3
Requirements and Alternatives
Introduction
As industries and governments around the world attempt to become more competitive in the world marketplace, affordability, defined as the ratio of system effectiveness to the cost of achieving this effectiveness, has become a vital criterion or figure of merit for determining success. Organizations are also trying to better understand the performance of new systems early in the design process, before a substantial fraction of program costs has been committed to a particular product design or mission concept. Rather than simply updating existing tools, the committee believes that many organizations must fundamentally change their engineering culture to take advantage of breakthroughs in advanced computing, human-machine interactions, virtual reality, computational intelligence, and knowledge-based engineering. The committee believes that achieving this goal and applying AEE technologies and systems across a wide range of government and industry activities will only be possible if AEE R&D is integrated into a coherent vision of future science and engineering. The remainder of this chapter discusses top-level AEE objectives, benefits, and requirements; components-level requirements; and alternative approaches for achieving the objectives.
Top-Level Objectives, Benefits, and Requirements
The committee identified the following top-level objectives that AEEs should satisfy:
- Enable complex new systems, products, and missions.
- Greatly reduce product development cycle time and costs.
In addition, AEE technology and system developers should devise a comprehensive, multifaceted implementation process that meets the following objectives:
- Lower technical, cultural, and educational barriers.
- Apply AEEs broadly across U.S. government, industry, and academia.
As described in the following sections, the top-level benefits and requirements of AEEs are closely linked to these key objectives.
System Objectives
Enable Complex New Systems, Products, and Missions
To remain competitive in the marketplace, manufacturing industries must continually develop products that offer new capabilities, improved quality, and/or lower costs. Although government agencies do not face the same competitive market forces as industry, technology-intensive agencies, such as the Department of Defense, the Department of Energy, the Federal Aviation Administration, and NASA face a similar challenge of developing new systems to maximize organizational effectiveness and accomplish agency missions.
Traditional processes for designing, developing, manufacturing, and operating the systems needed to satisfy the complex missions of industry and government are becoming increasingly unsupportable because of cost, schedule, and personnel requirements. The complexity of products and processes has rapidly increased, and the amount and heterogeneity of data required to define, manufacture, and maintain these products have increased dramatically. Global design, manufacture, and maintenance means that this large mass of data must be accessible and movable over long distances and at high speed. The risk of error, of course, increases with more complex systems, and the cost and time implications of mistakes are magnified by the speed and scope of product deployment. AEEs have the potential to improve accuracy and efficiency of engineering processes throughout the life cycle of products and missions. For example, AEEs would
enable industry to develop advanced systems more quickly with fewer personnel and at lower cost. Similarly, AEEs should enable government agencies and industry to accomplish missions and develop products that are not feasible using current processes.
Greatly Reduce Product Development Cycle Time and Costs
When complex new systems or products are developed using traditional methods, the bulk of a program's life-cycle costs are set by decisions made very early in the development cycle (the definition phase). Correcting errors made during this phase often involves costly and time-consuming design changes later in the process. For example, the initial product design may specify the basic system structure and a functional allocation for individual subsystems that commit the program to a particular direction. Later on, after subsystems are built and tested or, in the worst case, after the product is integrated, developers may realize that a particular subsystem or the integrated product cannot meet performance specifications. Corrective action may ripple throughout a number of subsystems, requiring extensive rework. Even if the individual changes are small, the net effect can be substantial.
In the commercial world, reducing product development cycle time helps manufacturers increase their market share by enabling them to create new and better products faster than the competition. In the government sector, reduced product development cycle time helps agencies complete projects sooner, thereby reducing costs and improving services or achieving mission objectives more quickly and freeing personnel and fiscal resources to move on to the next task.
One way to reduce product development cycle time and costs is to develop AEEs that allow designers to determine quickly and accurately how proposed designs will affect the performance of new systems and subsystems and how the changes in performance will affect the prospects for mission success. Integrating tools from all aspects of the mission life cycle would allow mission planners and system designers to do the following:
- Seamlessly integrate diverse, discipline-specific design and simulation tools that model and analyze components, subsystems, systems, and related processes from concept development to end-of-life disposal.
