ANIL MITRA PHD, COPYRIGHT © 1986, rev. 2004, 2009




1          Introduction

2          Details of the design process

2.1   Recognition of a need

2.2   Definition of the problem

2.2.1        Functional considerations - a holistic approach

2.3   Synthesis

2.3.1        Power system elements, mechanical components

2.4   Analysis and optimization

2.4.1        Optimization as a philosophy - a critique

2.5   Evaluation and presentation

2.6   Presentation

Document Status June 2003

Latest Revision and Copyright



Engineering design is sometimes called systems engineering or applied decision theory. To design is to formulate a plan to solve a problem. In engineering the problem is usually a need or want. Design begins with the recognition of the need. It ends with the formulation of the plan which should be detailed and accurate enough to permit execution. The process is a creative one. Many design problems have no unique or standard answers. The tools of design are the tools of engineering

The design process has many levels. We can identify overall design, sub-system design, and so on down to component design. In a sense the subsystem designs are individual designs. However, there will be iterations among the levels and this iteration may take place formally in a linear [step-by-step] way or creatively either by insight or analysis. Usually a combination of these processes will be present. The overall design will also have stages. Consider a product. In addition to design of the system [product] itself, the manufacturing process must also be designed. Although these designs may be done separately, there is an interaction: manufacturability is necessary; manufacturing cost is part of the total cost

Despite the interaction among levels of design, the principle of top-down design should be employed. In the total design process the total system should be designed in detail, first with the subsystems specified only in terms of their characteristics. Detailed design of subsystems occurs after the design of the total system. The process is obviously nested and iterative. There are ways to enhance the efficiency of the process and one is creative conceptualization. Also the details of the top-down process may be analyzed for efficiency. There is also a bottom-up process:

After the subsystems have been specified, the performance of the overall system may be analyzed. To a degree such a systems analysis may replace the iterative top-down process. This overall question of optimizing the efficiency of the top-down iteration, as well as analysis of complex systems is a part of systems engineering

The design process for a system or subsystem involves the following interactive phases:

Recognition of a need

Definition of a problem


Analysis and optimization iteration

Evaluation and interaction


The phases are interactive in the sense that any phase may iterate to any former phase. This standard model of the design process is linear only in the sense that each stage only occurs when the previous one is not iterated back. In reality design does not always follow the standard model even when it is the preferred one. For example solutions may occur before recognition or definition, either by creative synthesis or creative matching of systems with needs or by seeking a need for a “solution”. Design is not linear. However when the design is complex the linear-iterative process is a place to start. The standard model of design, embedded in a more general creative-analytic design, can be applied in any field of planning

What is the significance of the phrase creative-analytic? First note, the alternate phrase synthetic-analytic may be used. The actual approach to any problem is a synthetic-analytic one involving a dual use of synthesis and analysis in which synthesis-analysis is monitoring and evaluating, solving not only the problem but also the problem of the solution-process

Details of the design process

Recognition of a need

The need may be in any dimension - social, environmental or technical as in subsystem or product modification. Recognition of and phrasing the need is often a highly creative act, because the need may be only a vague discontent or feeling of uneasiness. Phrasing the need is also creative in that it will often be a specification of what function is required - at least in a qualitative sense. Sometimes the need will be quite obvious and specific

Recognition of the need is analytic in the sense that the need or function constitutes only one [or some] dimension[s] of the problem. Further dimensions will be specified in

Definition of the problem

This constitutes two phases. In the first, different dimensions or design considerations of the problem will be stated. This process involves creativity. In the second phase some of the dimensions will be quantified. It is always possible to quantify a dimension but it is not always reasonable to do so. The remaining dimensions will remain qualitative. The ranges of acceptability in the different dimensions will then be specified: these are the design requirements or specifications - any limitations on the designer’s choice. The design requirements will thus include both quantitative and qualitative aspects. Except when the treatment is necessarily quantitative, both qualitative and quantitative requirements can be treated at all subsequent stages of the design process. Often the analysis is quantitative and usually the optimization is quantitative. Thus the qualitative and/or affective dimensions are usually, but not exclusively, incorporated in the definition, synthesis and evaluation phases

