Analysis Of Biological Mimicry And Genetic Algorithms
Abstract
Deciding upon the best way to fathom out a solution to a complex problem or come up with ideas for optimal ways of working, novel types of materials or new technologies is one of the largest challenges faced by almost every organisation fervent in the construction industry today. There have been a number of ‘high-end’ tools that have embraced the teachings of nature. These tools have tried to apply natural theories to all manner of applications. This review will focus its analysis on comparing two techniques – Biological Mimicry and Genetic Algorithms. The study will provide the reader with a descriptive overview of the two approaches; an overview of some of the applications and an analysis of the similarities and differences between them.
Introduction
There is an overwhelming depth to the amount of information available on Biological Mimicry (BM) and Genetic Algorithms (GA), traversing a great number of sectors and industries. All of them draw upon an increasing number of natural sources but according to Julian Vincent at the Centre for Biomimetics at Bath University “at present there is only a 10% overlap between biology and technology in terms of mechanisms stale”. This means there is gigantic potential for further applications to be developed. The main focus of this essay will be to look at how these are applied to the Construction and Property Sector.
This essay is divided in to four progressive sections. Firstly it will look at outlining the fundamental basis (in simplistic terms) for making decisions. Next it moves on to describing BM and GA’s followed by looking at several applications of each. Finally it will compare the two techniques.
As a pre introduction to the two techniques the following passage aims to bring some degree of genuine world and historical context to the subject prior to setting off on the journey:
“A funny thing happened to business in the last two decades – it turned upside down. Suddenly, the old rules no longer applied – stability became liability, size an danger and command & control hierarchy an albatross. Business as usual became business as unusual: unpredictable, un-plannable and above all unmanageable”, Clippinger J H III Et al (1999)
This extract will, hopefully, fall into place as the essay unfolds. It highlights the point that GA’s and BM have advance about due to a shift in the way business is conducted.
Decision Making
As a precursor to defining the two techniques, it is for the relieve of the reader that we first outline a fundamental of decision-making principles.
A decision is essentially a “choice made from a set of alternatives”, Eaton (2004 \ 05) it logically follows that an assessment of these alternatives must take place to find the most helpful option(s) for the specific requirements of the problem. The following techniques relieve finding alternatives and help propel problems more efficiently towards a solution.
Biological Mimicry
Biological Mimicry (BM) is defined by the online free encyclopaedia, Wikipedia (2006), as “the application of methods and systems found in nature to the gape and design of engineering systems and modern technology”.
This is further emphasised by Eaton (2001) who states that “Biomimicry is the application of recognised biological concepts outside the discipline of biological science”. Biological Mimicry can be known by a variety of names, including but not limited to biomimetics, biognosis, bionical creativity engineering and bionics.
The word – Bionics – is an apt fusion of words to concisely sum up the principles of this topic. Major Jack E Steele first coined it in 1958; he was part of the Aerospace Division of the United States Air Force and wanted to promote it as a original path of science. The word is a portmanteau; the initial allotment of the word bio is taken from biology, which in turn is derived from the Greek word for Life. The second half of the word is taken from the latter part of the word electronics.
The act of transferring this information from nature into man-made scenarios is built on the premise that:
“There is a duality between engineering and nature which is based on a minimum use of energy. This is because animals and plants, in order to survive in competition with each other, have evolved ways of living and reproducing using the least amount of resource.” Vincent (1998)
This point is further emphasised by Wikipedia (2006)
“The transfer of technology between life forms and synthetic constructs is shapely because evolutionary pressure typically forces natural systems to become highly optimised and efficient”, Wikipedia (2006)
However, research undertaken by Vincent (1998) states that this “shouldn’t be taken as a panacea for problems but more as a portfolio of paradigms to draw upon”. The main method, in which BM will be thought of and dealt with in this essay, is in the context of the research undertaken at the Centre for Biomimetics at the University of Bath, by Richardson who states that:
“Biomimetic models can be used to create analogies in business. The analogies can be applied to provide inspiration as a precursor to innovation. This approach can then be hybridised with traditional programme governance models to create an integration to implementation method”
In layman’s terms he is essentially stating, that nature has a wide variety of solutions that can be drawn upon. They can be applied in business terms and have frequently inspired people to come up with new previously unconsidered solutions that they can merge (Hybridise) these new ideas with already established conventional methods to give a greater scope to potential solutions.