- Perform trade-off study sensitivity analyses early in the design process.
- Perform trade-off studies to assess appropriate parameters for each phase of the total life cycle.
- Reduce operational costs by ensuring that operational requirements are addressed early in the design process and in all trade-off studies.
- Lower manufacturing costs by
- making design for manufacturability an inherent part of the concept development and design process
- reducing the need to build physical test models of new designs
- reducing the need for design changes late in the cycle
The sophistication of simulations should be adjusted based on several factors, such as individual product or mission value and the size of the product run or number of missions planned. If the product is expensive enough to justify the cost, comprehensive simulations will be a worthwhile investment, even for small product runs. Launch vehicles and nuclear submarines, for example, certainly justify the cost of extensive simulations, even though they are not produced in large numbers.
Objectives of the Implementation Process
To realize the benefits of AEEs, the development of AEE technologies and systems must be coordinated with the development of a comprehensive, multifaceted implementation process. This process will have to include distinct elements tailored to the characteristics and issues associated with different AEE technologies and system components. The following two sections describe the top-level objectives of the implementation process.
Lower Cultural, Technical, and Educational Barriers
Because of barriers to change and innovation, old systems and processes are often retained long after more effective alternatives become available. These barriers may involve technical, cultural, educational, and/or economic factors. Technical barriers often involve the incompatibility of new systems with legacy systems, especially if an organization has a large investment in existing systems and infrastructure. For example, most of Boeing's existing in-house applications have been encoded using IBM's operating system (AIX), and it would be prohibitively expensive to make them compatible with other operating systems. This severely limits Boeing's ability to use computer hardware from other vendors.
To take advantage of changes in business practices and technology, organizations should develop systematic methods of encouraging innovation. Lowering barriers is an essential precursor to achieving the other benefits offered by AEEs. (See Chapter 4 for a discussion of specific barriers.)
Apply AEEs Broadly across U.S. Government, Industry, and Academia
Just as the development of AEEs should include an implementation process to maximize the benefits of AEEs for a
particular organization, development efforts should also be consistent with the broader objective of applying AEEs throughout U.S. government, industry, and academia. In other words, approaches to AEEs should not restrict their applicability to a small number of settings.
Consider the significant differences between the automotive and aerospace industries. An automobile manufacturer may produce several hundred thousand automobiles of a popular model, and each vehicle may have a value of $20,000 to $50,000. An aerospace company, on the other hand, may produce just a few satellites, a few dozen launch vehicles, or a few hundred airplanes of a given model, and the value of each vehicle may be well in excess of $100 million. AEE technologies and systems must be flexible so they can be tailored to improve product quality and reduce costs throughout both industries.
The widespread use of AEEs will also maximize their value to individual organizations. Complex products and missions typically are implemented by partnerships of many different organizations, and the AEEs adopted by one organization will be more useful if its partners use compatible AEEs.
To achieve widespread use of AEEs, industry and academia must establish appropriate training and educational programs for current and future workers. In fact, AEEs themselves can be used as effective training tools. For example, the astronaut training methods traditionally used by NASA include neutral-buoyancy training, part-task training, and full-scale simulators; make limited use of virtual environment technology; and are very expensive. Neutral-buoyancy trainers, which require safety divers and equipment operators, may have a staff-to-student ratio as high as 40:1. Also, as NASA has learned the hard way, differences in the physics of vacuum and water reduces training fidelity, and methods practiced in buoyancy tanks sometimes don't work in space. Furthermore, the committee believes that a stressful training environment is an important element of training for stressful real-world situations. The space station raises additional training issues because the number of planned operations is so extensive that traditional training methods will be impractical.
AEEs that provide shared virtual environments can address these issues. In addition, AEEs would enable team training even when all members of the team are not at the same location (i.e., reducing the need to station astronauts at Johnson Space Center for months or years of on-site training prior to each mission). For this application, the fidelity with which remote participants are depicted will be especially important, although the required level of fidelity is lower for participants who know each other. Virtual environments can include ''avatars," virtual participants whose actions and reactions are controlled by the training system in response to actions by the human participants.