When the optimization is quantitative, one of the quantitative dimensions is chosen as the objective to be optimized. This is often, but by no means always, the cost. Choice of objective is also a design requirement or specification. What are the usual design considerations? For complex systems the following design considerations are common:





In traditional design the functional considerations have often been considered primary and the design has been based on them. The human and environmental considerations [and others] have been considered secondary - to be applied at the end of the design process as an afterthought. However, from a simply pragmatic point of view such a patchwork approach is inefficient. As an example consider safety. Safety is a human factor [consideration]. Integration of safety into and design from the beginning and production of an inherently safe system is better than a potentially dangerous or hazardous one with safeguards added at the end: a society without nuclear arms is better than one with a vast array of nuclear armaments and even more devastating and/or sophisticated deterrents. The implied principle, that the nonfunctional design considerations should be an integral part of the design process can be applied to other designs such as social design

In traditional social design, environmental considerations would not be considered functional. However, since 1950, the world has learned that there are limits to growth and the environment has to be considered if society is to be a fit place in which to live. Even though this lesson seems only partially learned, concern for the environment has become part of the modern world culture. This is the pragmatic point of view

Going beyond the pragmatic, when we consider interest of the full human-life-society-environment system, the human and environmental considerations become functional. A well-functioning design is inherently beautiful and real beauty is functional. This principle is implicitly realized in the artisanship of traditional cultures. When the design of society and its implements [technology, institutions] does not place the human above humans or humans above environment, and when human needs are not compartmentalized, the false division of values into pragmatic and esoteric -- that is, functional and nonfunctional -- disappears. Such principles are implicitly realized in the traditions of primitive cultures

Realize the beauty in this picture. Real beauty is real economy. Real pragmatism is real esotericism: an economy of language and intellect. Esotericism and thought need not be estranged from pragmatism and action: an economy of life. Philosophers are engineers are… : an economy of institutions

One reason to keep the functional characteristics in a separate category is that in the overall design process, they are the ones that tend to vary most with the level of design. At the highest level of design, all valid considerations are functional. At the lowest level of design functional characteristics tend to be specific and quantifiable. Let us now consider the different classes of design consideration presented earlier, in their traditional categories

Functional considerations - a holistic approach

Functional considerations relate to the purpose of the design. Traditionally engineering design was functional design. However, we have seen, the nature of the functional considerations depend on the level of design. Since engineering is embedded in a larger context, all considerations that are valid are functional in an extended sense. The modern approach is that design divorced from the larger context is, even from the severely practical view, poor design. The best design is one which recognizes a hierarchy of nested levels. At any level the design will impact at least some considerations which are functional at a higher level and designers at that level must be able to incorporate considerations which are not strictly functional at their level. An alternate and perhaps better point of view is to not make distinctions between different types of consideration but between different levels of design. Designers at any level must be aware of considerations at all levels generally and specifically as they impact their work. In this sense all design considerations are functional at some level

Functional considerations at various design and planning levels


Balance and maintenance of magnitude, diversity and quality of populations, cultures and resources in society and nature-environment based on subsystem and mutual needs… Acceptance of nature’s independence from control and rejection of design for excess security; one hundred percent security is not meaningful or viable. Scenario and capacity analysis; equilibrium theory of open systems… Active and passive design - active is search and imaginative anticipation of problems. Holistic analysis, appropriate design, delivery systems; design is often called planning at this level… Can planning become excessive?

