National Academies Press: OpenBook
« Previous: 'As Is' to 'Future Perfect'
Suggested Citation:"Expected Future." National Research Council and National Academy of Engineering. 2000. Design in the New Millennium: Advanced Engineering Environments: Phase 2. Washington, DC: The National Academies Press. doi: 10.17226/9876.
×

3

Expected Future

This chapter extends the process paradigm in Chapter 2 to predict how the mission/product design process will look in 7 and 15 years. The chapter discusses underutilized elements of available AEE technologies by examining the gaps between typical current practices and what is possible with current technology. Major areas of improvement are identified by examining technical advancements that would enhance progress toward realizing AEE capabilities. Based on current efforts to advance AEE technologies and processes, expectations for the next 7 and 15 years are described and compared to the IMTR and ISE 15-year visions. The chapter closes with findings and recommendations for improving the 15-year outlook.

UNDERUTILIZED AEE ELEMENTS

The gap between practices that are typical today and practices that take full advantage of currently available AEE capabilities is substantial. Some of these gaps can be identified by comparing the second column with the last column of Table 2-1 . More robust mission requirements could be established if quality engineering methods were integrated with traditional systems engineering methods. Using virtual prototypes (i.e., computer models and simulations) instead of physical prototypes for development testing is another capability that could benefit many industries, as some architectural firms have shown. But this practice is not widespread. Systematic methods of comparing performance data of alternative designs could also be used more.

Several factors have contributed to the underutilization of AEE technologies. First, finding and accessing analytical data for many products and processes is difficult, and some attributes can only be compared heuristically. Second, institutional program inertia and the natural desire to create new designs rather than extend someone else's old design (even though the latter is almost always the result in any case), have also impeded the use of new AEE technologies.

Based on information collected for this report and the experience of individual committee members, the committee estimates that one-half to three-quarters of new parts could reuse previous designs and still meet product and process requirements. Reuse reduces the need for new product geometries and increases the use of previously determined features, parametrics, and embedded rules. At the preliminary design level, the committee estimates that about one-quarter of product and process attributes could be evaluated analytically; this number could increase to one-half during detailed design, depending on the attribute being evaluated. Analytical evaluations would greatly reduce the percentage of attributes requiring experimental refinement through prototypes. The fraction of manufacturing processes that require experimental refinement could also be reduced, from about one-half to about one-quarter.

Simulations of manufacturing processes could be greatly improved by using accurate, scalable human figure representations (virtual humans), which can be created by products such as Jack from Engineering Animation or ERGO from Deneb Robotics. Virtual humans can be scaled by gender, age, and size to evaluate the ability of people with given strength and joint-angle constraints to perform proposed tasks. For example, EB used ERGO extensively to verify electronically the human factors aspect of the Virginia class SSN. This, in turn, significantly reduced the number of downstream design changes that had traditionally been necessary in the manufacturing, testing, and operational phases.

Taking full advantage of today's AEE capabilities will require commensurate changes in organizational culture and educational practices. For example, available methods could better integrate the curricula in a given discipline (e.g., fluid mechanics and thermodynamics) and enable the teaching of system-level engineering beginning in the freshman year of undergraduate studies. Also, the committee observed that most organizations do not use distributed collaborative work

Suggested Citation:"Expected Future." National Research Council and National Academy of Engineering. 2000. Design in the New Millennium: Advanced Engineering Environments: Phase 2. Washington, DC: The National Academies Press. doi: 10.17226/9876.
×

teams, and very few organizations that do use them have extended their use beyond the preliminary design stage. Using AEE technologies to facilitate effective teamwork would constitute a significant step forward in organizational culture.

Finding 3-1. Industry, government, and academia can significantly improve their engineering practices by using available AEE technologies.

MAJOR AREAS OF IMPROVEMENT

Table 3-1 links improvements in AEE capabilities to steps in the product and process design and development process. The process steps listed on the left side of the table are the same as the ones described in Table 2-1. Improvement areas and methods of improving AEE capabilities are listed across the top of the table. These areas are described here briefly and in some detail below. Table 3-1 is based on information collected by the committee during both phases of the study.