The reason for defining the context of BM in this essay is because there is a whole world of Biomimicry outside of Construction and Property that could have a whole number of books written about it from bionic cars to revolutionary technology that eliminates the need for stout SCUBA equipment. For that reason that side of BM has been excluded from this paper.
Genetic Algorithms
Genetic Algorithms are described in different ways depending on the industry or field in which they are being used but for the purposes of this essay we will use a more generic definition that is commonly applied to all sectors:
“Based on the idea of genetic inheritance and Darwinian theory of natural evolution (survival of the fittest), GA’s work as stochastic search techniques to score an optimal solution to a given problem from a large number of solutions”, Shin and Lee, (2002)
This description is more concisely and simply described by Metaxiotis & Psarras (2004) as a “powerful and flexible means for obtaining solutions to a variety of problems that often cannot be dealt with by other, more traditional and orthodox methods”. There is a particular process that GA’s follow or rather a set of rules that they use to carry out the process of obtaining solutions.
In order to understand the process it is best to first define some of the terminology. To begin with a GA needs to have a ‘population’ to work with “a population is the position consisting of a definite number of chromosomes randomly selected”, Davis (1991).
To break this down further the population is comprised of ‘chromosomes’ these “are strings consisting of bits, which represent a point in the solution space. The set of bit symbols is called the alphabet. The most fundamental and popular representation of a chromosome is the string of binary numbers, 0 and 1″, Davis (1991).
To bring these two previous sections together somewhat, once the population has been created a “GA encodes the problem into a set of strings (chromosomes) each of which is composed of several bits (genes), it then operates on the strings to simulate the process of evolution”, Metaxiotis & Psarras (2004). More specifically it is based on Charles Darwin’s theory of Natural Selection:
“Natural selection acts only by taking advantage of microscopic, successive variations. She can never a great and sudden leap take but must advance by short and certain, through slow steps.” Darwin (1859)
Each of the individuals in the population is given a Fitness rating, “fitness is the performance evaluation of individuals in the population. The higher the fitness the better the performance of the individual and the greater the probability of survival.” Davis (1991)
To simulate the process of evolution GA’s use a set of basic operator’s \ variables. The most frequently employed operators in GA are Davis (1991):
(1) Reproduction. Reproduction is the process in which individuals copy themselves, according to the probabilities that are proportional to their fitness values. As result, individuals with higher fitness values will have higher probabilities of producing their offspring in the next generation.
(2) Crossover. Crossover is the operator that produces two current chromosomes (offspring) by exchanging some bits of a couple of randomly selected chromosomes (parents).
(3) Mutation. Mutation operates on a single chromosome with a very small probability. With this operation, one or more bits are chosen at random from the chromosome and are changed into a different symbol of the ‘alphabet’.
(4) Inversion. “This operator aims to mimic the property from nature that, in general, the function of a gene is independent of its location on the chromosome. The original proposal reordered genes through inversion of a fragment of the genome. With the more sophisticated crossover operators former nowadays, inversion is usually unnecessary.” GAUL (2006)
The process is followed through using evolutionary inspiration and “the best chromosome from the final population is taken as the solution to the problem”, Metaxiotis & Psarras (2004)
Applications of BM and GA
Rather than headline many applications of GA’s and BM this essay will look at a occupy number of applications in detail, pertaining to Construction and Property Industry.
