Conducive to Life

From Patterns from Nature
Jump to: navigation, search

Go back to the List of Patterns

Problem or Challenge

Healthy ecosystems evolve into balanced, rich, diverse and vibrant communities of species, in spite of natural selection. Natural selection works at the level of organisms; there is no generally accepted theory of natural selection at the ecosystem level. "Survival of the fit" suggests competition rather than cooperation and interdependence, yet many ecosystems display interdependent, cooperative or symbiotic behavior between individuals and species.

Web-of-life-copyright1.gif The Grinning Planet®


Ecosystems do not lend themselves to analysis by reduction. Ecosystems are not machines with easily identified causes and effects. They are:

  • comprised of complex arrangements of living and non-living entities
  • arranged in hierarchies of sub-systems and super-systems
  • interacting through a web of energy, material and information flows
  • spanning a wide range of spatial scales, from cells to the entire biosphere
  • spanning temporal scales due to iterative processes and feedback loops


Therefore

New analysis methods and tools are required to understand the principles underlying ecosystems and allow designers to apply these principles in solving societal problems. Non-equilibrium dynamics and complexity theory [34, 35] suggest that under certain enabling conditions, open systems through which high quality energy flows will display emergent properties and spontaneously self-organize.


These systems built increasingly more organized processes that in turn build and maintain new structure. As a result these systems are able to capture, utilize and dissipate more energy; cycle resources more effectively; increase respiration and transpiration; add more biomass; and increase diversity. As energy flow increases, systems are pushed further and further from equilibrium until a critical point is reached where the system can no longer adapt and behavior becomes chaotic. More than 80 years ago, Lotka summarized his observations of biological and ecological systems in much the same way [Lotka 1922]. He believed that evolution generates and selects living systems that maximize total energy flow per unit time subject to constraints.


Self-Organizing Hierarchical Open (SOHO) systems may exhibit multiple stable states. Changes in the system's environmental conditions may appear to have no influence on the system, due to regulating feedback loops. However, at a critical point, even a small change may cause the system to switch to another stable state. The change may be very rapid, discontinuous, and from our perspective, unpredictable. The state at which a SOHO system finds itself is a matter of history - the concept of a 'correct' system state is a matter of human preference.


The principles of SOHO systems suggest a number of implications for the design process. The structure and processes of societal systems depend on the resource flows from natural systems. Therefore societal systems must maintain the integrity of natural systems, by protecting their current well-being, ability to adapt, and ability to evolve. This requires not only a much deeper understanding of the structure and behavior of societal systems, but also the related natural systems and the interaction between the two. Narratives or scenario analysis may be better able to capture and represent the richness of information required by this process.


Kay proposes four design principles: [35]

  • we must interact with natural systems recognizing their limited ability to provide energy and absorb waste
  • large scale societal systems should be designed similar to the behavior and structure of natural ecosystems
  • where possible, natural subsystems should be used to deliver the functionality required by societal systems
  • non-renewable resources should be treated as capital expenditures to allow us to switch to renewable resources


Kay proposes two design strategies: [35]

  • adopt adaptive management: the ability to sense and adapt to change must be built into our designs
  • applying the precautionary principle: minimize our interaction with the natural system particularly in our use of resources such as energy, our production of waste, and our destruction of the natural environment on which we rely.


Kay emphasizes the importance of analyzing both the quantity and quality of flows. In most cases, we focus on the former, using measures of efficiency. Quality is related to effectiveness. Kay cites an example where electric radiant heat is 100% efficient, but is less effective than using natural gas. The high quality of electricity is more effectively leveraged in heat pumps or devices where heat is a by-product. Efficiency can also lead to optimization at the component level, on the assumption that this will make the entire system efficient. Effectiveness encourages optimization at the system level, accommodating non-linear interactions and feedback loops.


