Computer and the Mind of Man : Managers and Models (1962)
Sign in to track this film in your collection or want list.
Creator: A/V Geeks 16mm Films
Description: The film discusses the significance of mathematical and logical models in decision-making within business and science, emphasizing the role of electronic computers in simulating and optimizing designs before actual implementation. It illustrates how simulation techniques can predict outcomes and improve efficiency in various fields, including space exploration and chemical engineering. The use of mathematical models allows for better-informed decisions, reducing trial and error and saving time and resources. The film highlights the evolution of operations research methods and their application in modern business management, showcasing the increasing reliance on computers for handling complex information and optimizing processes. Keywords mathematical models, decision-making, simulation, electronic computers, optimization, operations research, business management, efficiency, trial and error, space exploration, chemical engineering Email us at footage@avgeeks.com if you have questions about the footage and are interested in using it in your project.
Transcription
[Music] traditionally the best and sometimes the only way of finding finding out whether or not a new design or product is practical and acceptable is to build it and try it out the same principle holds true for decision making in almost any aspect of human Affairs the test of the idea or decision is the test of reality however when we can build and test a model of a design or a real life situation we frequently can save a great deal of time money and grief and the most powerful of the many kinds of models are mathematical and logical can be manipulated and tested on the most versatile of modern man's inventions the electronic computer in this high-speed technological world of ours A MPL decimal point or an incorrect number can mean an error of gigantic magnitude but even more serious is the error of not having adequate current information available to [Laughter] us and in the highly competitive business world of today the penalty even for small errors is suddenly very high to an increasing degree we seek the ability to predict the consequences of our plans and actions before putting them into effect it is probable that many businesses today have considerably less latitude for trial and error in planning and decision making than they assume they have so much of our decision- making depends upon our ability to handle and to interpret information in science and in business our store of information increases more rapidly with each day by the close Market how are we to handle the sheer volume of information so critical to informed decision making within the past several decades methods and machines have been devised which have become invaluable extensions of our information handling and decisionmaking ability accurately and tirelessly electronic computers can trace out the consequences of a thousand possible actions can pick out the best design among thousands of possible designs thus providing a better base for the informed exercise of human judgment whenever the elements and operations of a system can be Quantified or expressed in measurable terms a mathematical model of the system can be constructed and manipulated on a computer this technique is known as simulation for example the performance of this the most powerful America's space Vehicles the design of this chemical plant and the operation of this the world's largest sugar refinery were simulated in advance of actual construction or operation by means of models made entirely of equations and numerical data manipulated on a digital computer where once we had to build something in order to test it we can now simulate and test the performance of machines of chemical and other processes and even complex human organizations by means of electronic computers on this computer at the Marshall space flight center in Huntsville Alabama mathematical formulations relative to the design and desired performance of the Saturn rocket were processed improved and reprocessed many times in other words a mathematical mockup of the rocket was tested and studied before an actual test result we have just witnessed a static test of the booster stage of the Saturn space vehicle the Saturn space vehicle is the largest one at the present time under assign in the United States this is Dr helmouth Hol head of the computation laboratory at the Marshall space flight center and it will have 1 and A2 million pounds of thrust the purpose of this test was to find out lots of things about the engines we wanted to know whether the engines really developed the men horsepowers and whether we had that many pounds of thrust and we want to know whether our fuel consumption was right right and whether we had excessive vibrations and what were the frequencies of those vibrations we also wanted to know lots of things about the interactions of the eight engines we have to propel the booster and things like stress and strain and heat protection and so on but we wanted to know more we wanted to find out whether our predictions arrived from a mathematical model or mathematical techniques for this test were right computers have played an extremely important but relatively unpublicized role in America's Space Program only a computer for example can handle the vast quantity of computations necessary in calculating a trajectory and in relation to the design work on the Saturn Dr Holzer commented if you can only save one single Saturn firing by using computers you can pay for quite a number of highspeed digital computers and this is why it is just absolutely necessary to use mathematical models and their Solution by computers to proceed in this design so the whole process works about like this we start out with some idea and we still don't know enough uh coefficients and enough actual values so we take the best thing we have and come up with a mathematical model then we start designing something or some parts and after we have the parts we test them and make sure in the test that either we have made the right assumptions or we gain new experience we then improve our equations and come up with new results make new tests improve our equations again and so in this using this cycle we finally then come up with a complete design and then after we have this we want to make sure that once this uh vle goes to the launching pet that the first firing will be a success the first launching of the Saturn was a success due in no small measure to the many simulated firings run on the computer before the actual launching of the bird a team of Engineers and mathematicians used a computer to select the economically Optimum design for this chemical plant in fact in a final test of the system the computer designed a chemical plant that would cost between 250,000 to 600,000 less to build than the best previous design of a group of Engineers the system is called chops for chemical engineering optimization system and is explained here by Dr Thomas Baron director of engineering research for the shell Development Corporation chops is a is a system of Designing chemical plants to start with I I probably should show you what a typical chemical plant looks like uh let's assume that we wish to