Environmental Considerations…

Considerations of conservation and replacement of depleting resources; multivalent nature of resources… efficient use of resources and planning for transition to non-depleting or renewable resources; is there a completely renewable resource system? recyclability and social measures, renewability and energy, entropy [and information, perhaps] flow; resource measures and availability; equilibrium analysis… environmental burden and quality; humans/life/environment/adaptability and diversity: needs for limits to agricultural and technical transformation of environment… resource analysis: estimation and search

Social Concerns

Equitable distribution of resources; provision for human needs; balance between humans and society; balance between society and environment; viability of population control planning and need; auto regulation. Political and macroeconomic considerations in relation to world order; appropriateness of categories in relation to global design; peace

Engineering systems

It remains true that functional considerations at any level are functional, when applicable, at the subsystem level… functional considerations specifically at the level of engineering systems are: performance; that is, function of system and integration of subsystems… life cycle analysis and performance… off-design performance… capacity analysis - includes failure mode synthesis and analysis deterministic and probabilistic; stability… weight, size, shape… response to environmental loading, enclosure, foundation and environmental monitoring and control… reliability, tolerance to faults, fault detection, warning and analysis, backup and dual systems… support systems… maintainability… control… human-”machine” interactions… assembly, manufacture and delivery

All of the considerations above are fairly direct in determining function/performance. The functions in traditional engineering are best defined by enumeration: they include transport or transformation or containment of matter and materials, motion, force, energy and power and information, or combinations of such functions and others to produce advanced functions or applications. Such applications include chemical processes, combustion, transportation, power transmission, production or multiplication of force and motion, pollution controls, fluid controls, data processing and storage, communications, and so on. A modern power plant is a composite example: in involves materials-processing and extraction of energy; transformation of energy from chemical [internal] to thermal [internal] to mechanical to electrical; data and information processing and control

The word transformation is to be appropriately understood. It includes transport, containment or storage as special cases. Thus storage of energy for later use or containment of impact energy are included. Containment or support for a load or force is included. Multiplication is included but does not imply, for example, creation of energy. In the context of power multiplication an input-output device with an external power source to provide excess of output over input is implied. Also implied are other mathematical functions such as addition, subtraction, division, integration and differentiation which are used in control systems and other systems

Mechanical systems and machines

For performance, see appropriate aspects of engineering systems… includes mechanical transformations such as transmission of motion or mechanical power or materials… kinematics and dynamic analysis of machines and mechanisms for forces and motion - theory of machines and mechanisms: determines reactions, includes deflection and vibration of elastic systems, stability dissipation of frictional heat, lubrication… strength… subsystems and interactions control… categories include containers, machines and mechanisms, robotic devices, electromechanical devices, fluid devices and controls, hydraulic systems, flow systems including pipes and HVAC

Pumps [and other mechanical flow converters], turbines [and other flow mechanical devices], some direct converters such as MHD

Thermal and power systems

Traditional systems for power such as fossil fuel fired power plants include elements of mechanical and thermal systems as well as electrical/electromechanical systems. Modern systems include bio-systems [biogas], solid state systems [photovoltaic], and wave systems [wave energy] and so on. Functional considerations are similar to thermal/mechanical

Mechanical component design

Often called machine design… design considerations for function include mechanical and corrosion behavior cost, processing and manufacture, reliability maintenance and replaceability… relevant considerations from mechanical/engineering systems

Mechanical properties include bulk behavior: strength, deflection or flexibility/ stiffness to static or dynamic loads, vibration response, toughness under impact loading, weight, size, shape, volume. Stability under static and various dynamic loadings; surface behavior: corrosion, wear, friction and heat generation, lubrication and cooling, surface finish. Materials properties and selection are an important part of component design

Other design considerations

By considering a hierarchy of design levels, we have seen that any valid consideration is functional. For example, appearance itself is not a functional consideration at the level of mechanical performance. However, when considering human-machine interactions in the context of quality of experience, it is. It is not as though human values are superfluous: they are of the essence. Single-minded concern about function in a limited sense is often reduction of life to survival. Consider also that a good functional design has an inherent aesthetic especially when the function integrates well into a higher level. Beauty and function become inseparable at the appropriate level of design.

