smagloworld


smafam


smaveg


smagroup


smanet


smalap


  Introduction  Policy  Traction  Extension  Growth  Incentives  Projects  Benchmarking  Models  Decisions  Transparency  Bioclimate  OQSI:1  PPPP  GBF  Contacts  Author


Since the 1940s computer-based models have been used to analyse complex technical, economic and financial issues. Most advanced operations research analyses cannot be performed without first of all defining a model of relationships in the form of a mathematical model. The following are some examples of the use of models in analysis and design:

  • Simulation of 3 factor agricultural sector economic growth model 1968-19691
  • Simulation model of plant biomass response to bioclimatic factors 19702
  • Simulation model of coffee crop images to design recognition system for coffee inventory and production prediction 19713
  • Simultion model of river basin precipitation run-off to reduce pollution due to agricultural production 19754
  • Simulation model for business processes to optimise design of processes 19905
  • Simulation model of farm production possibilities 20006
  • Simulation model of Real Incomes Approach to Economics (PAC model) 20017
  • Simulation model of Price performance Ratio impacts of Real Incomes (PAC model) 20068
  • Simulation of land consolidation system options 20149
  • Simulation model of commodity balances 201610


Models are a conceptual, graphic, written or oral representation of an existing or imaginary situation of inter-relationships between factors that influence something of interest. The structure of human thought and deductive processes that lead to decisions is based on the perceptions each person possesses based on their their individual instruction, information available and personal experience. People tend to structure their thoughts in the form of models of varying complexity.Deduction as a process that can lead to a decision is usually based on the formation of a mental model of relationships based on the existing knowledge of the subject, experience, and often, some assumptions. People will assess another person's propositions based of the confidence they have that the model proposed by that person reflects their own understanding concerning the issue being discussed. The basis for assessment is the degree to which the description of how a model operates, the quality of information used in constructing the model and the likelihood of events that affect the outcomes that would result from such as model.

Models are an essential analytical tool for complex subjects that involve by necessity a multi-disciplinary or systems group approach where many factors need to be taken into account. The first step in establishing good systems group communications is for participants to come to an agreement on the model that represents their common understanding of the topic of interest. Once the model is agreed upon it can then be used to analyse factors and be used to project different "scenarios" to identify, for example, a preferred solution to a problem.

It is common practice in technology domains to use models to identify ways to improve processes and because of the relative complexity used is made of computer-based models of the process to simulate different options to, for example, lower costs or improve quality of output. The main benefit of using models to evaluate options is well-establshed in terms of the need to avoid a commitment of resources to an imperfectly designed option, leading to a potential loss of considerable resources and time.



The use of models in analysis and decison-analysis encourages people to externalise their contributions in an explicit and clear fashion as to what they believe key relationships to be. If a model cannot be constructed then clearly the existing information and knowledge is insufficient for taking decisions in an evidence-based fashion. Quite often the effort of building a model will identify irrational assumptions because they cannot be featured in the model. In running simulations there are normally expections of what the results will show. When results are not as expected the situation can be that:
  • the model does not reflect the current state of knowledge
  • the model is correct but the state of knowledge is inadequate
This can identify missing information both in terms of quality and quantity.



Models tend to be built up based on physical relationships and the unit prices are introduced to create an economic model and the time dimensions converts this to a financial model.



The logic used to express how a model is built and operates has to closely parallel the way humans think and deduce otherwise it would be virtually impossible to understand them, let alone build them. The way human deduce and reason as a basis for taking decisions was described as a mathematical logic in 1854 by George Boole in his book entitled, "The Laws of Thought". This work provided the rationale and methodology for reducing complex logical relationships to simpler sets of relationships which can reproduce all of the possible relationships from which the set was derived. Some 83 years after the publication of this book, Claud Shannon identified the role of Boolean Logic in the optimization of the design of electrical switching circuits in a PhD thesis published in 1936 (University of Michigan). As a result Boolean Logic became the foundation of modern digital technologies (hardware and software). Because this is founded on the basis for human expression and deduction computers can mimic human logical processes as well as describe human understanding of complex systems. In the 1960s computers were controlled by a business oriented logic based on FORTRAN. In the 1960s, some 30 years after Shannon's contribution, Kristen Nygaard had become concerned with changing the syntax of programming languages to enable a more precise description of reality so as to create better simulation models. He was later joined by Ole-Johan Dahl, and togther they developed what is now known as object oriented programming (OOP) and a series of simulation models known as SIMULA. They both worked at the Norwegian Computer Centre in Oslo. OOP became more widely adopted after is was introduced to the popular C++ language in the 1970s by Bjarne Stroustrup, based on his PhD work at the University of Cambridge.



The important change introduced by OOP was that all phenomena are identified as objects (people, animate and inanimate things) which have specific properties (dimensions, weight, colour etc) and methods (what the objects can do or do and how they carry out these actions or the response of objects to impacts of other phenomena on that object). This basic desciption enables any real world phenomena to be described and coded as a computer program. The main constraint on progressing with such a powerful simulation oriented language is the lack of experience of many people working in many different domains in expressing their knowledge in an explicit format so as to be able to program this. Logical syntax, and expecially Boolean Logic of logical relationships, provide the solution to this problem but there remains a significant constraint caused by the lack of exposure of many people to this form of expression.

1  McNeill, H. W., "3D production function", Food Research Institute, TP, University of Stanford, 1968; and McNeill, H. W., and Jino, M., "Simulation of 3DPF", CNAE, National Research Council, Brazil, 1969.
2  McNeill, H. W., "A quantitative model of the interaction of bioclimate and plant biomass production", National Research Council, Brazil, 1970.
3  McNeill, H. W., "Coffee crop recognition algorithm simulation", Plant Production Division, FAO, Rome, 1971.
4  McNeill, H. W., & Serra, R. "The McNeill-Serra Model" for river basin catchment areas, Mogi Guacu River basin, Institute of Geology, SP, Brazil.
5  McNeill, H. W., "The CRESTFILM business process simulation model", Manpower Services Commission, Department of Employment, UK Government, Soft Horizons, Portsmouth 1990.
6  McNeill, H. W., "Farm production optimization system", Online system, Hungarian Agricultural Development Foundation, AFA_Agronet system, Budapest, Hungary, 2000.
7  McNeill, H.W., "The PAC Model", Online system, SEEL-Systems Engineering Economics Lab, Portsmouth, UK,
8  McNeill, H.W., "Price Performance Ratio Impacts of Real Incomes", Online system, SEEL-Systems Engineering Economics Lab, Portsmouth, UK,
9  McNeill, H. W., "Land Consolidation Co-operative Model - proof of concept", Online system, SEEL-Systems Engineering Economics Lab, Portsmouth, UK,
10  McNeill, H. W., "Commodity balance calculator series", Online system, George Boole Foundation, London, UK, 2016.

The Decision Analysis Initiative 2010-2015
George Boole Foundation