convert in a reactor which I'm just drawing a reactant which I will call R I want to make out of this reactant r a product p and this reaction occurs in this vessel now the difficulty is that not all of the reactant is converted into the product only part now the bigger the vessel the more of the reactant is converted into the product now somewhere along the line I have to separate the product that I have obtained from the reactant this is done and say a distillation column or some other separation unit the unconverted reactant is cycled back to the reactor so this is still reactant and out of here here comes the product now this is an oversimplified plant but a typical one the chops program works with three different sets of information first is the logic program containing the instructions for what the machine is to do the second contains information on the units that make up a chemical plant such as reactors and distillation columns there is also information on how the various process stream flow and interconnect along with the many limitations on operating temperatures and so on the third set of data contains technical information on the process streams and economic and legal information such as the cost of piping labor rates and even zoning restrictions uh typical plant would involve such such variables as pressure flow rates and a number of of maybe as many as 50 to 100 200 50 variables that we have to look at but we're always interested in finding that set of all these variables which makes the whole thing the cheapest Now by cheapest I mean not only that it's cheapest with respect to the the materials that are consumed and the labor that goes into building the plant but I'm talking about the amortised cost of the product which includes the cost of the vessel as well as the operating cost the plant is running so generally speaking then the problem is to minimize this amortized cost of the product and this will depend I will write this the way a mathematician writes it I will say that it's a function of but all I mean is that it depends on this will depend on a number of things and the number of things it depends on I will call X1 X2 X3 and as I say I may have several hundred of these X's with its bundle of mathematical information supplied by the engineers the computer following the instructions in the chops program works out its first design then following instructions the computer begins in effect to ask itself questions what happens if the temperature in the reactor is raised 5 de or what if the amount of chemical a in the feed stream is increased from 49 to 51% the computer designs a plant based on these changes if the design is an improvement it changes all of the appropriate variables in the direction of the Improvement and so on over and over again until further changes in the variables fail to produce a more economical plant the computer then has reached the optimum design it stops Computing and prints out the results is a system which includes the mathematics required for optimization for seeking the maximum or the or the minimum was all the things that can be said once and for all about chemical plants this includes uh instructions for the machine on how to accept data from Engineers on how to print out so that even Executives can understand what is being said uh on how not to violate obvious laws of nature for instance the number of pounds that enter this plant have to equal to the number of pounds that leave it the amount of energy enters a plant has to equal to the amount of energy that leave it and so now all these things are done once and for all this is Approximately 80% of the work total work that has to be done as far as instructions the machines are concerned now the other 20% have nothing to do with mathematics it simply represents feeding of data in which the engineer tells the machine what is different about this plan uh what what distinguishes this PL from all other plants uh this particular reaction from all other reactions and this particular flow scheme from all other flow schemes now I would say that the creation of this system this generalized system which we call chos took uh oh 15 20 man years the application of CHS the instructions of the to the machine in a particular case take anywhere from two man months to say a man year depending on the complications of the of the plant now the machine may take in a single optimization and seeking the optimum ones anywhere from a few minutes up to a few hours an accurate mathematical reflection of exactly what happens under a given set of operating conditions in the world's largest sugar Refinery can be worked out on a computer in approximately 30 seconds in this book there are several hundred representations each illustrating what happens in a Refinery under a slightly different set of operating conditions this information could not possibly be obtained if the refinery itself had to be specially set up for each such test these simulation studies are possible because of the development of a mathematical model programmed for the IBM 1401 computer and designed to reflect the average 10-day output of the refinery under a variety of conditions and levels of operation we asked Richard Brooks coordinator of operations research for the cnh sugar refining Corporation about this work during the past several years the production requirements that were placed upon the Crockett Refinery were such that it became virtually impossible to conduct various tests with in the refinery itself as a result top management uh authorized the development of a mathematical computer model of the refinery The Hope was to develop a type of analytical pilot plant for the refinery itself uh the study brought together the technical knowhow of the refinery Neil Pennington Chief chemist research for cnh with a competence in operations research Earl Isaac of the management consulting firm of fair Isaac and Company uh the final result was a simulation model of the refinery it involved the interaction of 500 variables and 100 product flows and uh takes 880,000 steps to complete with this simulation model we can now make uh test runs to determine whether or not a particular change in a process variable or an input variable will give us a improvement in our plant processing by describing the comp components individually in a mathematical sense and then describing the way in which they interacted also in a mathematical form it is possible to infer how the overall system will operate under conditions that have not been encountered in actual practice so in that sense it is a a pilot plant where the knobs can be twisted in this case the equivalence is changing the numbers within the computer to arrive at unknown operating points this simulation model is able to measure in terms of color Purity and quantity end products and intermediate flows the effects of variations in melt rate raw sugar Blends raw sugar polarization raw sugar color and sales demand it is in effect an analytical pilot plant where the effect of certain Refinery decisions can be measured individually and in terms of their effect on the whole refining process the computer prints out its answers on a flow diagram of the refining process each of the boxes represents a step in the refining process beginning with the transfer of the raw sugar from the bins to the mouse the sugar crystals