Table 3-1 AEE System Components and Characteristics
|
Computation, Modeling, and Software |
|
• multidisciplinary analysis and optimization |
|
• interoperability of tools, data, and models |
|
• system analysis and synthesis |
|
• collaborative, distributed systems |
|
• software structures that can be easily reconfigured |
|
• deterministic and nondeterministic simulation methods |
|
Human-Centered Computing |
|
• human-adaptive interfaces |
|
• virtual environments |
|
• immersive systems |
|
• telepresence |
|
• intelligence augmentation |
|
Hardware and Networks |
|
• ultrafast computing systems |
|
• large high-speed storage devices |
|
• high-speed and intelligent networks |
Component-Level Requirements
The AEE components and characteristics identified by the committee are listed in Table 3-1 and discussed in the sections that follow.
Computation, Modeling, and Software
The products and processes designed using AEEs will typically be large and complex, so AEEs must be capable of rapidly synthesizing and analyzing the performance of large combined hardware and software systems. Multidisciplinary tools will be required for analyzing attributes and subsystems, such as aerodynamics, thermal management, mission design, life-cycle costs, manufacturing, maintenance, and risk management. Traditionally, trade-offs among these attributes for shared resources (e.g., power or weight) have been made iteratively; to support rapid analysis and achieve "best" designs, AEEs will require a multifunctional optimization capability.
The models and simulations incorporated into AEE systems must effectively address the uncertainty and risk associated with the development of new systems. One way to reduce uncertainty is to validate simulations and models. Validation generally involves physical tests to verify that the performance of simulations and models is consistent with reality. In some situations, however, it is difficult to create high-fidelity models, and physical testing cannot remove all uncertainty and risk. Thus, AEEs must include methods for assessing the impact of residual uncertainty and risk. Deterministic simulation methods must be augmented or, perhaps, replaced by nondeterministic methods to account explicitly for uncertainties. For example, Monte Carlo techniques
might be appropriate for analyzing models that are very computationally intense. Accounting for uncertainties will be particularly important in analyzing multiyear missions.
Computation, modeling, and software systems will also have to facilitate interactions among multiple decision makers with different values. For example, the merit of proposed new products and/or missions may be perceived very differently by researchers, technologists, designers, manufacturers, suppliers, and customers. Multidisciplinary analysis and optimization tools should include flexible input and output capabilities to accept inputs and display results in terms that make sense to each of these constituencies.
From an operational standpoint, AEEs must provide an infrastructure for distributed collaboration in which the geographical location of team members is completely transparent to the design process. The organizational location must also be transparent, which implies a high degree of interoperability among analysis tools, data, and models. The system should be fully associative, so that data are entered only once and are then accessible everywhere their use is authorized.
Because of the evolving nature of AEE capabilities, software structures must be reconfigurable quickly and economically. Ongoing maintenance needs and future system changes should be anticipated in the design of the original system and in the design of hardware and software upgrades.
Human-Centered Computing
The purpose of human-centered computing is to increase the communications bandwidth among users and between each user and the AEE system. Virtual reality and/or immersive systems provide users with visual, audio, and, in some cases, haptic feedback,1 all of which increase the sense of presence or perceived reality of a simulation. With sufficient computing power, users can collaborate on redesigning in real time and experience the physical and performance results of their work through virtual interfaces. A less well known but equally powerful approach uses human-adaptive interfaces, which are adjusted by the system to suit the needs, skills, or mental state of the user.
Human-centered computing also includes intelligence augmentation, which uses software and hardware agents, decision-support software, knowledge-based systems, and autonomous learning systems to provide real-time design guidance. For example, AEEs should support decision methodologies compatible with multiple users who have different values and decision criteria.