Design paradigm refers to the trend away from regarding parts as the central theme of design and toward regarding assembly and, ultimately, function as the starting point for design and the focal point for all design attributes. Attribute evaluation is the means of predicting design characteristics and performance. Hybrid evaluation refers to the typical process of using a combination of physics-based and empirical models to make analytical predictions. Physics-based evaluation refers to analytical modeling based on first principles, which is possible in areas that are so well understood that empirical data are not needed during the design and development of new products and processes. This capability will have to be expanded to a wide range of disciplines and applications to achieve the first-of-a-kind design capability desired by NASA and others. With statistical/stochastic evaluation, all aspects of uncertainty and variability can be incorporated into performance predictions. In this context, statistical implies that the elements of uncertainty and their probabilistic distributions are known, whereas stochastic implies that little or no information is available about the variables involved or their behavior.

The two methods associated with visualization of product and process designs are scientific visualization and the developing field of networked virtual environments. Methods associated with guidance for designing products and processes progress from heuristics (including rules of thumb) to multifunctional optimization (MFO) and expert systems as means of advising the designer.

The entries (H [high], M [medium], and L [low]) at the intersections of the rows and columns in Table 3-1 show the importance of each improvement to each process step. For example, achieving a true physics-based analytical capability would greatly improve (H) the prediction of product attributes throughout the overall process and, thereby, move the product design process from the typical today state toward a future perfect state. Improved heuristics would have a highly beneficial impact (H) on target setting for new concepts, a moderate impact (M) on the development of new concepts and high-level system trade-offs, and a minor effect (L) on trade-offs at the component level.

EXPECTATIONS FOR THE FUTURE

The following sections summarize the committee's expectations for product and process improvements over the next 7 and 15 years. These projections are based on interviews by committee members with key providers and practitioners of the methods listed and the assumption that ongoing research and development efforts continue as planned, with no significant reductions or additions. Predicting the future, however, is always risky. In some cases, predictions turn out to be overly optimistic, as with past expectations about the future of artificial intelligence or intelligent highways. In other cases, predictions have been greatly exceeded (see Box 3-1 ). Thus, the committee considers these expectations informative, but also speculative.

In the following discussion, analytical methods of predicting attributes are referred to as either general-purpose or special-purpose methods. General-purpose methods are applicable to a broad variety of problems and/or problems with a broad user base. Examples include kinematics, material processing (e.g., machining, forging, or casting), electrical circuit analysis, and finite element methods of calculating strength and stress. The committee anticipates strong demand and support for improvements in general-purpose methods. Special-purpose methods, such as combustion analyses and predictions of total life-cycle costs, have either a narrow user base or require significant customization to meet each user's needs. Improvements in special-purpose methods will probably require targeted investments by affected companies or user groups.

Conceptually, projects can be grouped into less complex projects (i.e., the development of evolutionary or relatively simple products or missions) and highly complex projects (i.e., the development of new, complex, or highly interactive products or missions). Less complex projects include changes to the design of a well characterized motor or pump that would not depart from the fundamental operating principles of existing designs, even if the redesign significantly alters the product's geometry, electrical characteristics, or performance. Relatively simple implies that the product is probably either (1) a component or subsystem or (2) a system whose internal and external interfaces are relatively simple and can be well defined. By contrast, highly complex projects include the development of automobiles, whose system interactions are very difficult to define, and first-of-a-kind design of an undersea robot or autonomous space-craft. Highly complex projects also include systems of systems, such as the Federal Aviation Administration' s National Airspace System, which involves interactions of advanced

Suggested Citation:"Expected Future." National Research Council and National Academy of Engineering. 2000. Design in the New Millennium: Advanced Engineering Environments: Phase 2. Washington, DC: The National Academies Press. doi: 10.17226/9876.
×

TABLE 3-1 Improvement Areas and Methods of Designing and Developing Products and Processes

 

Improvement Areas:

Design Paradigm

Attribute Evaluation

Visualization of Product and Process Designs

Guidance for Integrated Design of Products and Processes

Collaborative Support

Steps in the design and development of products and processes

Methods:

Parts-toFunction Paradigm

Hybrid Evaluation

Physics-Based Evaluation

Statistical/ Stochastic Evaluation

Scientific Visualization

Networked Virtual Environment

Heuristics

MFO

Expert Systems

 
  1. Mission Requirements Analysis/ Product System Strategy

    • high-level systems engineering analysis

    • requirements definition

M

M

 

M

  1. Product specification

    • product strategy

    • voice of the customer

    • environmental and other regulatory requirements

    • planned product specification

H

M

 

M

  1. Concept Development

    • target setting

    • brainstorming on alternatives

    • development of product and process concepts

H

M

 

