which was fabricated in a collaboration with Blast Studio using their custom built 3D printer. They used a pulp made from recycled coffee cups which is used as a substrate to grow mycelium, so the final result is a solid structure made of mycelium hyphae.

Mycelium Vase Form, Andy Lomas & Blast Studio 2020

According to Andy, ‘nature shows systems of complexity with simple ingredients that are mixed together for astonishing results - and this is very Stuart Kauffman - what nature has used to make astonishing things, through a certain level of computation, becomes a possible thing for us to work with as well.’ He tells me that he believes this presents ‘an existence proof that shows it works - that shows there is a way, but maybe don't assume that's the best way or the most efficient way or the most natural way for us to work with it.’ An evolutionary system aims to reproduce the most successful species, therefore inherently a lot of the rich, emergent behaviour gets lost.

In fact, Andy tells me he is ‘addicted to rich emergent behaviour,’ explaining that this type of behaviour tends to ‘...occur close to the boundary between regularity and chaos.’ It can’t be easily predicted and is typically non-linear [3]. Like many of the Algorists who explored ‘…the relationship between rule-based systems and randomness’ [4], Andy is focused on the growth during the ‘tipping points’ in behaviour as it moves away from predictability.

His ‘Hybrid Forms: Cell differentiation’ is particularly illustrative of this unpredictability. Andy likens the green cells to broccoli and the blue to a brain. As the cells grow and multiply, we can see the ‘broccoli’ and ‘brain’ cells fighting to dominate the overall form. Eventually, we recognise their resemblance to brassica and cerebral images, but the most curious moments are, perhaps, when the forms don’t securely resemble anything familiar, they merely hint at familiarity.

Video: Hybrid Forms: Cell Differentiation, Andy Lomas 2016
Music: Max Cooper

It’s clear during the course of our conversation that Andy doesn’t aim to reproduce growth systems accurately, as a pure scientist might. He uses Species Explorer to quickly and systematically find forms using growth parameters which are modelled from nature but nurtured specifically to be visually complex. Scientists who have seen his forms often comment on them being inaccurate representations of nature, which seems to please him, he draws inspiration from nature without wishing to recreate its work.

What strikes me as I listen to Andy talk about his collaboration with The Bartlett, is that this particular strand of his work hasn’t resulted in him creating manufacturable products. But this makes sense. The artist I meet is, in essence, deeply curious, focused on finding the next question rather than any definitive answers and, in his own words:

‘This is me, doing what I what I like doing. The art context is a good discipline to work in. It's fun, it’s interesting, it's nice getting out there in exhibitions… It’s not real science in some ways, but it does have a science-like approach. There's a freedom to explore wherever you want to explore.’

Explore more of Andy’s work at www.andylomas.com

A full paper on his ‘Cellular Forms’ series is available here

See Andy soon at:

http://www.evostar.org/2020/evomusart/
http://www.eva-london.org/eva-london-2020/

References:
[1] Lomas, A., CellularForms HD (X-Ray version), https://vimeo.com/99948218
[2] Lomas, A., 2018. On Hybrid Creativity. Arts, 7(3), p.25.
[3]Lomas, A., 2020. Morphogenetic Vase Forms. [online] MIT Press Journals. Available at https://www.mitpressjournals.org/doi/abs/10.1162/isal_a_00215 [Accessed 5 April 2020].
[4] Phillips, A. (2011). The Algorists. [online] American Scientist. Available at: https://www.americanscientist.org/article/the-algorists [Accessed 5 Mar. 2020].

All images and videos used with kind permission from Andy Lomas

GENERATIVE ART: A FREEDOM TO EXPLORE