Hardware and Networks
Ultrafast computing systems will be needed to support the real-time analyses and interfaces described above. These systems will probably involve massively parallel processors and, ultimately, a distributed heterogeneous computing capability to maximize computing power economically. Because of the size and complexity of the products and processes likely to be designed with AEEs, large data sets will have to be retrieved, modified, transported, analyzed, displayed, and stored. This will require large high-speed storage devices and high communications bandwidths.
Component-Level Requirements
To investigate component-level requirements, the committee conducted a survey of personnel associated with the following companies and projects: Caterpillar (construction equipment), Ford Motor Company (automobiles), Simmetrix (analysis software), the X-38 Program at NASA Johnson Space Center (prototype spacecraft), Boeing Electromagnetics (analysis of electromagnetic interference for large systems), Boeing CAD Research (tools for managing design geometries of large systems), Boeing Applied Research and Technology (aerospace vehicle design), and Shell Exploration and Production (energy). Personnel at these organizations were asked to identify typical engineering processes that would benefit from the capabilities of an AEE, shortfalls of current processes, desired improvements, and the implications of process improvements with regard to specific functional attributes of AEE system components. Survey responses received by the committee are summarized in Tables 3-2 and 3-3.
As Table 3-2 shows, most requirements involve computation, modeling, and software. Relatively few requirements are related to human-centered computing or hardware and networks, perhaps because users are more familiar with requirements related to computation, modeling, and software than with requirements related to the other three areas. Also, the survey did not specifically ask for inputs related to human-centered computing.
R&D focused on the common themes listed in Table 3-3 would have the greatest impact on the highest priority processes identified by the respondents. Based on these common themes, the committee identified 11 specific opportunities for NASA to conduct broadly applicable R&D through the creation of partnerships with industry and academia (see Box 3-1).
Alternate Approaches
The committee identified three basic approaches for improving engineering processes. The first approach is an aggressive effort to maximize overall effectiveness by completely re-engineering existing processes and facilities using
Table 3-2 Survey of AEE Requirements
|
Organization |
Engineering Process |
Improvements Desired |
Computation, Modeling, and Software |
Human-Centered Computing |
Hardware and Networks |
|
Caterpillar |
preliminary and detailed design of whole vehicle system |
• 30x reduction in design-cycle time |
• 10 gigabyte CAD file |
• advanced visualization for engineering workstations |
• T-3 circuits (at 45 megabits/sec) between plants |
|
|
|
• standards for concurrent product and process development |
• parametric design |
|
• Internet-based communication with suppliers |
|
|
|
• collaboration among companies |
• system model with flexible body dynamics and all nonlinearities |
|
|
|
|
|
|
• distributed computing environment |
|
|
|
|
|
|
• probabilistic analysis |
|
|
|
|
|
|
• manufacturing and assembly simulation |
|
|
|
|
|
|
• eliminate CAD translation |
|
|
|
|
|
|
• standardized approach to systems engineering, component representation, assembly representation, test/analysis information, material properties, and boundary conditions |
|
|
|
Ford |
whole vehicle design |
• 3x reduction in design-cycle time |
• 500,000 element mesh |
• advanced visualization (stereographic, holographic, and immersive) |
• 10x increase in speed for common product information systems |
|
|
|
• improved quality |
• parametric and stochastic models |
• full 3-D animation of total vehicle with real- time analysis updates |
• 10x to 30x speed increase for pre- and post-processing |
|
|
|
• smaller number of prototypes |
• CAE associative with CAD |
|
• 20x increase in speed for analyses |
|
|
|
|
• sources of variability including material and dimensional stability |
|
|
|
|
|
|
• cost and manufacturability analysis |
|
|
|
|
|
|
• cross-attribute optimization |
|
|
|
|
|
|
• design for robustness |
|
|
|
|
|
|
• collaboration for global teams |
|
|
|
|
|
|
• standardized system engineering tools |
|
|
|
|
|
|
• standards for CAD and CAE |
|
|