M

M

H

H

M

H

H

H

H

H

  1. Preliminary Product and Process Design

    • high-level definition of product and process design

    • evaluation of product and process designs vs. targets

    • high-level system trade-offs

H

H

M

H

H

M

M

H

M

M

H

M

H

H

M

  1. Refinement and Verification of Detailed Product and Process Designs

    • development of designs for components, subsystems, and manufacturing processes

    • geometry creation

    • prediction and evaluation of all product and process attributes

    • tracking and trade-offs of subsystems and components

H

H

H

H

H

M

 

M

L

H

M

H

M

  1. System Prototype Development

    • experimental refinement of product attributes that do not meet tragets

H

H

H

H

  1. Prepare for Production

    • experimental refinement of process attributes that do not meet requirements

H

H

H

H

  1. Production, Testing, Certification, and Delivery

  1. Operation, Support, Decommissioning, and Disposal

 

M

H

NOTE: Entries indicate importance (high [H], medium [M], or low [L]) of improvement areas in moving process elements from the “typical today” state toward the “future perfect” state.

Suggested Citation:"Expected Future." National Research Council and National Academy of Engineering. 2000. Design in the New Millennium: Advanced Engineering Environments: Phase 2. Washington, DC: The National Academies Press. doi: 10.17226/9876.
×

BOX 3-1

Unpredictable Acceleration of Computing Technology

In 1989, supercomputers cost from $1 to $20 million, and the best commercially available machines, such as the Cray Y-MP, could perform about 1 billion (giga) floating-point operations per second (i.e., 1 gflop) using 8 interconnected central processing units. A National Research Council report published in 1989 predicted that in 1999 commercially available supercomputers would demonstrate 100 gflops using from 32 to 256 interconnected central processing units (NRC, 1989).

Improvements in general-purpose supercomputers have exceeded these predictions. By November 1999, the 50 most capable supercomputers in the world, which were built by six different manufacturers, could all perform more than 200 gflops. The top 15 could perform more than 500 gflops, and the most capable machine (the Department of Energy 's ASCI Red supercomputer, manufactured by Intel), could perform 2,400 gflops using 9,632 linked processors (TOP500, 1999). Commercially available machines, such as the Cray SV1, can be configured to provide up to 1,000 gflops (Cray, 2000a).

The 1989 report also described the technical difficulty of reducing the size of supercomputers. The Cray X-MP supercomputer, which was introduced in 1982 with a capability of 0.5 gflops, occupied about 100 square feet—and was much smaller than earlier generations of supercomputers (Cray, 2000a, 2000b). The 1989 report predicted that supercomputers smaller than the Cray X-MP might be developed, but cooling would be a “fantastic problem.” A “suitcase-sized supercomputer may dissipate a couple thousand watts of power,” and there “would be an instant meltdown in case of a cooling malfunction” (NRC, 1989). Today, the most capable supercomputers are still massive devices with large cooling requirements. At the lower end, however, children's toys that feature impressive capabilities come in small packages with no special cooling requirements. Unlike general-purpose supercomputers, the special-purpose PlayStation® 2 video game console conducts single precision computations. Nonetheless, the PlayStation 2 is capable of 6.2 gflops, takes up just 0.25 square feet, consumes 15 watts of power, and sells for about $300 (Sony, 1999).

technologies for air traffic control with advanced technologies for aircraft.

Product Technologies

This section describes 7-year and 15-year expectations for product technologies related to design paradigms, attribute evaluations, design/process visualizations, guidance for designing products and processes, collaboration support, and education and training.

Design Paradigm

Seven-Year Expectations. The committee predicts that approximately half of CAD work will be done on Web-based systems communicating through open standards. Exchange of parts will be reliable up through “dumb” solids (which contain no information about features, parametrics, or design intent). It will be possible to design 75 to 90 percent of simple parts from generic, reusable design templates (wizards). For these parts, the traditional laying of lines or creation of geometry from scratch on CAD systems will be eliminated. For example, rather than creating the geometry of a structural rib one feature at a time, the user will specify a minimum set of unique dimensions, and a generic template will automatically create the rest of the product specification. Design tables will be used for part regeneration, and design templates for processing. Automated part creation from functional descriptions will be emerging, but only for simple, common parts and assemblies. Graphical interfaces will replace dialogue boxes. Haptic feedback will be used to display the results of analyses to experienced users for some tasks, such as evaluating fit and assembly processes.