One may debate the rationale of the groups, but the idea is convenient, because the mechanical designer tends to neglect the “other considerations”. Note that these groups correspond to the more appropriate categories suggested earlier:

Functional considerations

Human and environmental considerations

Other considerations

It is essential for the modern designer at all levels to consider those factors which do not impinge function as traditionally understood. Of especial importance are safety, ecological and societal considerations. These are discussed by Robert Juvinall who suggests the following procedures to incorporate such considerations:


Total life cycle review and off-design performance, accident scenario and case review, balanced approach to systems capacity, make safety integral, fail-safe design through redundancy and safeguards, check government standards such as OSHA, ANSI and pertinent technical literature, provision of hazard warnings and labels, human factors analysis including capabilities, communication and cooperation


Consider all aspects of basic design for soundness and alternatives, modular design for replacement [vs. obsolescence], design for recycling and renewal including maintenance, consider ecological factors in selecting materials, processing and packaging: consider pollution - hazards and degradability, durability and recyclability


Robert Juvinall States that “the basic objective of engineering design as well as other human pursuits is to improve the quality of life in our society and… That might be measured in terms of a life quality index”. He lists nine factors: physical health, material wellbeing, safety, environment, cultural, educational, treatment of disadvantaged groups, equality of opportunity, personal freedom, and population control, and suggests a weighted quantification. He accepts that the resulting lqi [life quality index] is rough and simple but emphasizes the need of this kind of thinking for sound judgments

Possible criticisms of his approach are: it does not focus on the human-environment system; with which I believe we form an essentially integrated system: the environment is to be valued not merely because we depend on it; rather we are it… Is us and this is where value lies quietly and implicitly. The approach is list oriented rather than holistic. Quantification and index analysis are unnecessary and in a way devalue value. Such quantification is only necessary when single objective quantitative optimization is to be undertaken

Alternatives are multi-objective/qualitative design point well within capacity. In a sense performance and optimization is absurd. It implies a need arising out of a burdened environment and a situation not benign to life, experimentation, freedom, fecundity and adaptation. The objective should primarily be to live within “carrying capacity” at the global level and live freely within that capacity. If progress is anything, it includes experiment and adaptation arising out of such freedom, conceived and guided by our design as well as by chance and accepting of the fall of natural systems as much as the rise

This is not to imply that optimization of certain inherently quantified systems is inappropriate. Further, the emphasEs on safety, ecological and societal considerations in Robert Juvinall’s book are overdue

It is clear that the question of design considerations is multifaceted and is a ripe field for the application of real philosophical perspective. In fact we may say that social-environmental [LEMS[1]] design is the crucible of style and philosophy

Once design considerations have been identified, the requirements or specifications, affective and quantitative should be made. This is an interactive process based on the previous step, need for design, and the next: synthesis - in which solution concepts are formulated. Practice and experience are useful


The designer’s next objective is to come up with plans for the solution which is best in terms of the specified objective and the design resources available. Based on these resources and the nature of the problem, this could be done in one step under appropriate circumstances but is usually resolved into a number of interactive steps: synthesis - conceptualizing of new and/or standard subsystems and/or components in new and/or standard arrangements as a potential solution - a number of syntheses may be done in seeking the “best” solution; analysis and optimization consists in modeling the syntheses, often analytically, to see if the design requirements are met, and optimization to arrive at the best solution

The process of synthesis is creative. There is no general algorithm

According to which synthesis can be done. In this context the study of creativity is of interest and there are certain general ideas on enhancing creativity. Each designer will have individual approaches to such enhancement. There are aids to synthesis - top-down synthesis, modeling certain aspects, building physical models of certain aspects, block models, graphic modeling including computer graphics, reference to literature/other individuals for past solutions and solution elements

It should be recognized that a design problem is never a perfectly defined problem. There are always potentially new design considerations; given a well-defined set of considerations, there are always potentially new syntheses. Design has affective dimensions. Models and tests can never be one hundred percent validated. A designer works in presence of some unknown elements. Because of this, design involves risk and creativity; and design problems, however carefully defined, do not have unique solutions

The standard model of the design process is itself a solution to a design problem - the problem of making design more reliable and efficient. The model presented here is a standard synthesis. I have discussed certain aspects of improving design performance in a qualitative way. Aspects of this question can be studied quantitatively, as mentioned earlier, as a topic in systems engineering analysis. The topic of management of design and planning is itself a design and planning problem of some importance

