complexity

Complexity and Bio-Inspiration

There is growing interest in exploring how bio-inspiration could tackle more complicated challenges.  A key goal of bio-inspiration is to re-balance our relationship with nature and the ecosystems that support us – simple solutions are unlikely to have a significant impact.  However, Ashok pointed out in the May 27th B3D Webinar that the tools and methods we have used to emulate organisms or simple natural phenomena are not well suited to dealing with systems.

The Cynefin framework (David Snowden and Mary Boone) is a useful way of looking at different degrees of complexity, from simple through complicated to complex and finally chaotic.  Each level has specific characteristics and requires unique approaches.  Incorrectly assessing the complexity of a situation or using the wrong approach can lead to poor, unexpected or catastrophic results.  As the complexity increases, understanding the situation and using an appropriate process become increasingly important, in contrast to simpler situations where specific knowledge is more effective.  Rather than looking for specific solutions in complex situations, the goal is to find ‘safe-fail’ interventions, constantly adjusted based on results. 

This message is reiterated in Donella Meadow’s Thinking in Systems.  Again, rather than proposing solutions, Meadows emphasizes finding the right leverage points.  The book provides extensive examples of modeling systems and seeing how the model reacts to changes.  Terry Love has taken this further to develop very complex models that made unexpected predictions.

Where does bio-inspiration fit it?  Although programs like Cosmos: A Spacetime Odyssey emphasize the interconnectedness of all life, we would benefit from a clearer appreciation of how our existence depends on natural systems.  Another approach is to look at how evolutionary processes could be applied to design, although time scales, history and context can make this challenging.  We could also apply the ideas of Snowden/Boone and Meadows to understanding the evolving dynamics and relationships between humans and the natural world. 

In addition, there may be generalizable attributes of complex systems that can be deduced from studying a wide range of human and natural systems.  Meadows in Thinking in Systems (pages 51-58) describes the influence of delays on systems with multiple feedback loops.  Although it seems intuitively obvious that reducing delays would improve the overall efficiency of the system, Meadows shows how this can actually result in instability.  The Complex Mathematics of Robot Wrestling describes how the researchers modeling wrestlers discovered that introducing a short delay into the response to inputs reduced the level of complexity in the simulation.  Perhaps the germ of an underlying principle?

Stuart Kauffman in At Home in the Universe  claims to have identified optimal degrees of ‘connectedness’ as the number of actors in the system increases.  Too few connections (as determined by the number of influencers on each actor) and the system is unable to adapt.  Too many connections and the system becomes unstable.  Julian Vincent speculated in A Comparison of Biological and Technological Systems that being embedded in a system reduces complexity because you only need to deal with a limited part of the system. 

Are you aware of other examples that suggest ‘lag’ and ‘connectedness’ might be useful attributes to explore?  Are there other examples of attributes identified in natural systems that might be applied to technological systems?

Syndicate content