Fifteen-Year Expectations. Monolithic CAD systems will no longer exist. Data exchange will be reliable, even for “featured” solids, which include information on function and design intent. Function-based design will be common for simple parts and will be emerging for complex parts and assemblies. Some forms of unique parts will be created using functional descriptions. Geometry optimization will be common at the part level. Concept design with detailed, automatically generated models of complex systems will be possible. Rules-based algorithms embedded in the design generator will ensure that the specified part is satisfactory for all design conditions (e.g., thermal pressure and impact loading). This will reduce the difficulty of analysis when parts are at different levels of design definition (the mixed-phase design problem). Virtual reality interfaces will be common for aspects of functional design, and they will be integrated with real-time analysis of functions, at least for general-purpose methods.

Suggested Citation:"Expected Future." National Research Council and National Academy of Engineering. 2000. Design in the New Millennium: Advanced Engineering Environments: Phase 2. Washington, DC: The National Academies Press. doi: 10.17226/9876.
×
Attribute Evaluation

Seven-Year Expectations. The speed of general-purpose methods will be 10 to 30 times faster than today, significantly automated, and easier to use. Evaluations will be more accurate and provide high levels of product definition. These methods will be at least half Web-based, with reliable data exchange and interoperability of related analytical streams. Most verification testing will have been eliminated for these methods. Back-end analyses, such as processing and assembly, will benefit from industry's emphasis on rapid service to the consumer. Special-purpose methods will advance unevenly, depending on the efforts of individual companies or user groups. The utility of design analyses with a low level of detail definition will show limited progress. Improvements in physics-based analyses will also be limited for most attributes. Dynamic analysis methods are likely to show the most progress.

Fifteen-Year Expectations. The growth of rules-based automated methods of designing parts based on functional descriptions will lead to highly automated general-purpose design methods for most applications. These methods will account for product and process variabilities, including manufacturing variability, and will provide better predictions of performance than prototype testing. Special-purpose methods will have advanced similarly, especially in areas of special interest to individual companies or user groups that have been active in supporting the development of advanced design methods. Mixed-phase analysis will be automated, enabling seamless integration of highly detailed parts with concept-level parts, and general-purpose methods will be as useful for concept development as for detailed design. Also, some of today's special-purpose methods, such as vibration or impact analysis, may be so easy to perform 15 years hence that they will be considered general-purpose methods.

Design/Process Visualization

Seven-Year Expectations. Multimodal systems will have graphics that operate at more than 300 million polygons per second, with some spatial audio interfaces1 and haptic interfaces (on tool handles). Computer-generated agents will be available to guide human designers, but the agents will not operate autonomously. Computers will be 200 times faster than today.

Fifteen-Year Expectations. Graphic systems will have high speed and quality (one billion polygons per second, 60 frames per second, images indistinguishable from reality). Interfaces with full spatial-audio effects, full-hand haptics, and olfactory displays will also be available. With computergenerated autonomy, agents will do much of the design, guided by high-level specifications. Computers will be 30,000 times faster than today.

Guidance for Designing Products and Processes

Seven-Year Expectations. Easing constraints will be a tempting and popular alternative to true optimization analysis. MFO (multifunctional optimization) will be incorporated into general-purpose methods, including analyses of control, stress, vibration, and assembly, but will not be widely used for linking different general-purpose methods. MFO will not be incorporated into special-purpose methods, cost, weight, or complex material processing. Enterprise resource management will have moderate optimization capability but will not include life-cycle cost. Recursive design will be the dominant method, with faster iterations and more heuristic guidance than today. Expert systems will be used with general-purpose methods of raising the performance of less skilled users to a predictable, consistent level, but they will not replace experienced designers or engineers when creativity is required. Expert systems will play a role in moving from geometry-based to function-based design but will be confined to repetitive design tasks. Heuristics will be extensively used in conjunction with general-purpose methods and will provide substantial benefits in the concept development stage of design.

Fifteen-Year Expectations. MFO will be available across the more common general-purpose methods. For example, producibility may be combined with stress and control performance. Special-purpose methods will be included in MFO for selected situations. Enterprise resource management will be highly developed, with moderate life-cycle cost capability. Heuristics will be so well integrated with design and MFO packages that heuristic methods will be virtually indistinguishable from rigorous optimization. Nonrecursive design will be possible for less complex projects and generalpurpose methods, but recursive design with design guidance will still be required for more complex problems and specialpurpose methods. With expert systems, unique parts and simple assemblies will be created automatically based on functional descriptions.

