Contribution to Virtual Manufacturing Background Research

By

Edward Lin, Ioannis Minis, Dana S. Nau and William C. Regli
Institute for Systems Research
University of Maryland
College Park, MD 20742

For

Lawrence Associates Inc.
5100 Springfield Pike, Suite 509
Dayton, Ohio 45431

Under Contract Number F33615-92-D-5812

May 1995

Delivery Order Final Report for Period February 1995 through May 1995

Prepared for: Manuacturing Technology Directorate
Air Force Wright Laboratory
Air Force Systems Command
Wright-Patterson Air Force Base, Ohio 45433-6533

Executive Summary

In this document, we identify, assess and categorize research and applications relevant to key aspects of Virtual Manufacturing (VM). We identify gaps in these research and application efforts, and present an outlook for the future of VM technologies. Below is a brief summary of our results:

Contents

List of Figures

List of Tables

1 Introduction and Background

 

This document presents the results of work performed by researchers at the University of Maryland as part of a subcontract to Lawrence Associates in support of the ManTech contract Manufacturing Technology Special Advanced Studies.

Three major paradigms have been proposed for VM: Design-centered VM, Production-centered VM, and Control-centered VM [36]. Figure 1 depicts the relationships among these three types of VM with respect to the virtual product life-cycle. In this figure, the three blocks represent the three types of VM; the relevant interactions (or information flow) are represented by directed arcs. For example, the information (such as product models) provided by Design-centered VM to Production-centered VM, is represented by a directed arc from the Design-centered VM block to the Production-centered VM block.

These three different types of VM have the following characteristics:

  
Figure 1: Virtual Manufacturing Product Life Cycle [POSTSCRIPT IMAGE HERE]

In order to identify, assess and categorize research and applications relevant to key aspects of VM, we did a detailed study of ten different questions dealing with various aspects of VM. The specific questions are listed in Section 2.1; and our approach for gathering and assimilating information about them is described in Sections 2.1 and 3.

Section 4 contains the results of our work, including detailed definitions, background, and discussion for each question. Based on these results, Section 5 identifies gaps in these research and application efforts, and present our outlook for the future of VM technologies.conclusions.

Additional material is contained in the appendices, and at the following web site: http://www.glue.umd.edu/~lin/vmproject.html

This web site is a temporary location. Once we have finished some cleanup work on the material stored there, we will move it to a permanent location, and publicize the new URL.

2 Objective and Approach

 

2.1 Objective

  The goals of this project were to review several different aspects of Virtual Manufacturing (VM), with the following objectives in mind: to identify, assess and categorize research and applications relevant to key aspects of VM; and to identify gaps in these research and application efforts, and present our outlook for the future of VM technologies. The particular questions on which we focused were the following:

  1. Design-Centered VM.
    Question 1.1. To what extent can manufacturability (determined via modeling & simulation) be used to make performance tradeoffs, at conceptual design and detailed design?
    Question 1.2. What levels of detail in VM are necessary to support making these manufacturability assessments?
    Question 1.3. To what extent can existing models of parts/processes be used to support the manufacturability assessments using VM? Are there any models that describe plant-specific process capabilities?
    Question 1.4. What are the alternative mechanisms for displaying manufacturability results?
    Question 1.5. What is the reliability of cost, schedule, and quality estimates made using VM during the conceptual and detailed design stages?
    Question 1.6. Are/should different types/levels of models necessary for using VM to make manufacturability estimates versus cost/schedule/quality estimates?
    Question 1.7. Can VM effectively support process planning?

  2. Production-centered VM.
    Question 2.1. Will VM enable a stronger movement toward distributed manufacturing? What advantages accrue from distributed manufacturing? Is it any different from current supplier relations? Does the availability of an ``information highway'' offer opportunities to significantly change the way manufacturing is accomplished? Does VM have a role in facilitating distributed design?

  3. Generic VM issues.
    Question 3.1. What are the relationships between VM, Virtual Prototyping, the Virtual Enterprise, etc.?
    Question 3.2. What is the role of object oriented technology in VM?

2.2 Approach

  The proposed approach for achieving the objectives described in Section 2.1 was as follows:

3 Description of Effort

 

Personnel.

The contract for this work was initially awarded to Ioannis Minis and Dana Nau, faculty in the Institute for Systems Research at the University of Maryland. However, in order to carry out this work more effectively, we also recruited the help of two other University of Maryland personnel: Edward Lin and William C. Regli. Appendix B contains brief biographical summaries for all four of the investigators.

Steps Taken.

To carry out the proposed work, we did the following:

Having gathered all this information, we performed the assessment described in Section 2.2; the results of this assessment appear in Section 4. We also set up a World-Wide Web site to contain the information that we gathered and the results of our assessment, so that it would be available to everyone working on the project. The temporary URL for the web site is http://www.glue.umd.edu/~lin/vmproject.html

As of the press date for this report, we are still doing some additional cleanup work on this web site. We intend to reformat some of the documents so that they are more readable and include more hypertext links. In addition, we intend to include a link to an HTML version of this report. Once such things have been done, we will move the web site to a permanent location, and distribute the new URL to interested parties.