|
Simmetrix |
simulation- based design of component and system |
• CAE for design, not verification |
• generic parametric CAD model linked to attribute-specific CAD models |
|
|
|
|
|
• faster analysis |
• open CAD systems |
|
|
|
|
|
• better linkage of models |
• geometry-based adaptive mesh |
|
|
|
NASA X-38 |
design of a space rescue vehicle |
• faster analysis, more effective collaboration, and better configuration control |
• single CAD package |
|
• low-cost video |
|
|
|
• reduced testing |
• linkage of CAD to in-house analysis packages |
|
|
|
|
|
|
• single analysis package |
|
|
|
|
|
|
• integration of structural analyses |
|
|
|
|
|
|
• integration of thermal, aerodynamic, and manufacturing analyses |
|
|
|
|
|
|
• collaboration for global teams |
|
|
|
|
|
|
• reliable CAM transmission to vendors |
|
|
|
Organization |
Engineering Process |
Improvements Desired |
Computation, Modeling, and Software |
Human-Centered Computing |
Hardware and Networks |
|
Boeing Electromagnetics |
analysis of electromagnetic radiation effects for large systems |
• faster analysis |
• rapid model creation |
• less involvement by users in managing analyses |
• gigabit/sec data transmission |
|
|
|
• better 3-D representation |
• 1,000,000+ element mesh |
|
• higher speed for analyses |
|
|
|
• optimization capability for large models |
• adaptive mesh refinement |
|
|
|
|
|
|
• integrated solvers |
|
|
|
|
|
|
• handle 10x larger problems |
|
|
|
|
|
|
• rapid design updates |
|
|
|
|
|
|
• optimization integrated with analysis |
|
|
|
|
|
|
• support for analytical "design of experiments" |
|
|
|
|
|
|
• binary standard for data transfer |
|
|
|
Boeing CAD |
management of geometry and configuration for very large systems, particularly at the concept stage |
• shorter process |
• elimination of data translation |
• visualization of large data sets on low-end workstations |
|
|
|
|
• increased model size |
• product information management for configuration control |
• real-time collision detection for 15 to 20 gigabyte data sets |
|
|
|
|
• better process management |
• process modeler for management of dependent processes |
• high-speed haptic interfaces for gigabyte data sets (e.g., for assembly sequence verification) |
|
|
|
|
• faster, more capable trade- off studies |
|
• knowledge-based engineering tools |
|
|
Boeing Applied Research and Technology |
whole aerospace vehicle design |
• 2x to 10x reduction in cost and cycle time |
• rapid model updating |
• decision support capability |
• platform-independent environment for analyzing and viewing |
|
|
|
• control of product cost |
• design reviews of 10,000+ part assemblies |
• better visualization tools for exploring design space |
• Internet-based information delivery |
|
|
|
• improved distance collaboration |
• design cost linked to CAD |
• haptic interfaces |
• virtual collocation |
|
|
|
|
• nonlinear solvers |
• natural language interfaces |
|
|
|
|
|
• multifunction design optimization |
|
|
|
|
|
|
• cost trade-off studies |
|
|
|
|
|
|
• open environment for planning, design, and analysis |
|
|
|
|
|
|
• distributed decision making |
|
|
|
|
|
|
• generalized product data structures |
|
|
|
|
|
|
• standard approach to rapid modeling/generative design |
|
|
|
Shell Exploration and Production |
evaluation of prospecting data from potential underground energy source |
• 10x improvement in speed of evaluation |
• 3-D graphics @ 3 million polygons/sec |
• interactive visualization |
• interoperable platforms |
|
|
|
• collaboration across remote sites |
• 4 gigabyte data sets |
• photograph-quality maps (1,000 dpi) |
• satellite transmission from offshore prospects |
|
|
|
|
• 10x increase in analysis speed (one week turnaround) |
• real-time sectioning and texturing |
• collaboration across sites |
|
|
|
|
|
• 30 frames/sec animation |
|
Table 3-3 Common Themes
|
Engineering Process |
Improvements Desired |
Computation, Modeling, and Software |
Human-Centered Computing |
Hardware and Networks |
|
broad-based design, development, and support of complex products |
• reduced cycle time |
• gigabyte file size for pre- and post-processing of 10,000+ part assemblies |
• stereographic, holographic, immersive environments |
|
|
|
• reduced risk and higher quality |
• rapid modeling, model modification, and preparation for analysis |
• desktop visualizations |
|
|
|
• reduced total cost |
• 1,000,000+ model (grid) size |
• less involvement by user in managing efficiency of analysis |
|
|
|
• improved collaboration |
• parametric design |