Collaboration Support

Seven-Year Expectations. Video walls and smart meeting rooms will be widely available. Grid-like networking infrastructures connecting designated user groups with secure high bandwidth will enable routine exchanges of large data files and group interactions. Shareable distributed sets of extant data, objects, and tools will be readily available. Event

1  

When using headphones, spatial-audio interfaces make sounds appear to come from particular points in space outside the listener's head. With ordinary stereo headphones, sound generally appears to originate somewhere on a line between the listener's ears. Spatial-audio effects increase the sense of being immersed in a virtual environment, especially when sounds are generated at more than one location in virtual space.

Suggested Citation:"Expected Future." National Research Council and National Academy of Engineering. 2000. Design in the New Millennium: Advanced Engineering Environments: Phase 2. Washington, DC: The National Academies Press. doi: 10.17226/9876.
×

capture will be commonly used to provide access to knowledge bases with group memories. A large selection of interoperable tools and techniques will be available for education and training of individuals involved in new collaborations or projects.

Fifteen-Year Expectations. Immersive telepresence and roomless meetings will become practical. Large-scale collaborations with broad scopes will be available through ubiquitous networking technologies, not just specially equipped user locations. Heterogeneous interoperability will enable the transparent sharing of data, objects, and tools among collaborators. Event guidance and smart capture to facilitate collaborations will create an accessible knowledge base and group memory. Automatically adaptive, reconfigurable tools will be available for teaching individuals involved in new collaborations and projects.

Education and Training

Seven-Year Expectations. A large number of Web-based degree programs will be available. Some AEE links will be established with NASA and industry. Design and system engineering courses will start in the freshman year and will be partially integrated into the curriculum.

Fifteen-Year Expectations. Fundamental courses will be linked with AEEs. Team teaching in engineering, mathematics, and science will be common. Case studies of projects using AEEs will be used as examples. Cross-disciplinary design and system-engineering courses will be fully integrated into the curriculum.

Processes
Seven-Year Expectations

For less complex projects, the concept development and preliminary design phases will collapse into a single process step, especially for general-purpose attributes. The time required for these steps will be reduced by as much as one-half. The most difficult aspect of achieving this goal will be improving the accuracy of predictive methods of refining detailed designs. Improvements can be expected in selected areas with the development of heuristics, physics-based analysis, and expert systems to support early design.

For complex projects, preliminary design and detailed design will still be distinct tasks. The level of available design detail will vary at different points in the design process. Therefore, the accuracy of performance predictions will also vary. For less complex projects and general-purpose methods, predictions will be very accurate during the later design phases and manufacturing. This increase in accuracy will almost completely eliminate the need for experimental refinement and physical prototypes (but only for less complex projects and general-purpose methods). The time for steps 5 through 7 of the design and development process (detailed design through preparation for production) will decrease by as much as a third.

Highly complex projects will benefit from part reusability and limited, function-based geometry creation. Special-purpose methods could achieve many of the same improvements as general-purpose methods if the demand and corresponding investments are sufficient. However, process steps are unlikely to collapse into a single step, as they will for less complex projects.

Semiautonomous, computerized, digital design advisers will be created for specific purposes. Limitations will include insufficient physics-based analysis capabilities, a limited ability to break down complex systems, and difficulties in dealing with nonlinear and subjective phenomena. Reductions in total process time will be limited to about 20 percent, with similar reductions in cost. Although physical prototypes will still be necessary for highly complex projects and special-purpose methods, the speed of iteration and improved visualization should reduce their number by at least 20 to 30 percent, with similar reductions in the amount of rework associated with experimental refinement during prototype development and preparation for production (steps 6 and 7).

Also during the next seven years, improved collaboration processes will enable work in all design process stages to be distributed more broadly across teams and organizations. However, participation in a distributed design collaboration will still depend heavily on physical location, and effective anywhere-to-anywhere collaboration will continue to be a future goal. Improved capabilities for sharing data, objects, and tools will be evident throughout distributed work on less complex projects, particularly projects that rely on general-purpose methods for which open standards will have emerged. Improvements in the distribution and sharing of digital designs will mean that human efforts to integrate and synthesize work outputs will occur largely at the system and subsystem levels of design. At the part and component level, integration and synthesis will be mostly automated.