One element of efficient design is experience - of the process as well as of the elements and of previous designs. Much experience is stored in the cataloging of standard design elements, analyses, procedures and syntheses. This information is available in standard design and planning texts, manufacturer’s catalogs, planning reports and evaluations. I will consider examples of such lists below:

Power system elements, mechanical components

In synthesizing efficient power systems and power planning, it will be useful to have a catalog with characteristics of power systems and elements: sources and modes of energy, conversions, storage and transmission and distribution, uses; elements and syntheses of such systems; energy planning, plans and agencies; nests of design related to power… it is similarly useful to have a catalog for mechanical components… these will be provided at another time

There may also be something to be said for bottom-up thinking as a philosophy of design… and its basis in evolution

Analysis and optimization

Each synthesis has some freedom: sizes, shapes, factors that are unspecified or partially specified in the requirements. However, the design requirements may not be consistent even though this freedom exists. That is, the synthesis may not be realizable in a real system which satisfies the requirements. This is because real systems have natural laws of behavior which imply additional constraints. Such laws may be physical for physical [sub] systems, social for social [sub] systems, and so on. The validity of such laws is usually a question in science. In many cases when “pure” scientists are not concerned with the laws, they are studied by applied scientists. The laws apply to [interpretations, as states, of] the system and may be analytic-mathematical or geometric-graphical and so on. The form is usually a relation among considerations of design or design variables. Graphic modeling is common in engineering which deals with physical/natural systems. Mathematical modeling is increasingly common in many areas of design and planning. This is largely due to the efficiency of computers in processing large amounts of data and consequent treatment of large-scale problems. Physical models may also be used but are generally expensive and are often used when the number of potential solutions or syntheses has been reduced to a few. This is especially true of large-scale prototypes. The process of determining consistent or realizable syntheses is analysis. Computers have become a very commonly used tool, hence computer graphics/design/manufacturing/ engineering/planning and so on. Because of the great power and sophistication and utility of computer models, there may be a tendency to overrate them. It is true that when appropriate they are powerful and useful. There is a wide range of questions associated with reliability/validity of computer and/or mathematical and/or graphic models even when quantification is appropriate and hence model verification - a necessarily incomplete process - is an important part of model development. These questions range from the lowest to highest levels. At the lower levels programming quirks may escape detection due to complexity and at a high level we project past into present and future in assuming regularity of nature. Granting nature a measure of regularity to permit life to have meaning, and granting validity of the models employed, it remains that most designs will have dimensions that are not appropriately quantifiable. At lower, physical levels of design, the implication of such questions may be deferred explicitly or implicitly to a higher level. Thus in the use of models, judgment will be called for in the selection of model [simple or sophisticated], validity of the model, and evaluation of the qualitative

Aspects of design

It is not fully correct to say that qualitative aspects are not reducible to analysis, just because they are not quantitative. There is a perfectly rigorous mathematical theory of certain aspects of qualitative behavior as in topology and topological dynamics. Further mathematics and sequential thought is capable of adequately modeling at least certain aspects of non-sequential behavior as in modeling of evolution of continua and systems. However there also seem to be dimensions which are practically or inherently affective; that is, even if they are quantified or analytically reduced, the relation of such a dimension to the quantitative dimensions is not capable, practically or inherently, of analytic or quantitative representation, and judgment is called for

It might be mentioned that, over time, attempts at reduction of affective dimensions have been attempted. A recent one, the theory of “fuzzy sets” has had some success - critics claim that it is not essentially new - but human judgment has not yet been replaced. Based on mathematical theories of solvability, deductibility and reducibility which might also be called theories of insolvability, non-deducibility and irreducibility, we might expect that human judgment will never be replaced. However such theories are based on models of the thought process and not thought itself and hence do not preclude reduction of judgment. It remains true that judgment has not been replaced and continues to be important in design, generally in terms of the value of design considerations and specifically in deciding whether a specific synthesis which satisfies the requirements is realizable