• high-speed haptic interfaces for gigabyte data sets |
|
|
|
|
• "open" CAD systems linked to CAE, CAM, product information management |
• 100 megabit/sec communication |
|
|
|
|
• accommodation of legacy data |
• 10 to 30x speed increase for pre- and post-processing |
|
|
|
|
• support for manufacturing applications |
• 20x increase in speed for analyses |
|
|
|
|
• integration of specific analyses |
• high-speed data transfer |
|
|
|
|
• multiple-attribute analysis simultaneously from a single model (function, cost, support, and manufacturing) |
• Internet-based communication with suppliers |
|
|
|
|
• incorporation of probabilistic analysis, nonlinearities, material and dimensional stability over time |
|
|
|
|
|
• support for analytical "design of experiments" techniques |
|
|
|
|
|
• multifunction optimization integrated with analysis |
|
|
|
|
|
• using standards to eliminate translation between different applications |
|
|
|
|
|
• imbedded intelligence |
|
|
|
|
|
• support for global teams and distributed decision making |
|
|
|
|
|
• open distributed environments supporting total product definition, configuration management, and lifetime support |
|
|
|
|
|
• standards for CAD; systems engineering; component and assembly representations; test and analysis data; and geometry, properties, boundary conditions, and results |
|
|
AEE technologies and systems. Premature implementation of developmental technology in an operational setting, however, could be risky and expensive. If the new technology doesn't live up to expectations, the result could be ineffective and counterproductive. In fact, long-term damage could be done to the prospects for incorporating AEE technologies into operational settings because a bad experience with immature technology would make it more difficult to justify the use of advanced technologies in the future.
The second approach would adopt and integrate AEE technologies with existing systems and practices gradually, using staged implementation. Improvements made one at a time would be tested in practice and optimized before the next innovation is implemented.
The third, and most conservative, approach would defer the use of AEE technologies until a proven, comprehensive system has been developed. In the meantime, this approach
Figure 3-1
Approaches for improving engineering processes.
|
Box 3-1 Opportunities for NASA-Industry-Academia Partnerships (based on a survey of seven private and public organizations) Gigabyte file sizes and processing of assembly data sets with more than 10,000 parts. AEEs need the capability to display and modify large data sets. The space shuttle, the international space station, and many NASA payloads have more than one million parts. Rapid modeling, model modification, and preparation for analysis. Desired reductions in engineering cycle time will require AEEs to generate models of new products and processes rapidly, to ensure that models of modified products and processes are consistent with models of the original products and processes (to facilitate comparative analyses), and to validate new models and capabilities before they are released for operational use. Models with more than one million grids. To construct engineering models for current and future NASA projects, AEEs must have the ability to support models with large regular and irregular grids. Accommodation of legacy data. Although AEEs will be used for the development of new products and processes, they will not operate in isolation from the current inventory of products, processes, and support services. Thus, specific attention must be given to the ability of AEEs to accept legacy data sets and interface with legacy systems for analysis and other engineering activities. Support for manufacturing applications. NASA, like many large engineering organizations, conducts manufacturing operations using both in-house facilities and contractors. AEEs should have the ability to export data to and interface with manufacturing systems operated by other organizations. Visualization. Collaborative (i.e., large-scale and/or distributed) visualization has demonstrated its ability to streamline the elements of the engineering enterprise that cross disciplinary and geographic boundaries. Thus, AEEs should provide immersive and nonimmersive three-dimensional displays and, potentially, nonvisual simulations (e.g., auditory, haptic, and vestibular simulations).1 Integration of individual analyses and simultaneous analysis of multiple attributes using a single model. Today, analyses of system functions, structural