Fifteen-Year Expectations

For less complex projects and general-purpose methods, steps 3, 4, and 5 (concept development, preliminary design, and detailed design) will be collapsed into a single process, which will be one-quarter to one-half as long as today's processes. This reduction will be attributable to a combination of computerized design advisers, rapid iterations of design and analysis, physics-based analyses, MFO, and advanced visualization, all of which will reduce the number of design iterations, although the process will still be recursive for all but the simplest products. Experienced staff will be required to manage the design flow for all but the most repetitive designs because the selection of appropriate

Suggested Citation:"Expected Future." National Research Council and National Academy of Engineering. 2000. Design in the New Millennium: Advanced Engineering Environments: Phase 2. Washington, DC: The National Academies Press. doi: 10.17226/9876.
×

methods from available tools will have to be tailored to the individual problem. Many mundane functions, such as geometric design, specification of tolerances, assessment of processing feasibility, and prediction of general-purpose attributes, will be completely automated. Physical tests will be unnecessary to validate product attributes and processes, except where special-purpose methods are needed. Physical testing of special-purpose prototypes will only be needed when special-purpose methods are not as capable as general-purpose methods.

For general-purpose methods applied to highly complex projects, it will be possible to collapse steps 3 and 4 (concept development and preliminary design) into a single step for all but truly unique products. Tools will be augmented with function-based design and concept design methodologies that include detailed part definitions to make this collapse possible. Little or no hardware evaluation will be necessary during concept development or preliminary design. In step 5 (detailed design), evolving design maturity will produce changes in the design that will require redefinition, reevaluation, and trade-offs. These changes will probably not be automated for new or complex systems. Augmentation of experienced personnel with computerized advisors will reduce the time and improve the accuracy of detailed design. For prototype development and preparation for production (steps 6 and 7), physical tests will be limited to the evaluation and refinement of system interactions and will, therefore, be reduced by at least half.

For special-purpose methods applied to highly complex projects, progress will depend entirely on the degree to which special-purpose methods have been developed. Concept development and preliminary design (steps 3 and 4) will probably be partially combined, depending on the maturity of the special-purpose methods. In the absence of highly developed special-purpose methods, the ability to conduct concept design with detailed parts and the assistance of computerized advisors will be critical to increasing the speed and accuracy of the early design stages. Depending on the maturity of special-purpose methods, perhaps one-third of prototypes used during preliminary design will be eliminated, with a similar reduction in testing during experimental refinement of prototypes and production methods in steps 6 and 7. These reductions will be less pronounced for projects that require nonphysics-based analyses and for projects with complex system interactions for which decomposition techniques are not available.

Advances in computer systems are likely to be so great that users will not be limited by the capabilities of computing hardware. Voice interfaces will be common, and distributed computers and wireless interfaces will be embedded in a wide variety of products. Immersive environments will be commonly available for entertainment and work. The demand for skilled technicians to manage AEE systems will increase, and the educational system will adjust curricula to meet this need.

Finally, constraints on the physical location of personnel engaged in complex collaborations will have been largely eliminated for nearly all stages of the design process (with manufacturing likely to be the most notable exception). Further improvements in the sharing of standards-based and heterogeneous data, objects, and tools will facilitate distributed work on all types of projects, whether they rely on general-purpose or special-purpose methods. For all less complex projects, and for some highly complex projects, moreover, human work in integration and synthesis of information will be confined to the system and metasystem levels of design. All lower levels of integration and synthesis will be handled by autonomous AEE technologies.

COMPARISON OF EXPECTATIONS TO THE 15-YEAR VISIONS

This section compares the committee's expectations for future technologies and processes (described above) to the 15-year visions defined by the ISE and IMTR initiatives (summarized in Table 2-1 ). Based on current and projected efforts by industry and government, the committee does not believe that all elements of the 15-year visions are likely to be achieved. Because of political pressure to constrain discretionary spending by the federal government, obtaining support for major new initiatives is very difficult. As a result, advancing the state of the art of AEE technologies and systems rapidly enough to achieve the 15-year visions of the IMTR or ISE initiatives is highly unlikely. For example, physics-based, first-principles analyses will not be able to predict the reliability and performance of first-of-a-kind products and missions with enough certainty or accuracy to preclude testing. Shortcomings will also persist in modeling interactions in highly complex systems.

Life-cycle optimization will be limited to general-purpose and selected special-purpose methods and will not accurately predict cost or risk. Enterprise resource management will predict comprehensive life-cycle costs with limited accuracy.