Fuzzy Optimisation of Labour Allocation by Genetic Algorithms (GA)
Genetic Algorithms can be used to create a “GA optimisation model to allocate workers of different skill types to complete a defined amount of trade works in a defined time frame”, Tam et al (2003). However the particular model proposed by Tam et al (2003) assumes the following points:
§ Available numbers of workers in each skill type
§ Daily quantities of each job type
§ Workers daily wages of each type
§ Hourly productivity
In addition to this Tam et al (2003) sets the following parameters (1) Population size, (2) Crossover rate and (3) Mutation rate. The initial population uses “continuous chromosome representation, as they generate the largest number of combinations for representing the percentage allocation of resources”, Tam et al (2003). The flow chart that follows highlights the overall process:
Flow Chart of the GA Model
The model outlined above has various advantages and strengths over other optimisation techniques. Tam et al (2003) states that it “can identify a near-optimal solution in a short period of time without need to search for the optimal solution over a longer period of time”. The process is also “structured in a multi-level format. Decisions made at one level will influence the factors leading to decisions made at another level”, Tam et al (2006).
In conclusion their findings provided useful and “practical examples that can result in the reduction of the total labours costs” used on construction projects.
Predicting Corporate Bankruptcy (GA)
There are many prediction models weak for calculating or gauging the potential for companies to become bankrupt, Aziz & Humayon (2006) split these into three main categories:
i) Statistical Models
ii) Artificial Intelligent Expert System Models (AIES)
iii) Theoretical Models
GA Models fall into the AIES category. Aziz & Humayon (2006) compiled a table summarised from “89 published empirical investigations that were collected from more than 180 sources” they have been grouped into the aforementioned framework of the three categories.
In summary, of the 89 investigations listed out in their research GA’s are cited 4 times, despite the low number of times it appeared in the selected investigations it still scored higher on average than most other categories.
In the overall averages as a category it came 5th out of sixteen but the four other predicting methods above it had lesser amounts of samples.
This Table shows its overall average raking in Overall Predictive Accuracy (OPA)
Rank
Type
OPA (%)
Nr of Samples
Nr 1
Gamb
94.00
1
Nr 2
Credit
91.15
2
Nr 3
RS
90.60
3
Nr 4
Probit
88.85
2
Nr 5
GA
88.65
4
Nr 6
BSDM
87.75
Nr 7
RPA
87.12
Nr 8
Logit
87.05
Nr 9
CUSUM
83.70
Nr 10
CBR
83.02
Nr 11
Univariate
81.41
Nr 12
LPM
81.03
Nr 13
Par Adj
80.49
Nr 14
MDA
79.17
Nr 15
NN
76.99
Nr 16
Cash
67.32
The following table shows it compared in terms of ranked against GA’s highest sample, its top two highest samples and its tops three highest samples:
OPA (%)
GA Top 1
97.30
GA Avg of Top 2
96.15
GA Avg of Top 3
91.63
As this second table clearly shows, when directly compared on a level basis that GA’s out performed other predictions techniques and shows that “AIES models have slightly better average predictive accuracy than statistical models”, Aziz & Humayon (2006)
Value Ecologies (BM)
The term Value Ecology (VE) comes from the evolution of Supply Chains. Traditionally Supply Chains “take the necessary steps a product takes from origin to consumption” Hearn & Pace (2006, Pg 56) and looks at it in terms of cost minimisation.
The next evolution from Supply Chains are Value Chains, which go further than supply chains in that they “emphasise cost optimisation and value maximisation” Walters & Lancaster (2000).
However, Hearn & Pace (2006) suggest that VC’s “rest on a simplified notion of ‘value’. For example they assume value remains ‘in the product’ ignoring externalities (i.e. product value derived from the relationship of the product to a system or other products”. This appears to suggest that the process needs to evolve further and is not fully evolved as a system:
“Like many metaphors, the view of a value chain is at once useful (e.g. because it clarifies key processes of product delivery and emphasises value creation), and limiting, because it hides the dynamism of value creation.” Hearn & Pace (2006)
This is where the evolution to Value Ecologies (Networks) comes in to it own. It essentially breaks the chain metaphor based on the linear theory. The Value Ecology looks at things in terms of organic networks, this was more accurately terms by Moore (1998) as an “extended system of mutually supportive organisations”, and this is also where he coined the term “business ecosystems”.