properties, thermal properties, cost, and/or manufacturability are usually conducted independently. AEEs should have the capability to do simultaneous analyses of specified sets of multiple attributes. This would be facilitated by the use of a single model from design to manufacture. Multifunction optimization integrated with analysis. In addition to analyzing a product's performance against each attribute, AEEs should be able to define an optimal balance of performance when attributes compete for the same resource, such as power or weight. Traditional iterative methods are slow, costly, and inexact. Multifunction optimization provides a rigorous approach to finding the best balance. Although multifunction optimization is developing rapidly, it must be tailored to the specific attribute being analyzed. The needs of industry and government agencies, such as NASA, overlap in this area, so collaborative work could be fruitful. High-speed haptic interfaces for gigabyte data sets. Haptic interfaces provide users with the opportunity for direct, feedback-enabled interactions with models. These interfaces can enhance the ability of design engineers to manipulate rapidly multicomponent objects of very large size. Support for analytical "design of experiments." The use of a designed experiment (e.g., Taguchi methods) is a powerful statistical technique for simultaneously studying the effects of multiple variables on a system. For example, this technique is a powerful tool for establishing the robustness of a design or process in response to noise factors. To minimize the need to evaluate complex products and missions experimentally, AEEs should have the capability to rapidly create analytically designed experiments that operate seamlessly with the models and analyses used to evaluate attribute performance. This capability is especially important to the development of autonomous vehicles. Open, distributed environment supporting total product definition, configuration management, and lifetime support. This is essentially the principal goal of AEE R&D conducted by NASA. It can be realized through the use of common object models in combination with hardware and software that demonstrate a high level of interoperability. |
Table 3-4 Estimated Effectiveness of Alternative Approaches
|
|
Alternative 1 Aggressive Implementation of AEE Systems |
Alternative 2 Staged Implementation of AEE Technologies |
Alternative 3 Evolution of Conventional Technologies |
|
Risks |
|
|
|
|
Technology insertion |
high |
medium |
low |
|
Information complexity |
high |
medium |
medium |
|
Cultural impacts |
high |
medium |
low |
|
Total cost of implementation |
high |
medium |
medium |
|
Time required for implementation |
3-5 years |
1-3 years |
< 1 year |
|
Costs |
|
|
|
|
Software tool interoperability |
high |
medium |
low |
|
Legacy systems migration |
high |
medium |
low |
|
Commonality/standardization requirements |
high |
high |
low |
|
Training |
high |
medium |
low |
|
Information infrastructure requirements |
high |
medium |
medium |
|
Supplier interfaces requirements |
high |
medium |
low |
|
Maintaining systems effectiveness |
high |
high |
medium |
|
Potential Benefits |
|
|
|
|
Seamless interfaces |
high |
medium |
low |
|
Standardization |
high |
medium |
low |
|
Tools/information management |
high |
medium |
low |
|
Collaboration capabilities |
high |
medium |
low |
|
Real-time assessment |
high |
medium/high |
low |
|
Life-cycle management |
high |
medium |
medium |
|
Compatibility of product targets |
high |
medium |
low |
would rely on evolutionary improvements in conventional technology. This approach would unnecessarily postpone the benefits of implementing available AEE technologies.
Figure 3-1 is a conceptual comparison of the three alternatives. In the long-term, aggressive implementation of complete AEE systems has the largest potential payoff, but it also has the largest risks and highest costs. Staged implementation of AEE technologies is likely to outperform evolutionary improvements in conventional technology. Staged implementation is also likely to outperform the aggressive implementation of AEE systems, at least in the near-term and midterm, because it avoids the cost, time, and complexity of reinventing complete processes with unproven technologies. Table 3-4 characterizes the committee's estimates of the risks, costs, and benefits of these three alternatives. Before an organization selects an alternative, it should assess the subelements listed in Table 3-4 and their potential impact on the organization in question.