These concepts lead to a different approach to design. Rather than creating static 'solutions to a problem', design "must be seen as setting in process the evolution of a built environment which evolves to meet the evolving needs of users and which adapt so as to fit into changing environmental conditions." [35]


Vincent [15] estimates that "... in technology, the manipulation of energy can account for up to 70% of the solutions to technical problems, whereas in biology energy never figures more than 5% of the time." Our ability to tap into fossil fuels has allowed us to deploy large amounts of energy when and where we need it. Kay and Gribbin describe how increased energy flows result in "period doubling", where the number of states and length of the state cycles increases dramatically until the behavior of the system becomes chaotic. However, they do not provide guidelines for evaluating the state of complex systems. Are societal systems in the 'chaotic' range, where increasingly complex control mechanisms (driving even higher energy demands) are required to enforce stability? Would reducing our energy consumption allow inherently stable states to emerge?


Such questions command particular attention in light of recent findings concerning the mass-specific metabolic rates of organisms ranging from bacteria to whales. Makarieva and coauthors found that organisms spanning 20 orders of magnitude in body mass displayed specific basal metabolic rates from 1-10 W/kg [Makarieva et al 2005]. This finding led them to speculate that a universal specific metabolic rate might exist. If a universal rate exists for organisms, does it mark the energetic constraint mentioned by Lotka and the chaotic boundary discussed by Kay and Gribbin?


But

The mechanism of how systems self organized is unclear. Furthermore, we have only rudimentary tools to document, analyze, and make predictions about the complex structure and processes of self organizing systems. Although Network Thermodynamics and graph theoretic techniques allow us to capture the quantity and quality of energy flows, we lack similar tools to measure the quality of material flows. Our ability to analyze information flows is even more limited.


Kay points out that "humans have a set of priorities which will cause them to find a different balance, between the need to make good use of resources while coping with a changing environment." In contrast to natural ecosystems, "We value human life and try to minimize the 'hardships' felt by members of our species." [35] We cannot blindly copy solutions from natural systems as though we do not exist or are able to isolate from those natural systems. Like beavers building dams, we influence and are influenced by the ecosystems in which we are embedded.


Approaching problems from a systems perspective often requires that the problem be re-stated or re-defined, introducing uncertainty into the design process. In addition, the emergent characteristics of self-organizing systems can result in potentially unacceptable unpredictability. "In general, we do not accept unpredictability in technical systems; indeed we avoid it." [15] Vincent points out that we face this challenge with existing designs, "... since nearly every technical system is actually a combination of a technical system in the narrow sense, and a living (usually human) system which is the operator of this technical system." [15] Isolating the technical from the broader system "... can lead to reduced effectiveness, ... technological catastrophes and/or social tension and unrest." [15] However, without tools and methods to reduce this uncertainty, designers will continue to follow 'tried and true' methods that they are familiar with.


See Also (Optional)

See Multi-Functional Materials for implications of the 'quality of materials'.


New.gif Applications

Self-Organization, Embodiment, and Biologically Inspired Robotics

"Self-Organization, Embodiment, and Biologically Inspired Robotics" (Rolf Pfeifer, Max Lungarella, Fumiya Iida) describes an approach to designing 'embodied' robots that, like biological organisms, are able to "perform and survive in a world characterized by rapid changes, high uncertainty, indefinite richness, and limited availability of information." The term 'embodied' refers to the distribution of control and processing throughout the robot, including incorporating aspects of the interaction between the robot and the environment. In contrast, robots based on centralized, internal control mechanisms tend to demonstrate reduced adaptability and energy efficiency.


The authors have abstracted four principles:

  1. Behavior is affected by the environment in which the system is embedded, the morphology of the system (shape, placement of sensors and effectors) and the material properties of the system's elements.
  2. The interaction of the system and the environment is regulated by physical forces, which can help improve stability, maneuverability and efficiency.
  3. Coupled sensory-motor activity can introduce statistical regularities in information flow, enhancing processing.
  4. Treating an embodied agent as a complex, dynamical system allows self-organization and emergence to be incorporated into the design


The article examples studies of insect motion that show how the structure and orientation of the legs results in coordinated positioning without any central control. Insects can also adapt to surface irregularities through "passive compliance of the insect's muscle-tendon system and the slack in its joints." 'Walkers' have been built that can walk down a slope without requiring any control systems or actuators, by carefully matching the elements and morphology of the walker to the physical characteristics of the environment. Intriguingly, adding actuators and simple control systems to these walkers allows them to learn how to walk on level surfaces much easier than traditional approaches. Other examples include the dynamic shape of fish fins and insect wings that maximize propulsion (or lift) with minimal energy consumption.


