| A/E/C SYSTEMS Computer Solutions | Winter-Spring 1992 | page 6 |
by Charles R. Carroll. Jr., FCSI
Much has been written, pro and con,over the past few years about the value of computer-assisted design and the managementofthe design function. Arguments have been assembled by proponents of each point of view to prove the validity of their positions. An indication of concern: Business Weekrecently reported that in the last 10 years the total increase in U.S. office productivity was less than 2%, even though a third of all capital expenditures have been for information processing. As Peter Drucker points out in a recent issue of The Harvard Business Review. "The single greatest challenge facing managers in the developed countries of the world is to raise the productivity of knowledge and service workers. This challenge, which will dominate the management agenda for the next several decades, will ultimately determine the competitive perform- ance of companies. Even more important, it will determine the fabric of society and the quality of life in industrialized nations."
Drucker's prescription for a l990s version of the Industrial Revolution is a pragmatic series of steps to work smarter" using all the tools available, such as the computer and other electronic devices.
Lest anyone in the design professions think these problems do not apply to them and that architecture and engineering are not businesses, don't believe it. A prominent member of AIA, Charles Luckman, FAIA, put the problem into focus in a pamphlet published in 1968, Architecture IS A Business. The points he made then are as valid today. His major thesis was that "while design is of utmost importance, it must be a partner to the total concept. It must be brought into focus with the total process of architecture and engineering, which includes design, fine arts, engineering, construction, and the business and economic aspects of the project." To survive, compete and prosper, a business must be productive. As Tom Young pointed out in his article in the Autumn, 1991 issue of Computer Solutions, 'We would not be discussing computers ... if we did not believe we could use computers to improve efficiency. profit and competitive edge."
Why then, with all the computers and computer-based workstations in place in A/E firms in this country, is A/E computer-based productivity still poor com- pared with similar computer use in other industries? The use of computer-sup- ported and CADD-based design programs in the aerospace in- dustry, electronics, shipbuilding and similar design-oriented businesses has resulted in large increases in design productivity. What is it about the design professions, architecture in particular, that makes their use of computers less productive?
A number of studies have been made to determine the reasons for this failure. Architect Charles Thomsen, President of 3D/I, a large Texas firm, put it succinctly when he said, "Someone comes in and automates a manual approach to save time. That doesn't work until you do the task differently. " Many firms have been applying the power of the computer to existing tasks without redefin- ing the way the task should be performed, e.g.: Don't use the computer as a faster typewriter or fast 2-D drafting machine. Use it as a source for an automated specification database or as a 3-D design tool. Detailed investigation in numerous firms revealed that the real problem is the way computers are being used, often solely to automate manual pro- cesses such as word processing, ac- counting or drafting. Computers are used for some application that previ- ously was done manually, and the same procedures are being followed.
Examples of computer misuse abound. Some would be funny, were they not so tragic. An example is the prominent design firm currently designing a large project for an overseas client. The designer refuses to use CAD but is having all the drawings converted from paper to CAD at the client's demand. So why not create the design in CAD in the first place? If nothing else, the project would benefit from improved accuracy and reduced cost. More than likely, a better design would result because more options could be investigated within the pa- rameters of fee and time allocated.
However, in the many firms where productivity and creativity increases have been found, benefits are uniformly attributed to computers used at managerial or high skill levels to enhance user abilities and thinking. For A/E firms this means better project control and more opportunity to explore alternative design solutions.
Computers Don't Make Decisions, they Support Them.In his PC Productivity Book (Simon & Schuster, NY 1991), Jim Seymour points out that computers should support design and decision-making, not be decision-mak- ing systems. He illustrates his point with these comments from a Fortune 500 company executive: "The one thing I really hate about computers in general is how they encourage people to come in and tell me what 'the computer says.' I don't give a damn about what the computer says! We bought the computers for people to use to become better managers. I don't want to know what the computer says; I want to know what they say."
It is said repeatedly that many people don't like to use computers because com- puters require training. We don't spend megabucks learning how to use electron- ic devices like telephones and photocopiers. Why do we spend so much time and money learning how to use computers? Instead of people having to learn to think like computers, why can't computers "think"like people? A number of groups have been thinking along these lines to develop solutions, which range from speech recognition and higher levels of artificial intelligence to the most basic linking of default options.
How does this relate to productivity and creativity--not only design cre- ativity but all of the other construction areas that require assembling infor- mation, making decisions, and arriving at an end result? A review of efforts currently under way to make computers more useful supporters of design and creative decision-making reveals some interesting players and programs.
One well-known commentator on the design professions contends that architects and engineers are truly "assemblers" of information. They take existing graphic and alpha-numeric information and combine it to produce projects. Whether information is accessed subconsciously as the design is worked out manually or a large computer database is accessed by some expert system, an information database is in use. Some say the genius of creativity lies in the ability to bring together existing information, ideas. concepts and designs in original arrangements. This point of view certainly seems to have merit, particularly if you consider how difficult it must be for anyone to come up with a totally new idea or design concept.
How can computers help A/E's achieve these design and construction goals? The more information that is easily available ror review and integration into the development of a design, the more freedom the professional has to explore solutions and to exercise appropriate creativity.
The key to productivity, however, is having information easily available in an organized format. This is where computers can earn their keep.