One way of dealing with inherently affective dimensions is to use them at all stages of analyses: before, to reject obviously unacceptable syntheses, and after to see if they “sit right” and see what modifications or additions might be made. Such judgment includes engineering judgment such as processing, available size, experience - to replace extensive analysis, use of standard approaches; social and environmental considerations and so on

Another consideration that is qualitative is uncertainty. A designer or planner does not have complete knowledge of the behavior of the actual system, in part because such knowledge depends on measurement, nor of the circumstances of its operation, in part because these circumstances are not under our control and in part because knowledge of circumstances is based on prediction from measured data. Measurement and prediction are almost invariably accompanied by uncertainty of validity and/or interpretation, and such uncertainty can be highly affective. This is especially so when the dimension of uncertainty has an impact on security. Uncertainty can be made quantitative to within reasonable levels of acceptability when the impact on security is not too severe and/or high quality data enables the designer to remove the dimension of uncertainty in relation to insecurity. When the data permits the designer to assert safety with a probability above a certain safe level, the design may be acceptable. For some designs no level is acceptable as safe. Thus for large segments of society no level of safety is acceptable for nuclear power plants. Statistical considerations may be used in designing parts manufacturing systems for assembly. The parts are designed to make the parts fit properly with a probability less than one hundred percent. Parts that do not fit may be rejected, and the material recycled, but the cost is more than compensated by lower expense of the production process. In component design a safety factor may be used to reflect acceptable uncertainty arising from variations in the manufacture and processing and in the actual loading of the component

Once the existence of realizable syntheses has been determined, optimization may be done. This is the process of determining the best synthesis in terms of the objective. As mentioned, the objective or objective function is usually one of the design dimensions or considerations, such as cost, which has been singled out for its importance. It is not always possible or desirable to do this rigorously. Further, because of qualitative and affective dimensions, judgment is called for in the decision on selection and/or modification of the final solution

Optimization has often been done implicitly or by trial and error over a period of development application. When a class of syntheses is parametrizable in terms of a number of continuously varying quantitative design dimensions, alternatively called parameters or variables, and the objective can be expressed as a function of the parameters, the process of optimization can be reduced to analysis; that is, the methods of calculus. For each parametrizable class, an optimal solution is obtained and from these optima, a single overall optimum can be selected by informal process

Modern optimization has done the following: extended the theory to incorporate constraints and certain types of discrete parameters; produced efficient computer implemented algorithms which, when coupled with commensurately powerful simulation [analysis] models, permit very large scale optimization. When the evolution of a system is to be optimized, the optimization is called optimal control theory. Efficient algorithms also exist for this. It should be noted that in most algorithms the guaranteed success of the optimization depends on the satisfaction by the objective function of certain restrictive hypotheses such as linearity, convexity of the objective function as well as its domain of definition, etc. In some cases convexity can be hard to demonstrate. General algorithms exist, but are less efficient and do not guarantee success in finding the global optimum. A number of extensively developed and tested programs exist. These programs have various characteristics as to size of system acceptable, hypotheses on objective function, efficiency of algorithm, computer systems on which implementation is possible. Technical judgment is necessary in the selection and successful application of an optimization program and associated simulation model

A number of objections can be raised against single objective programming as outlined above. One is that only a single objective can be considered. This objection can be removed by combining objectives as in ratios, such as strength/weight, which is important in transportation or cost/benefit which has been used in social welfare and health care and other systems. Other ways of combining values are possible through weighted objectives. In cases where it is felt that the dimensions being combined are not comparable or that the implied compromise is unacceptable, it may be possible to employ multi objective programming. Multi objective programming works acceptably well [in giving a unique result], when the multiple objectives have a measure of independence. This is not always true

Despite persistence of objections, optimization programs and models are powerful and have significant potential for resource policy and analysis. However, they do depend on quantification and, for the same reasons as in modeling, judgment will remain essential in evaluating the validity of results and modification due to affective dimensions