This evolution has led to a shift in the terminology used; Hearn & Pace (2006) provide a list of five (5) key areas where this terminology shift is more evident:
1. The shift from thinking about consumers to thinking about co-creators of value
2. The shift from thinking about value chains to thinking about value networks
3. The shift from thinking about product value to thinking about network value
4. The shift from thinking about simple co-operation or competition to thinking about complex co-opetition
5. The shift from thinking about individual firm strategy to thinking about strategy in relation to the value ecology as a whole.
The main benefits and points to be drawn from Value Ecology in terms of business analogy from biological mimicry, according to Hearn and Pace (2006) are:
§ Value ecology encapsulates emerging understandings of the knowledge economy operates and how business strategy can be derived from this understanding.
§ Helps us understand how value is understood, identified, and/or created.
§ Directs us to understand the vitality of a system not fair in terms of its inherent characteristics but also in terms of its relationship to other ecologies, both material and ideational.
§ The population dynamics of a value creating ecology teach us about strategies that succeed.
In conclusion Hearn & Pace (2006) sum up the spend of VE thinking as a contemporary business practices as follows:
“If firms can learn to apply ecological models metaphorically to analyse their business processes, they might also come to see their business outcomes in terms of ecology in actuality – and this might be an even more necessary shift in thinking.”
Building and Protecting Corporate Reputation (BM)
This particular application of biomimicry is much broader in context than Value Ecologies; it is at the top end of strategic planning but filters all the way down the line. Reputation is an element of survival or as Firestein (2006) puts it “a risk to its reputation is a threat to the survival of the enterprise.”
To develop a good corporate reputation in today’s modern world it is necessary to be more in tune with natural rhythms and structures. This could be looked at in the map a rain forest balances its ecosystem in that it has to manage its self efficiently and effectively in order to survive.
In the same scheme a company has to have a strong internal culture that “forges a positive concept of the company by successfully coping with both expected changes and unanticipated challenges”, Firestein (2006).
To highlight the importance of a corporate reputation it is best to peek at some examples of seemingly strong corporations who suffered from and what Firestein (2006) describes as “subtle ethical drifts in corporate culture, made in increments so small as to go almost unnoticed, brought companies into damaging confrontations with investors, regulators, and the public.” This looks and sounds very much like Darwinian Theory of Natural Selection.
Firestein (2006) gives several examples of reputations being destroyed and some that bounced back, he describes them as “reputational compromises that can lurk in any corporations DNA”, again another example of a biological metaphor being used in a business situation.
In his paper Firestein (2006) highlights the way in which certain companies have evolved to have a learnt response to planning towards potentially damaging future situations. For example:
“BP is notable for having reputation risk management in its global enterprise well before ENRON and others put it on the front page. The company has the experience to know that fierce opposition comes with the territory. So, it attempts to structure the planning of such operations to acknowledge critical questions before they are asked. Its business plan is designed to create a relate of responsibility well before it is challenged”.
As this extract shows companies have been burned in the past and have evolved themselves into complex networks or systems that pre-empt hazardous situations from learnt responses.
That gives a very brief insight to how biological mimicry has been applied to corporate branding and reputation building. Not specifically a Construction Sector related topic but in the world of contemporary UK and Global multi-disciplinary Consultant practices; Major Contractors; Joint Venture Consortiums; Special Purpose Vehicles, PFI \ PPP projects it is a complex world where image and reputation are primary elements of working relationships, trust and securing future work.
Compare and Contrast
Following the analysis of Biological Mimicry and Genetic Algorithms it is determined that the boundaries between the two techniques are somewhat blurred, but in contrast there are limits to how far the boundaries are overlapped.
A worthy map to highlight this initial statement is to recognize at where the two techniques stem from. Obviously both seize flight from the nest of nature but they are definitely not the same creatures. Genetic Algorithms are based on a single theory. That theory is Darwin’s theory of ‘Survival of the Fittest’ whereas Biological Mimicry takes its inspiration from anywhere and everywhere in nature. This includes BM taking from Darwinian Theory such as those highlighted by Eaton (2006) who included Variance under domestication, hybridisation, Genetic Drift etc.