The committee believes that attribute prediction will be almost completely analytical for general-purpose and selected special-purpose methods, but physical testing will still be required to refine some attributes, especially for complex system interactions and first-of-a-kind missions or products.

The time needed to complete some steps of the design process will be reduced by more than the factor of 200 predicted in the ISE 2015 vision. However, the time to complete the overall design and manufacturing cycle will not reflect a similar reduction because some process elements will have made less progress. For complex, first-of-a-kind missions, cycle times will be reduced by a factor of 10 at the most, and actual gains are likely to be somewhat less.

In 15 years, the interchange of geometry and geometry-related data will be seamless, but the same cannot be expected for generalized product data. For example, a truly

Suggested Citation:"Expected Future." National Research Council and National Academy of Engineering. 2000. Design in the New Millennium: Advanced Engineering Environments: Phase 2. Washington, DC: The National Academies Press. doi: 10.17226/9876.
×

integrated approach to system and mission design must include electrical systems and software. Most AEE activities, however, seem to be focused on structural design issues, with relatively little effort devoted to electrical system design or software development. Software engineering is particularly important because mechanical design methods are very different from software engineering methods. Also, software-intensive systems, such as avionics, account for an increasing percentage of the total cost of complex systems, especially in the aerospace sector. Validating the performance and reliability of complex software is difficult, expensive, and time consuming (NRC, 1999). As a result, validation often misses software problems that subsequently cause product or mission delays, accidents, or failures.

The committee believes that the entertainment industry may prove to be an important source of advanced technologies in areas such as immersive environments, which are likely to become widely available as part of entertainment systems (NRC, 1997). Large corporate and government entities may then adopt these systems for many of their own applications.

RECOMMENDATIONS FOR IMPROVING THE 15-YEAR OUTLOOK

The following recommendations will enhance the ability of research and development by government, industry, and academia to achieve the 15-year visions described in Table 2-1 .

Finding 3-2. Government agencies (such as NASA) acting alone will not be able to achieve the 15-year visions of the Integrated Manufacturing Technology Roadmapping Initiative or NASA's Intelligent Synthesis Environment (ISE) Initiative. Similarly, actions such as linking the Army's Simulation and Modeling for Acquisition, Requirements, and Training (SMART) Program with the ISE initiative, although a step in the right direction, are unlikely to achieve these visions unless the partnerships are expanded to include other government, industry, and university programs with additional resources.

Recommendation 3-1. Until additional funding is made available to invigorate the recommended national partnership for AEEs, government agencies should make the most efficient use of the limited resources now available for the development of AEE technologies and systems by focusing their efforts on the types of missions and products that can benefit most from AEEs and on functional areas that are lagging behind the rapid advances being made in other areas, such as the speed, size, and cost of computer hardware. Research and development should be focused on the following areas:

  • comprehensive processes for project design and development that integrate the design of mechanical systems with electrical system design and software development

  • general-purpose methods of analyzing cost and determining the effects of risk and uncertainty to reduce the need for project-specific cost and risk analysis tools

  • physics-based analysis of mission-specific phenomena if first-of-a-kind missions are a high priority

Recommendation 3-2. The federal government should carefully assess how the limited resources available for AEE research and development (such as the Intelligent Synthesis Environment Initiative) are allocated between the development of (1) general-purpose research with broad application (to improve engineering processes throughout the United States), and (2) engineering processes of particular relevance to agency missions (including activities by industry and academia). The guidelines in Recommendations 9 through 12 of Advanced Engineering Environments—Achieving the Vision (the Phase 1 report), which call for action by NASA and other federal agencies, remain relevant.2

REFERENCES

Cray . 2000a . Cray SV1 Scalability . Available on line at:http://www.tera.com/products/systems/craysv1/scalability.html June 2, 2000 .

Cray . 2000b . A History of Cray . Available on line at: http://www.cray.com/company/history.html July 5, 2000

NRC (National Research Council) . 1989 . Supercomputers: Directions in Technology and Applications . Computer Science and Technology Board . Washington, D.C. : National Academy Press . Available on line at: http://www.nap.edu/catalog/1405.html May 3, 2000 .

NRC . 1997 . Modeling and Simulation: Linking Entertainment and Defense . Computer Science Telecommunications Board . Washington, D.C. : National Academy Press . Available on line at: http://books.nap.edu/catalog/5830.html June 2, 2000 .