Increasing attention is being paid to a combination of systems in which the user assembles large amounts of organized information and reviews it to determine the most applicable pieces. Such a system can result in an "expert" system or in a teaching system for use by those less expert than the person who is developing the expert system.
Computers Organize and Assemble InformationWork is proceeding on at least three general system types that use computers to assemble and make information available for decisions that otherwise would be difficult to reach as rapidly. The time and effort saved in assembling inforrnation often gives the user enough time to make more accurate decisions. He or she can investigate design and information options much more broadly within available budget and time constraints. Among the systems being used and developed are some that combine graphics and other information; some that develop specifications, estimates and other alpha-numeric information; and others that carry the concepts to the logical end result, such as expert systems and artificial intelligence.
One basic semi-expert system concept is illustrated by the Specification Gen- eration Package (SPECGEN) developed by Intergraph using MASS-II word processing software. In this system the user develops a DEVICE/ DIRECTORY, which contains the master specification and the equivalence file as created by the system user. The equivalence file is just what it sounds like: it matches specifica- tion sections, paragraphs or other text to the details created in the CAD system and sends the text to the word processor for project specification output. This sys- tem is an example of a first step users often take in developing computer- assisted systems: creating an equiva- lence file between standards, drawings or details (the graphics file) and the master specification.
Others have attempted to develop similar systems on a national basis, but until just recently, the lack of a sufficient number of standard details and graphics related to available mas- ter specifications has made such programs difficult to implement. There are those who feel that a combination of graphics, speci- fications and cost estimating should be developed un- der an expert system. Some firms have done this in house, developing an integrated data base of costs, graphics and specifications. Autodesk and Timberline are addressing these integration needs jointly and have several programs under development. CSI and Vertex also are working towards these ends, with some programs becoming commercially available.
Greatly increased productivity and better management of drawing development and use can come from standardization of drawing format, just as it has from use of MASTERFORMAT, for other construction documents. Onkal "Duke" Guzey. AIA, and Jim Freehof, AIA have pointed out how the design professional would benefit from a systems approach to the linkage of CAD, specifications and cost estimat- ing. The system they have developed is named ConDoc. Guzey quotes Buckminster Fuller: "You don't change the old by resisting it, you change the old by making it obsolete." This belief may be responsible for the Freehof/ Guzey success in developing acceptance for the ConDoc system and its increasing adoption by A/E's for use with drawings and specifications.
NAVFAC, the Navy construction group, and several government agen- cies are involved in various levels of development for expert systems. How- ever, a first step taken by many is in the direction of simple equivalence, where the selection of a standard de- tail or graphic triggers selection of a specification section covering that item, as in the CSI/Vertex system.
Little professional effort is necessary, comparatively speaking, to develop such an equivalence since many decisions are obvious or have been made previously. The computer has to be told about these equivalencies in order to make the system work. This alone is a giant step forward in pro- ductivity and accuracy, since it gives everyone concerned an automated control of the match-up between specifications and standard graphics.
Other areas of activity in the development and application of expert systems range from medical facility design through the maintenance and operation o f machinery to cook Campbell soup. The basic principle behind expert systems is the application of existing human expertise to identifiable decision options, storing that expertise and knowledge base in a computer program based on some- thing like a menu selection grid. The hardware requirements for expert sys- tems can be accommodated on the typical microcomputer. It's the software that has changed--and makes the difference.
There are three essential differences between data-processing and knowledge processing software:
Futures are fine, you may be thinking, but how can I improve the management, productivity and profitability of my firm today, particularly when the compe- tition is fierce, and work is so difficult to get?
If we can agree that architecture and engineering involve putting people, ideas and information together to produce a building or other project, anything that will improve that process and make it more efficient should be useful. Comput- ers promise to be able to do just that, "but only when they are used as infor- mation machines," as Dr. Kaiman Lee said so well. We have all the hardware we can use. The things we must look at carefully are the programs and proper use of computers.
Proper computer use is something each firm must address for itself in order to benefit from the machine. If you use them only as drafting machines or simple word processors, investigate ways of creating project databases of specifica- tions, drawings, cost estimates and other project information. Network your computers so that everyone involved on a project can have access to all project information as it develops . Put your best people at the computer workstation. It's wasteful to put the power of a $20,000 workstation in the hands of a draftsman or clerk.
Investigate the programs under de velopment by APEC, AIA, CSRF and others for project design and informa- tion management. Take a look at the expert systems now available. Save time developing details, and increase uniformity by using the programs sponsored by CSI and Vertex, ASG and ASI.
Some time ago a speaker at a National Academy of Sciences symposium gave us a look into the future and a few projections about the future of computers. One of his salient points was an argument made by Toffler: technology can give us individual choice and increase our degree of freedom in doing our thing. He illustrated the point by showing slides of a wide variety of construction proj ects, which he said were extreme examples of in- dividual choice. The speaker's message was that computer technology can allow the architect/engineer the self-expression he illustrated, even more fully than it has in the past. From the applause it was apparent that the audience agreed.
The Greeks called this the heuristic approach, where the pupil or machine proceeds along empirical lines, using rules of thumb to find solutions or answers. With this sort of computer application, we will find the computer to be a tool that will increase both productivity and creativity and will help us manage our business better. It can even help us produce better designs by giving us the time and resources to evaluate more options.