Optimization as a philosophy - a critique

It has been proposed that optimization could be used at the highest level of design - at the global level. In order to achieve a further perspective on optimization, let us consider this possibility. One possible objection is that there would be too many parameters. This is unlikely. The number of parameters could be kept within bounds by subsystem analysis. Further, it is known there is no lower bound to the amount of energy required to do a computation

A more serious objection is that the design dimensions would be highly affective involving uncertainty and values of life and environment expressed through humans. Further, how would control be applied? What relative consideration would go to humans and environment? These questions are difficult. Consider, however, the implications of optimization and/or optimal control. The underlying assumption is that there is competition for limited resources of space, materials, energy, natural environment and, perhaps, time for individual men and women to develop a natural and sustaining relationship with the nature of which they are an inherent part. The planet - humanity, society, environment, and life - is a system of mutual interactions. How can it compete with itself? At this level, the idea of competition is false - without meaning, and so its attempted application is absurd. Clearly, it is possible to plan to live within the carrying capacity of the environment. When civilization and environment live well within carrying capacity, there will be room for abundance and fecundity as well as acceptance of the equality of nature with human design and consequent adaptability and adaptation, hence a value of living well within carrying capacity. Within such bounds a civilization based on renewable and/or non-depleting resources should be possible. Modeling in the scientific sense would be a reasonable enterprise with a dual, but limited, role in culture/application. Optimization could be applied in instances of local shortage and imbalance, but would not the long-term goal be restoration of such balances and not optimization? Does it not seem that insofar as it seems essential, that modern optimization at the global level would be a frantic effort to “shoestring” an implicitly mismanaged world?

At this writing no definitive definition of the associate problem of global design, or of its solution synthesis/analysis, is attempted

The problem has been considered informally to enhance understanding of the role of optimization. In this light, consider an analogy from ecosystems: predator-prey [p-p] relations in a stable ecosystem. Such populations have oscillations which can easily be explained qualitatively and quantitatively. If we are to have a useful interpretation, we should be careful about applying the world “competition”; for then we can intellectually extract a message of competition, absorb the idea at an emotional level, and apply it without discrimination. The interpretation of competition as a philosophy is ours, not that of the system, nor of predator-prey. The process may be looked on as one of mutual beneficence in which weaker members of both populations are eliminated and the gene pool is kept healthy. The combined level of the population is maintained in a rough balance by availability of resources and natural factors such as disease. If either population is completely eliminated, the consequence is starvation for the other… what are the implications for global planning? What are the origins of such populations? What factors affect the size of the population swings? What mechanisms keep the mutual predator-prey and the ecosystem? Is the performance of predator-prey in any sense optimal, or is the key interpretation one of equilibrium?… are global population and resource levels on a true exponential rise or on a shift of equilibrium levels? Is the implied size of oscillation too large in the sense that the stability of the ecosystem is being affected? Would not optimization at the group level enhance the size of such swings? Although optimization seems to have a role in the relevant subsystem analysis, what does this imply for the wellbeing of the individuals living in and as an integral element of such subsystems? We have implied that the idea of optimization and the associated competition lack meaning and reasonable application at the global level. Is this true at the group level, too? What is the structural connection between values at global and group levels? Is there an appropriate balance between optimization and wellbeing at the group level? Is not the need for optimization felt because we have been too efficient in exploitation and survival and too wasteful in use? Is there not an economy in healthy ecosystems that we should like to imitate? What is the relevance of this in the sphere of social interaction and what are the implications for human values? Can the question of stability or stable oscillations of acceptable magnitude be formulated as a problem in optimal control? Need it? Should not nature itself come before theory?