Genetic Algorithms although described as ’stochastic’ or ‘random’ search techniques they are based on mathematical principles i.e. Algorithms being “a finite dwelling of well-defined instructions for accomplishing some task which, given an initial state, will terminate in a defined end-state”, Wikipedia (2006). This leans towards the thinking that GA’s although they mimic natural selection and randomness they are in fact calculated from numbers and figures; as such they could be described as being Quantitative in their approach.
Compare this to Biological Mimicry, which is viewed as more feeling orientated or as harmonising with nature. This principle is encapsulated in a quotation cited by Benyus (1998) in her book entitled Biomimicry – Innovation inspired by Nature, by Swan & Swan (1994):
“Nature has evolved systems over billions of years that work in harmony with each other, that build from bare, rocky, thin soil to lush, green forests. Without human intervention the processes of nature have evolved self-regulating forces of beauty, grace and efficiency. Our challenge is to learn how to honor them and be inspired by their truth to create new cultural values and systems.”
That said it could be viewed or argued that BM should be described as a Qualitative advance to spot solving. This is quite a distinct difference between the two techniques.
The previous inequity leads on to what appears to be a similarity shared by both techniques, in that both are methods of solving problems and finding solutions but there is also a contrasting viewpoint in this apparent similarity. When investigated to a deeper level it is distinct that this is a difference between biomimicry and genetic algorithms.
To highlight this point, Shin & Lee (2002) state that “GA’s aim to find an optimal solution to a given problem” suggesting ‘the best’ solution but according to Tam et al (2001) it is only a near optimal solution in a short space of time”. In opposition to this biomimicry solutions have evolved over billions of years and have “become highly optimised and efficient”, Wikipedia (2006) precise solutions.
This can be illuminated in terms of a real world situation by looking at the previously outlined applications. Take ‘Predicting Corporate Bankruptcy’ v ‘Value Ecologies’, we can see that predicting corporate bankcruptcy using genetic algorithms is confined to the rules governing GA’s whereas value ecologies have evolved from basic supply chains into ‘organic networks’ hence they have evolved to mimic biological networks that are in nature and have evolved over billions of years all the way from single cells to complex organisms.
As previously stated GA’s find a “near optimal solution in a short space of time”. It is worth noting that GA’s “often require a lot of tweaking”, Wall (2004) but with little or no effect on the overall result. The advancement of IT systems and more specifically computer power has helped with processing larger and larger amounts of data but hardware and software can still be expensive to purchase.
Conversely a benefit of genetic algorithms is that they use historical data and evolve it to learn from past experieriences, this could be of major adavantage to the unique construction industry where standardisation and repeat projects are more approved place, for example large national rollout programmes and house builders.
In conclusion, the use of BM and GA’s is widespread in the higher levels of construction and business management. Contemporary construction demands a greater degree of accuracy and sureity
They both have a justification for their individual existence due to the their inerrant strengths towards particular applications and weaknesses to others. This can be summarised by looking at reasons why they are suited to the applications highlighted previously:
Application
BM or GA
Reason for Suitability
1. Fuzzy Optimisation of Labour Allocation
GA
§ This application relies heavily on historical data
§ Historical working patterns and experience are feeble as part of the calcualtions
2. Predicting Corporate Bankruptcy
GA
§ This application relies heavily on historical data
§ Corporations are typically judged on how they have performed in the past and GA’s take this and organically evolve it to forecast into the future.
3. Value Ecologies
BM
§ It directly mimics natural networks
§ There is no eventual outcome it is a system to continually strive for Value within a culture.
4. Building and Protecing Corporate Reputation
BM
§ There is no eventual outcome it is a system to continually strive for Value within a culture.
§ It mimics a combination of natural elements and is not limited to purely Darwinian theory.
References
Aziz & Humayon (2006), Predicting corporate bankruptcy – where we stand? Corporate Governance, Volume 6 Number 1, p. 18-33
Benyus J M (1998), Biomimicry – Innovation inspired by Nature, Harper Perennial, Unusual York
Clippinger J H III et al (1999), The Biology of Business – Decoding the Natural Laws of Enterprise, Jossey Bass Wiley, San Francisco
Darwin C, (1859), “On the Origin of Species by Means of Natural Selection, or the Preservation of Favoured Races in the Struggle for Life,” p. 162.