NRC . 1999 . Trust in Cyberspace.Computer Science Telecommunications Board . Washington, D.C. : National Academy Press . Available on line at: http://books.nap.edu/catalog/6161.html June 2, 2000 .

Sony . 1999 . Sony Computer Entertainment Announces World's Fastest 128 Bit CPU (Central Processing Unit) “Emotion Engine” for the Next Generation PlayStation® . Tokyo, Japan : Sony Computer Entertainment . Available on line at: http://www.playstation.com/news/press_example.asp?ReleaseID=9521 April 20, 2000 .

TOP500 . 1999 . Top 500 Supercomputer Sites, November 1999 . Available on line at: http://www.top500.org/lists/TOP500List.php3?Y=1999&M=11 April 20, 2000 .

2  

See Appendix B of this report, which lists the recommendations from the Phase 1 report.

Suggested Citation:"Expected Future." National Research Council and National Academy of Engineering. 2000. Design in the New Millennium: Advanced Engineering Environments: Phase 2. Washington, DC: The National Academies Press. doi: 10.17226/9876.
×
Page 21
Suggested Citation:"Expected Future." National Research Council and National Academy of Engineering. 2000. Design in the New Millennium: Advanced Engineering Environments: Phase 2. Washington, DC: The National Academies Press. doi: 10.17226/9876.
×
Page 22
Suggested Citation:"Expected Future." National Research Council and National Academy of Engineering. 2000. Design in the New Millennium: Advanced Engineering Environments: Phase 2. Washington, DC: The National Academies Press. doi: 10.17226/9876.
×
Page 23
Suggested Citation:"Expected Future." National Research Council and National Academy of Engineering. 2000. Design in the New Millennium: Advanced Engineering Environments: Phase 2. Washington, DC: The National Academies Press. doi: 10.17226/9876.
×
Page 24
Suggested Citation:"Expected Future." National Research Council and National Academy of Engineering. 2000. Design in the New Millennium: Advanced Engineering Environments: Phase 2. Washington, DC: The National Academies Press. doi: 10.17226/9876.
×
Page 25
Suggested Citation:"Expected Future." National Research Council and National Academy of Engineering. 2000. Design in the New Millennium: Advanced Engineering Environments: Phase 2. Washington, DC: The National Academies Press. doi: 10.17226/9876.
×
Page 26
Suggested Citation:"Expected Future." National Research Council and National Academy of Engineering. 2000. Design in the New Millennium: Advanced Engineering Environments: Phase 2. Washington, DC: The National Academies Press. doi: 10.17226/9876.
×
Page 27
Suggested Citation:"Expected Future." National Research Council and National Academy of Engineering. 2000. Design in the New Millennium: Advanced Engineering Environments: Phase 2. Washington, DC: The National Academies Press. doi: 10.17226/9876.
×
Page 28
Next: Overcoming Barriers »
Design in the New Millennium: Advanced Engineering Environments: Phase 2 Get This Book
×
Buy Paperback | $47.00 Buy Ebook | $37.99
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

America is changing. Many of the most noticeable changes in day-to-day life are associated with the advancing capabilities of computer systems, the growing variety of tasks they can accomplish, and the accelerating rate of change. Advanced engineering environments (AEEs) combine advanced, networked computer systems with advanced modeling and simulation technologies. When more fully developed, AEEs will enable teams of researchers, technologists, designers, manufacturers, suppliers, customers, and other users scattered across a continent or the globe to develop new products and carry out new missions with unprecedented effectiveness. Business as usual, however, will not achieve this vision. Government, industry, and academic organizations need to make the organizational and process changes that will enable their staffs to use current and future AEE technologies and systems.

Design in the New Millennium: Advanced Engineering Environments: Phase 2 is the second part of a two-part study of advanced engineering environments. The Phase 1 report, issued in 1999, identified steps the federal government, industry, and academia could take in the near term to enhance the development of AEE technologies and systems with broad application in the U.S. engineering enterprise. Design in the New Millennium focuses on the long-term potential of AEE technologies and systems over the next 15 years. This report calls on government, industry, and academia to make major changes to current organizational cultures and practices to achieve a long-term vision that goes far beyond what current capabilities allow.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    Switch between the Original Pages, where you can read the report as it appeared in print, and Text Pages for the web version, where you can highlight and search the text.

    « Back Next »
  6. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  7. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  8. ×

    View our suggested citation for this chapter.

    « Back Next »
  9. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!