I feel some justification in concluding that the role of optimization in global and large scale planning is limited but otherwise open. There are valid uses of optimization, probably in situations where a certain type of competition is appropriate, but in balance with other dimensions. The present use of optimization, not merely as a tool but also as a philosophy, seems to be an attempt to extend the activity and life of inherently unhealthy attitudes with a probable result of possibly catastrophic postponing the inherently self-limiting action of such attitudes. Present global goals could be twofold: first, living within the carrying capacity of the planet and attempting to define and adopt healthy interactions at various levels and, second, identification, action and planning for potentially catastrophic but controllable threats such as nuclear self-destruction. Long-run goals for optimization include: characterization of systems and situations such as crisis and/or short-term, and contexts such as global design, systems for which optimization is of value; extension of this domain by appropriate choice of objective or multi-objective, programming including optimal controls and introduction of system time constants and dynamic optimization or feedback; reflection, observation, analysis and formulation of essentials of relevant dynamics, not for mere performance, but for understanding including meaning and applicability of quantification, appropriate system levels, categories and interactions, adequacy, inherent recognition of carrying capacity, values such as appropriate insecurity, living in and not above nature, recyclability and renewability; need for monolithic vs. appropriate theories; understanding and insight vs. prediction; recognition of affective dimensions and interaction with quantitative

Evaluation and presentation

Evaluation is a final proof and evaluation of the success of the design by the designer and/or design staff. Evaluation often involves testing of a prototype in the laboratory or the field and is necessary because no amount of planning and analysis can guarantee, among other things, that a complex synthesis will perform as expected or predicted or that all aspects of behavior, synthesis, function and malfunction have been covered in the analysis. Thus evaluation often includes final evaluation for function including reliability of a prototype over as much as can be afforded of the total anticipated life cycle of the system. The questions to be asked are: does the design really satisfy the need[s]? Is it reliable and safe? Will it compete successfully with similar products? Is it economical to manufacture and use? Is it easily maintained and adjusted? Can profit be made from its sale or use? Essentially, evaluation is a review of all design considerations and requirements in the context of a situation which is as close as affordable to the actual application or use - for example, a prototype in a simulated life cycle. In fact evaluation is an ongoing process and continues after sales have been made and is applied [with further consideration of need, problem [re] definition, synthesis and analysis and optimization as necessary] to product modification or replacement. Viewed in this sense, design can be seen as an ongoing cooperative social interaction between manufacturer and user. This is somewhat different than the traditional view in which the production and use phases are separate and in which there is an element of competition between producers and users. Viewed from a social design perspective perhaps both aspects are desirable: production and application are both aspects of an inclusive process and the risk as well as checks [evaluation] and profits could be shared by both manufacturer and user. The share of responsibility and profit and risk need not be equal, but for any party to be assigned all of the risk or evaluation would be unbalanced in a social design of mutual benefit. Since manufacturers control the design and process, they should be assigned a greater share of responsibility for risk and perhaps a greeter share of profits. The over-system shares in profit through taxes. Also, focus on profit measured financially is incomplete. If the product fits into an overall social design, then society benefits by enterprise. Total control and total freedom of enterprise probably represent inappropriate polarizations. Therein there is the question of type of control: through enforcement and/or encouragement. A mix would be appropriate. Such questions have been considered by many including Marx, economists, political, social and global leaders and theorists. The ideas of Schumacher and the approaches in Japan which include the concepts of shared risks, responsibilities and benefits as well as the idea that the means of production, the workplace, should be practically aesthetic - harmonize with man’s nature in the context of an overall socio-environmental design, are interesting in that the application has demonstrated a potential value and deserve further consideration and experimental application on various scales


The final phase of design and planning is presentation to the next level of decision making. The objective of a good presentation is to make people feel and think well about a good design. This includes a balance of accurate but understandable and attractive presentation as well as persuasion, advertisement and ongoing leadership. Occasional failure is expected, especially in an ideal environment of critical appraisal. The formal means of communication are written, oral and graphical, and development of these skills requires practice and experience. Actual communication includes formal communication as well as human and social aspects

Document Status October 2009

May be useful if I return to engineering

Details may be useful for Journey in Being; however, essential ideas absorbed to Evolution and Design and to Journey in Being



Anil Mitra PhD, Copyright © 1986, revised Thursday, October 15, 2009


[1] Life, environment, man and society