Davis L (1991), Handbook of Genetic Algorithm, Van Nostrand-Reinhold, New York, NY.
Eaton, D (2003), Genetic Algorithms in Construction: An Introduction: University of Salford, Price Bidding, Bid Evaluation and Financial Management lecture notes.
Eaton, D (2004/05), Decision-Making under Certainty and Decision-Making under Risk: University of Salford, Impress Bidding, Bid Evaluation and Financial Management lecture notes
Eaton D, (2006), The Origins and Interactions of Biomimicry Concepts in Innovation and Change: University of Salford, Effect Bidding, Bid Evaluation and Financial Management lecture notes
Firestein P J, Building and Protecting Corporate Reputation, Strategy & Leadership, Vol. 34 No. 4.
GAUL (2006), Introduction to Genetic Algorithms and GAUL [Online] Available at: http://gaul.sourceforge.net/intro.html
Hearn & Pace (2006), “Value-creating ecologies: understanding next generation business systems”, Foresight, Vol. 8 No. 1, pp 55-65.
Metaxiotis K & Psarras (2004), The contribution of neural networks and genetic algorithms to business decision help – Academic myth or practical solution, Management Decision, Vol. 42 No. 2 pp 229-242.
Moore, J.F. (1998), ”’The rise of a new corporate form”’, Washington Quarterly, Vol. 21 No. 1, pp. 167-81
Office of the Director of National Intelligence, 2006, The Global Course of the Information Revolution: Political, Economic, and Social Consequences – Chapter 9 – Europe, Viewed 26th October 2006, http://www.dni.gov/nic/PDF_GIF_research/consequences/HTML/cf154.chap9.html
Shin, K., Lee, Y. (2002), “A genetic algorithm application in bankruptcy prediction modelling”, Expert Systems with Applications, Vol. 23 No.3, pp.321-8.
Swan & Swan (1994), Bound to the Earth: Creating a Working Partnership of Humanity and Nature, New York: Avon.
Tam et al (2001) Genetic Algorithm Models for Optimising use of Labour. Construction Management and Economics Journal. Vol 19 Nr 2, Pg 208 – 215.
Vincent JF (1998), Smart by nature – In: Lightness; the inevitable renaissance of minimum energy structures, eds. A Beukers & E v Hinte, 010 Publishers, Rotterdam, pp 42-47.
Wall, M. (2004) Introduction to Genetic Algorithms [Online] http://lancet.mit.edu/~mbwall/presentations/IntroToGAs/
Walters, D. and Lancaster, G. (2000), Implementing value strategy through the value chain, Management Decision, Vol. 38 No. 3, pp. 160-78.
Wikipedia (2006), Algorithms [Online] Available at: http://en.wikipedia.org/wiki/Algorithm
Wikipedia (2006), Bionics [Online] Available at: http://en.wikipedia.org/wiki/Bionics
Bibliography
Egan, J et al (1998) Rethinking Construction. Construction Task Force.
Obitko, M. (2004) Genetic Algorithms: Available from the World Wide Web http://cs.felk.cvut.cz/~xobitko/ga/main.html
Powell, C et al (1998) The Challenge of Change. Quantity Surveyors Division of the Royal Institution of Chartered Surveyors
Sushil J. L (2004) Genetic Algorithms and Design: Available from the World Wide Web http://www.cs.unr.edu/~sushil/papers/thesis/thesishtml/node2.html
Varetto, F. (1998), “Genetic algorithms applications in the analysis of insolvency risk”, Journal of Banking and Finance, Vol. 22 pp.1421-39.
Wall, M. (2004) Introduction to Genetic Algorithms: Available from the World Wide Web http://lancet.mit.edu/~mbwall/presentations/IntroToGAs/
Related Posts
Filed under Declare Chapter 7 Bankruptcy by on Jan 15th, 2012. 