The article explores how 'active sensing' can simplify the challenge of dealing with large quantities of real-time sensory information. Coupling sensory and motor activity introduces 'structure' into the information and can also control when sensory information is received. Bionics: International Scientists are developing robotic rats describes research into rat's whiskers as active sensors that contain feedback loops which in turn control the whiskers movement.


The article goes on to explore designing robots that can alter their morphology to match the environment or function required, similar to how organisms co-evolve their brains and body. Modular robots have been designed with components at centimeter scales. Future advances in self-assembly may allow for microscopic modular robots. Another approach is 'collective robotics' that explores cooperation and coordination between independent agents, based on research into similar phenomena in nature ("the formation of trails and bridges; sorting, flocking and schooling behaviors; communication; and dominance interactions").


A key message is that "actual behavior emerges from the interaction dynamics of agent and environment through a continuous and dynamic interplay of physical and information processes." Control and processing is distributed, often using structure, shape and component qualities as alternatives to traditional information processing and control mechanisms (possible linkages to Multi-Functional_Materials. The article does not directly discuss energy flows, but rather explores the factors that in combination with 'free energy' can lead to complex SOHO systems. The principles listed above could be sub-patterns, expanding on and linking the ecosystem patterns.


Regen Energy Power Controllers

Regen Energy has developed autonomous power controllers that use the principles of 'swarm theory' to minimize peak power consumption driven by cyclical electrical devices. The devices are easy to install, self-organize, learn the characteristics of the system (no requirement for programming) and require no management. For an overview, see The Power of Ants and Bees in the December 2007 BioInspired! Newsletter


"Emergence in Times of Emergency" (Steve Barth, May 17/2007)

Steve describes numerous examples of where 'top-down' systems have failed in copying with emergencies, while 'bottoms-up' or emergent community responses have (in comparison) been more effective. What is unclear are the conditions that either encourage or inhibit community-based responses. I suspect the degree of 'official' structure can be a major factor - we tend to defer to experts, even if they are not effective. Robust, distributed communications is clearly another. One of the comments talks about pre-existing trust relationships. A willingness and ability to deal with uncertainty helps. Emergent Design talks about alternatives to traditional 'training events'. A number of comments talked about presentation effectiveness being inversely related to preparedness. Rather than just talking about "tools you can use to self-organize and react to the unknowns you’ll likely face", the training session could have been a simulation of a disaster where participants get to use the tools and become more confident in their abilities to deal with real disasters.


At the risk of turning into a 'hammer' that treats everything as a nail, I was intrigued by a comment to Barth's post about Charles Perrow's research: "disasters often involve the release of massive amounts of energy, which results in events unfolding in ways that simply cannot be foreseen." The Next Catastrophe: Reducing Our Vulnerabilities to Natural, Industrial, and Terrorist Disasters talks about concentrations of energy sources, population and economic/political power as increasing our vulnerability. Natural disasters also involve releases of energy that overwhelms our ability to cope. On the one hand, concentrations build up gradients that eventually must be dissipated. On the other hand, emergent responses could be seen as a means of dissipating the effects of the energy release.


"Hungry like the wolf" (Business Organization)

Julian Wilson of Matt Black Systems contributed an article to the February 2008 BioInspired! Newsletter that contrasts the traditional 'command and control' methods of business with the 'independent, responsible agent' model that he believes is more prevalent in natural systems. Wilson has successfully implemented a system that puts more responsibility and authority on individuals within the company: each person is largely responsible for the full lifecycle of a project, from scouting the opportunity to delivering the final product. Although inefficient according to traditional management models that encourage 'economies of scale', the result has been higher profits in a difficult market, greater quality and better 'on time' performance.


Miscellaneous Ideas

Retrieved from "http://sinet.ca/patterns/index.php?title=Conducive_to_Life&oldid=2128"