News CV Texts Input Home

My Ph.D. Dissertation: Ambient Intelligence Ecologies - Toward Biomimetic IT Link to text

My research has focused on issues in biomimetic IT design, i.e. IT designed with support from models of living and complex systems. Biological approaches to IT design are blooming these years and I am convinced that we are one the verge on completely new ways of understanding, creating and using IT. My Ph.D. project is an investigation into this new field and comprises analyses of the historical and scientific roots of the present occupation with biological models, perspectives for the biological approach in relation to design, use and conception of IT, investigations of biomimetic IT systems and much more.

However before I fixed my project on biomimetic IT design I had to walk a curved path. After five years of strictly philosophical education I began my Ph.D. project by studying biological theories of cognition of relevance for the design of better IT interfaces. However convinced of the profound truth of interactivist theories on complexity, self-organization and adaptivity, as developed by my co-supervisors Prof. Mark Bickhard and Ph.D. Wayne Christensen, I started focusing on new ways of designing IT to meet growing demands for flexibility, robustness and adequate services.

Biomimetic Ambient Intelligence

I am investigating to which extend we can apply a biologically inspired approach on the design of Ambient Intelligence (AmbI). I prefer the notion of AmbI since it covers the interesting elements by future IT: 1) In opposition to Pervasive or Ubiquitous Computing it denotes the functionality – intelligent assistance – and not some means to obtain it, while still capturing the highly distributed nature of IT. 2) ‘Ambient’ indicates the ‘present at hand but non-intruding’ behavior we seek from an optimally functioning technology captured under the slogan ‘If there is to be computers everywhere they’d better get out of the way’. 3) ‘Intelligence’ not only concerns the behavior of the technology but intrinsic characteristics of recursively adaptive systems, i.e. their structural and functional development and organization.

Efforts in the biomimetic direction are currently being pursued and we will see a lot of new types of technology in the decades to come. But it cannot happen successfully without investigating a range of central topics concerning technology and design. We will have to deal with some very interesting questions, some of which I will sketch here.

Change of perspective

To follow the possibilities that a biomimetic approach offers will take a change of perspective both within IT research and - of equal importance - a new view on assisting technology from the users. In relation to the research because we will have to let go of the teleological idea of complete control in the design process. Major aspects of the systems we aim for will have to develop 'on their own' by means of evolutionary computing techniques, simply because these methods seems to facilitate the most stable and tolerant systems - as in nature. Besides studies show that we will reach a limit in decreasing the size of hardware. Identical components will simply perform uniquely due to micro-physical effects leaving circuits with idiosyncratic characteristics. Thus we will have to rely on a structural and functional coupling emerging from the development of the systems themselves instead of being designed top-down.

Designing will have longer perspectives than we are used to. Creating technology will be a matter of both development and evolution, or ontogenesis and phylogenesis. Ontogenetic factors will be new to us. Today we are only used to the 'history' of artifacts as patina and wear, which is not an actively adaptive feature (it can be a passive adaptation though). Instead a new type of technology will emerge that is not ready for use when it leaves the production line, but that needs to learn, mature and perhaps even grow. Artifacts will develop and evolve - and thus get 'designed' - on the fly with their use. We are more familiar with the notion of phylogenesis of design, albeit in a metaphorical sense as improvement of artifacts over several 'generations'. However this process will be of even greater importance when we start 'breeding' on successful systems. We cannot rely on old fashioned design and production because we will probably not be able to analyze vast AmbI systems completely do to their complexity. In its extreme this evolutionary tendency could also lead to the end of mass production born with industrialism. An IT design strategy resting on historical and developmental factors, which are intrinsically contingent, would leave the idea of large scale production of identical products uninteresting.

Instead of purchasing finished static 'products' we will start buying into dynamic 'produces' continuously developing by responding to their use. Produces will be more like a service providing the right kind of developmental possibilities than a commodity. Produces will obtain a functional coupling with their users in a mutually forming adaptive process. Adaptation does not mean accommodation to fixed circumstances as some seem to think but actively seeking optimal fit by interactive negotiations. In such a process AmbI produces might 'train' the user in some kind of functional deal. Thus produces harbors the possibility of highly idiosyncratic practices and very personal relations between people and their technology.

Users and consumers on the other hand should prepare for a different technology where features, speed and 'squeaky-clean' aesthetics is stressed less than reliability, adaptivity, self-maintenance and calmness. This part of a 'bio-mimetic turn' in the design of technology will be the most controversial due to the social and symbolic nature of man. The reason being that our technology is a cultural phenomenon that serves other and sometimes much more important purposes than strict functional assistance. We communicate by way of our artifacts, and seem to like them flashy, to have a lot of features and to facilitate identity. Some even suggests that technology culturally serves as the antidote of the creepy, slimy, dangerously subversive forces of nature.

A need for a new kind of consumer ethics will probably rise (for products that is not use-and-through-away). If the informational or software level and the materiality or hardware level of systems starts to merge we probably just cannot continue the informational promiscuity we impose on our computers now by installing and uninstalling software, downloading all sorts of files and programs etc. If properly functioning technology starts resting on integrated development, we will start to be more critical about the long term treatment of our assisting devices and not just rely on anti-virus programs and cleaning from time to time. Maintenance of especially vital devices will start being more therapeutic or historically sensitive.

Why biomimetics?

One important question is the nature and rationality of the biomimetic quest. Why is it so interesting to look to nature in the design of technology at all? Without denying that the bio-focused tendency we experience these years is sometimes hyped beyond bearing (due to various more or less historical factors such as neo-romanticism, the environmental movement and the ever implicitly legitimized reference to the norms of nature) I think there is some good and sound reasons for studying natural systems as well. The basic argument is our limited cognitive capacities as Home Sapiens. We were never promised a more privileged access to reality than an ecological balancing survival and energy costs reasonably. Thus we are perhaps superior in building tools and artifacts compared to other animals, but when it comes to really complex constructs we just cannot overview every aspect. Hence we try to 'import' some of the mechanisms we think we have unveiled in natural complex systems and processes in order for them to take care of business when we cannot ourselves. Plus many complex systems are intrinsically dynamic so that we just cannot 'build' them because they are heavily historically constrained (this aspect could stem from our limited insight though). The strategy is to take advantage of the huge amount of 'empirical knowledge' that nature possesses after a long, slow and thorough evolutionary process. A story about how life by evolution came to exploit the goodwill of the universe, i.e. the tendency in living systems to organize and order various processes by recursively using the spontaneous (law-like) pattern-generating abilities inherent in the universe. A capability we would very much like to adopt and benefit from faced with the great technological challenges that lies ahead.

Another reason to study biological processes is the need for a deepened understanding of the informational or semiotic aspect of the natural orchestration of system organization, in order to approach questions like: What are the informational processes involved in the flow of natural systems? Are information processes a singular phenomenon, like a natural law, or is it a range of loosely related dynamic phenomena? How much semantics is involved in informational processes, i.e. are these processes more related to causality than meaning? Can information processes be adequately treated without including the structure and materiality of systems? How does information relate to patterns in general? I think a lot of questions like these will have to be addressed in order to gain properly from a biomimetic approach. A quest I have heroically thrown myself into