Machine Wilderness at Transmediale

Posted Feb. 25, 2017 by Theun Karelse

Machine Wilderness investigates the nature of technology in our landscapes. We think that an investigation of those hybrid ecologies is important in addressing our present situation of global environmental crisis. Complex machines have been part of our environment for many centuries. Pioneers like al Jazari already made programmable automata around 1200AD. Machines came to dominate our landscapes dramatically since the Industrial Revolution. The word that comes to mind is brutality.

During these many centuries we have shared our landscapes with a myriad of all kinds of creatures; spiders, kangaroos, salmon, turkeys, ants, dolphins, hermit crabs.. all kinds of beings, each with their own perspective on things. But all those centuries we've designed our technologies in that shared environment to cater for the needs and preoccupations of only one species; a particular kind of ape. When you really think about that, it reveals an unbelievably self-absorbed design practice. And it has come at a price.

Edward O. Wilson describes our current age of mass extinction as the 'Age of Loneliness' and in many ways our technologies in these shared environments have been technologies of loneliness.

Machine Wilderness takes a radical turn towards the diversity in our shared environments. We identify the recent appearance of autonomous systems in biomes such as coral reefs, rain forests and rivers as sign of change in human technologies. They appear at a time of deep environmental crisis often dealing directly with the collapse of natural processes in ecosystems. In a way the crisis is bringing them out, it is forcing us to find design practices that are inclusive of non-humans.

This first generation of autonomous systems - the drones, ROV's, UAV's - are almost exclusively aimed at solving single issues in damaged ecosystems. They operate in a narrow bandwidth of landscape management. Machine Wilderness states the need to take our design processes much further, beyond crisis management or biomimicri, towards participation. We are interested in exploring the question of what design methods would look like for engaging with full environmental complexity.

Machine Wilderness is not about promoting the use robots in our landscapes or even environmental engineering in general. We identify the autonomous robot as a prime candidate for learning how our technologies relate to natural processes, and to learn this relationship can move towards holistic participation or even co-evolution. Robotic systems can move, sense and act, so they manifest themselves in the landscape almost like animals do. We think this exposes more of the issues of relating to a given landscape than, say, sustainable architecture.

Our program is centered on field-work sessions where we develop and prototype systems that try to engage with the local environmental complexity. The landscapes serve as case studies. This enables us to engage with real circumstances and stay away from generalizations and theoretical abstractions.

In this workshop during Transmediale we ask if the behavior of a local animal can be a guide for designing technologies that are adapted to that environment. Through observation we want to build up a picture of a typical daily routine of an animal is really like.

[what are it's daily goals?]
[what are it's daily challenges?]
[what opportunities is it looking for?]
[how / where does it get information to act on?]
[what social communities does it belong to?]
[how does it avoid danger?]
[what is it's energy source?]
[how does it celebrate it's existence?]

By observing the phenology - the full spectrum of behaviour - of a local animal we aim form a picture of what it means to participate in the environment locally. This workshop investigates if the daily routine of an animal can serve as inspiration for designing autonomous systems to deal with these same issues. Can observation be the basis for designing technologies adapted to local conditions?

Ecologists are experienced observers of animal behaviour. Matthew Creasey is a behavioural ecologist who specializes in bird behavior. He introduced the idea of the ecological niche - the position or role an organisms holds within the ecological community - and he gave an overview of observation techniques used in ecology:

Focal-Animal Sampling - record all of the actions of one individual for a specified time period.

Continuous Group Sampling - record all of the behaviours that occur while the group is being watched, e.g. preening, feeding, flying, displaying, and the time & duration of the behavior.

Instantaneous Sampling - record the behaviour of an individual at predetermined time intervals.

Scan Sampling - record the behaviour of all group members at predetermined time intervals.

Sampling Occurrences of a Specific Behaviour - record each time a chosen behaviour is observed during a specified time period.

Ethogram a catalogue or inventory of behaviours or actions exhibited by an animal.

Then we went outside for fieldwork around the Haus Der Kulturen Der Welt, starting with some excursuses in direct perception. After this we collectively explored some of the surrounding park with Matthew to get a sense of the observation methodologies. This was followed by an hour of individual exploration.

In the direct vicinity of the Haus Der Kulturen Der Welt participants observed mushrooms, trees, mussels, many different species of birds, raptors, moths, moskitos, bamboo and woodlice. Various tracks revealed activities of animals like woodpecker holes, hares, mice.  Many processes and relations in the landscape were observed: relationships between trees, the dance of moskitos, Participants became aware of different scales of space, different scales of time - a butterfly may sit on my hand for a few minutes, but for that butterfly that may be a significant part of its lifetime. We considered the nocturnal world, seeing organisms as evolving processes rather than fixed. How do animals view us? Does weather affect their psychology? And what looks like playful behaviour in a pond in the summer, we may associate very differently in the cold of winter. Participants tried to understand behaviour by copying it, flicking leaves on the floor with a small twig. Tried to feel what it is like to sing like a bird continuously and ongoing - what the duration of it does to your psychology. What is it like when a bird makes eye contact, or when you meet a mouse? Could we learn form corals how to live in heigh density of neighbours? How do the bodies of animals influence the way they relate to others - we relate differently to birds that mushrooms which don't have heads with eyes. Could we bridge the different sensory worlds between species with technology? Even similar senses can be very different between species - birds seeing gravitational patterns.

To close the workshop we asked if the exercises with observation seemed relevant to the context of machine wilderness. Participants felt that the exercise exposed how hard it is to move beyond our given senses as humans and our own domains of experience, but that the challenge is enjoyable and relevant. They felt is not so interesting if it literally takes an animal as the starting point of a design method: it is not so interesting to replace an animal with a machine copy. It's potential lies more with looking at behaviour - the social dynamics of groups of organisms or interspecies interactions in a landscape and finding new modes of relating that fit within that. In conclusion we feel that real observation needs more time and that it would be interesting to do the same kind of workshop over a period of a week or more. Then you can really start to get a deeper understanding of the animals. This is something to try out in a future session of fieldwork.