February Newsletter

'Patterns from Nature' Update

The ‘Patterns from Nature’ project was kicked off in late 2006 through an article in the BioInspired! Newsletter and a post to the Biomimetics listserver. Eighteen people representing a broad range of interests and disciplines volunteered, of which eight have been part of the core team from the beginning. The group identified a compelling need for a tool to organize knowledge about biological systems in a structured fashion and facilitate inter-disciplinary communication, particularly with researchers and practitioners not fluent in biology. ‘Pattern language’ emerged as one way to accomplish this.

Strengths of Pattern Language

The pattern language approach was developed in the late 1970s by Christopher Alexander to capture the characteristics that made good architecture ‘whole’ and ‘alive’ Pattern languages combine a unique set of features that make them particularly useful for complex and evolving knowledge domains:

• A simple yet powerful structure that describes the problem and the steps involved in developing solutions (individual patterns)
• A process for creating and using these patterns
• Rich content about the domain of knowledge represented by the patterns
• The ability to use ‘common language’, helping a broader audience gain value from the content
• The ability to handle breadth as well as depth, to whatever level is appropriate
• The ability to solve larger problems through a series of smaller solutions developed using the explicit linkages between patterns (the pattern language)

Ecosystem Patterns

Although pattern languages are usually built ‘from the bottom up’, the group decided to start with ecosystem principles, with the goal of identifying ‘deep patterns’ that could be applied to human challenges. To date, the team has explored three ecosystem principles:

• Ecosystems tend to use materials effectively for multiple functions
• Ecosystems tend to create conditions favorable to sustained life
• Ecosystems tend to be made up of interdependent cooperative and competitive relationships

The ‘Multi-Functional Materials’ proto-pattern contrasts the tendency of living systems to use a small set of materials for multiple functions, whereas we employ chemistry to develop new materials for each desired performance characteristic or function. Research by Julian Vincent (University of Bath) suggests biological manipulation of the embedded structure of materials dramatically reduces energy requirements, compared to our “heat, beat and treat” approach that uses energy to remove and then impose structure (Biomimetics listserver post). An example of leveraging the inherent qualities of materials is the JANO Dual Bike, designed by Roland Kaufman. He discovered that wood is significantly stiffer than fiberglass or even Kevlar/epoxy composites. By using wood veneer, Kaufmann was able mold wood into shapes that combined lateral stiffness with the shock absorption of carbon fiber and the responsiveness of steel.

The ‘Conducive to Life’ proto-pattern looks for mechanisms underlying the balanced, rich, diverse and vibrant communities of species that we see in healthy ecosystems. On the surface, natural selection working at the level of individual organisms would appear to encourage competition and declining diversity, attributes we often ascribe to human systems. Complexity theory and research on Self-Organizing, Hierarchical, Open (SOHO) systems suggests that open systems through which high quality energy flows tend to display emergent properties and spontaneously self-organize. ‘Free energy’ appears able to create ‘order out of disorder’, if other enabling conditions are satisfied. Regen Energy uses self-organization and principles of swarm theory to develop autonomous, self-organizing agents that collectively manage peak power consumption (see the December 2007 BioInspired! Newsletter).

The group found substantial overlap between the ‘Interdependence’ and ‘Conducive to Life’ proto-patterns. In order to differentiate them, the team is exploring more complex networks such as food webs. The complexity of these systems appears at odds with their resilience - the group is exploring a number of avenues to reconcile the apparent paradox. Early results reinforce the importance of maintaining the overall health of a complex ecosystem, rather than targeting a specific issue. One example might be the work of Ducks Unlimited to restore wetlands and reverse declining waterfowl populations.

Implications

Energy is woven through all of the three patterns. New research in thermodynamics suggests that energy is different from other resources. With the exception of atmospheric gases lost to space, materials on Earth can be recycled indefinitely, although they may at times be trapped in forms that are not easily accessible over the short term. On the other hand, energy cannot be recycled completely. Work can be turned into heat, but heat cannot be converted into an equivalent amount of work. Ultimately, we rely on energy from the sun and, to a lesser extent, geothermal energy from radioactive decay deep within the Earth. Any other forms of energy are prone to depletion unless usage is matched to the rate of replenishment.

One way in which humans have been able to skirt energy issues is through heavy reliance on 'heat pathways', an ability rarely seen in other living systems. John Reap (Georgia Tech) suggests there may be natural limits on energy usage by organisms across a wide scale. Total human energy consumption significantly exceeds this limit, especially when compared to the low metabolic rate of humans and large primates. By tapping into fossil fuel, we have been able to temporarily ‘step out of the game’, driving our explosive development at the cost of dependence on shrinking supplies.

Project Status

The team started looking at 'pattern language' as a tool or a method to solve a specific problem: how can we organize, communicate, and make biological information available to other disciplines. As we began to develop patterns around the ecosystem principles, the pattern language methodology encouraged us to gain greater insight into what each ecosystem principle means and how this knowledge could be applied to understanding human systems.

Pattern languages can capture the evolving knowledge about complex systems as well as the gaps in that knowledge. Such a framework could help practitioners leverage this incremental knowledge and also encourage research to close the gaps. Pattern languages could also support design in complex, dynamic environments by encouraging practitioners to explore problems in depth, rather than jumping immediately into finding a solution. The multi-scale aspect of pattern languages helps practitioners look at how their solutions 'fit' within the larger context, while the communication and collaboration aspects help them deal with multiple stakeholders.

The work of the team has been presented at the 2007 Society of Experimental Mechanics, with the team’s first paper published in the conference proceedings. An update was presented at Biomimetics 12 in Bath (UK) late in 2007. The latest information will be presented at the Institute of Biological Engineering conference in early March. Discussions are underway to submit a paper to the peer-reviewed Journal of Biological Engineering.

For more information, please visit the Patterns Wiki. The Wiki is accessible by anyone, although you need to login to download most documents and create or modify content. Please contact me through the link below if you have questions or are interested in being part of the project. We are particularly looking for expertise in non-equilibrium thermodynamics, exergy (a measure of energy quality), ecosystem dynamics and self-organizing systems.

Norbert Hoeller