are then dissolved in large melt tanks and the liquid is then then sent to the Char House to undergo two types of filtration from the Char House the liquid sugar is separated into various grades and sent to the vacuum pans for crystallization the refined sugar is then dried in large granulators sorted according to Crystal size and packaged in a variety of packaged sizes and grades this simulation model of the refinery is not the first model to be used at cnh as Richard Brooks indicates we started off three or four years ago and developed a simulation model for the shipment of raw sugar from the Hawaiian Islands to cocet where our Refinery is located uh this past year we have just completed a simulation model of the refinery itself and we have on the drawing boards uh ideas for models which will uh predict sales determine the net Financial return to the corporation our general thought is that perhaps uh in The Fairly near future we will integrate all of these models to have a truly systems wide model of the entire operation of the company designers Architects engineers and scientists have been using three-dimensional models for a long time these for example are models of various kinds of viruses but the kinds of models we are interested in here are Ma mathematical models of operating or dynamic business and Industrial Systems the books that a company accountant keeps are a simplified model of the flow of goods and services through a business organization from the books it is possible to measure such things as rates of flow to measure the performance of various parts of the company and to determine the values or profits produced businesses have always had models of this sort designed to give a current picture of a company's condition but in addition to wanting to know where the company stands today the manager also wants to be able to predict where the company may be in the future did managers make decisions based on mathematical models prior to the recent scientific developments in business management we asked Dr Ernest kbur professor of industrial engineering at the University of California well the funny thing is they must have used them whether they recognized it or not uh I have heard a manager say that well we really don't don't use any forecasting we just say that next year's business will be like this year's which is the equivalent of making a forecast uh and generally when they make decisions about changes they are doing it in terms of a model which may be very poorly defined and Ill defined but somewhere in their mind they have a picture of the effects of their action the modern business executive has been described as an information converter that is a man who converts information into decisions which trigger actions designed to obtain certain objectives frequently his main problem is to decide which information Source has priority over another when there is irrelevant information and when there are a great number of minor decisions required it is altogether too easy to lose sight of the possible effects of a single decision throughout the total company operation ideally the executive needs a method which enables him to study the parts of the business in relation to the whole operation of the company and a means of tracing out in advance the effects of decisions throughout the system the point is that the president of the company is involved in day-to-day business and he has to consider actions on a day to-day basis he doesn't have time to sit back and consider the relationships and Inter relationships within the business he generally doesn't have the ma mathematical background to develop a model that is accurate and workable even if he he even if he could write a model he couldn't he couldn't uh carry out the necessary calculations I'm sure he can understand the result once he got them but first the results must be produced under the generic title of operations research methods first developed by scientists in relation to military strategic and tactical problems in World War II are today being applied to many business operations particularly the petroleum industry I don't think there is a petroleum company that is not making extensive use of operation research particular linear programming and extensions of linear programming they have to to stay in business linear programming is a mathematical technique used in operations research whereby a business system is analyzed into its basic components or activities and these elements are then manipulated in such a way as to optimize profits or minimize costs for the system and the high-speed computer is an essential tool for this kind of practical business research mathematical modelss for business problems that can be fed to a computer range from what combination of raw materials and Manufacturing resources will result in the most economical production of a given product to the problem of selecting sites for new plants and stores so as to maximize profits and meet customer demand frequently a single mathematical model can be applied to many different kinds of businesses as professor Professor George danig widely heralded as the father of linear programming commented well a standard joke in my house is that uh when I come home in the evening my children ask me Daddy what industry are you in today this gives some index of the broad number of applications most management decision making involves many Alternatives rather than a simple choice between one decision that is right and another that is wrong to today the sheer volume of information to be considered and the complex interdependence of the factors involved are more than the uned human being can handle electronic computers together with techniques such as linear programming are creating a revolution in the field of business management and decision making without the computer there are many problems involving information and control which we simply could not solve increasingly the computer is being used as a kind of electronic library making available information which would otherwise be too costly and timec consuming to obtain it is the computer that gives the manager the ability to predict the consequences of decisions and the future state of a business system and it permits the manager to make better informed decisions all of which adds up to increased power and knowledge which is a property of men not machines within 10 years it will be obvious to everybody that a revolution has already taken place you can couch a problem in directly in computer terms and arrive very economically and readily at a solution so you're thinking in broader principles and structures and Inter relationships rather than in the small and very arduous uh component limited thinking that you're restricted to before the problems that that high level management will space will continue to exist and computers will make some contribution to them but I don't think that we're going to replace the manager and not all the middle level managers are going to be replaced either so that uh it's a revolution but let's say it's a it's not a not so bloody it's a got some friendly [Music] characteristics e [Music] [Music] this is NE National educational television
Online Copy: https://www.youtube.com/watch?v=_d9FjqQf634
Metadata Source:YouTube
No holdings listed.
No related films.
Record added: 2026-05-28 17:56:47