4 Assessment

 

This section contains a detailed assessment of the questions stated in Section 2.1, which relate to the Design-centered and Production-centered VM. For each question, we have included a definition of the basic issues involved, some background material, and a discussion.

4.1 Design-Centered VM

4.1.1 Question 1.1

To what extent can manufacturability (determined via modeling & simulation) be used to make performance tradeoffs, at conceptual design and detailed design?

Definition.

Evaluating the manufacturability of a proposed design involves determining whether or not it is manufacturable with a given set of manufacturing operations---and if so, finding the associated manufacturing efficiency.

Background.

The idea of analyzing the manufacturability of designs date as far back as World War II [152]. Scarcity of resources, and constant social and political pressure to build better weapons in the shortest possible turnaround time were the main motivating factors behind the tight integration of design and manufacturing activities. Many of the successful weapons of that period were designed by small, integrated, multi-disciplinary teams [152]. However, with the the post-World War II era of prosperity and the resultant rapid growth of industry size came the segregation of design and manufacturing activities, which resulted in a sequential product development environment, with little attention to manufacturability during the product design phase.

Increasing global competition and desire to reduce lead time led to the rediscovery of the design-for-manufacturability philosophy in the late 1970s. Some of the earliest attempts involved building inter-departmental design teams that consisted of representatives from the design and manufacturing departments. In these design projects, manufacturing engineers participated in the design process from the beginning and made suggestions about possible ways of improving manufacturability [51,69]. Such inter-departmental design teams did not always work harmoniously and many management-related problems existed when building and coordinating such teams [105].

In an attempt to increase the awareness among designers of manufacturing considerations, leading professional societies have published a number of manufacturability guidelines for a variety of manufacturing processes [8,10,12,116,140]. Some companies produced and used their own guidebooks for designers---one of the pioneers was General Electric [38]. These guidelines mainly enumerate the design configurations that pose manufacturability problems and were intended as training tools in design for manufacture (DFM). To practice DFM, the designer had to carefully study these guidelines and try to avoid those configurations that result in poor manufacturability.

The availability of low-cost computational power is providing designers with a variety of CAD tools to help increase productivity and reduce time-consuming build-test-redesign iterations. Examples include tools for finite element analysis, mechanism analysis, simulation, and rapid prototyping. The availability of such tools has become a driving force for research in concurrent engineering where various product life-cycle considerations are addressed at the design stage. As the advantages of concurrent engineering are being realized, more and more downstream activities associated with the various manufacturing aspects are being considered during the design phase, and DFM has become an important part of concurrent engineering [8,146].

One of the primary goals of concurrent engineering is to build an intelligent CAD system by embedding manufacturing related information into such systems. In intelligent CAD systems, DFM is achieved by performing automated manufacturability analysis---a process which involves analyzing the design for potential manufacturability problems and assessing its manufacturing cost. Inadvertent designer errors, such as missing a corner radius or excessively tight requirements for surface finish, that go undetected during the design stage may prove costly to handle in a fully automated CAD/CAM system (i.e. the system might select an expensive manufacturing operation to achieve that erroneous design attribute).

It is anticipated that a systematic methodology for manufacturability analysis will help in building systems to identify these types of problems at the design stage, and provide the designer with the opportunity to correct them. Such systems will alleviate the need to memorize and study the manufacturability checklists, therefore allowing the designers to focus on creative aspects of the design process. Moreover, as the manufacturing resources or practices change in an organization, the knowledge bases of these intelligent CAD systems could be updated automatically with least interference with the design activities of the organization.

It has become evident that the task of manufacturability analysis requires extensive geometric reasoning. As the field of solid modeling has matured, functional and architectural improvements in modelers have facilitated increasingly sophisticated types of geometric reasoning. The closed architecture solid modeling systems of the late 1980's did not allow easy access and manipulation of geometric and topological entities, most of the computer-aided DFM tools developed in that period did not rely on extensive geometric reasoning. This, in turn, limited their capacity for handling complex design shapes. In recent years, the functional capabilities of commercial systems has been vastly improved. These new enhancements, coupled with the advent of open architecture solid modeling systems [137], facilitate implementation of the complex geometric reasoning techniques required for realistic manufacturability analysis.

Discussion.

Existing design technology is mainly focused on supporting detailed design activity. CAD software systems, while providing increasingly sophisticated means of manipulating shape and form represented in the computer, are poor at representing the information critical at conceptual design. For the most part, conceptual design information is lodged in the minds of the members of the design team in the form of recollections of previous experiences with similar products and etchings in notebooks. Existing software tools for conceptual design, such as the cost estimation packages available from Boothroyd and Dewhurst Inc. (BDI), provide valuable means to get rough estimates on cost and related trade-offs without extensive analysis of geometric models.

For VM to have an impact on conceptual design, it will be crucial for in next-generation systems to seamlessly integrate conceptual-stage design analysis tools (such as those from BDI) with some form of Computer Aided Conceptual Design (CACD). While there has been a flurry of research activity surrounding topics related to modeling non-geometric design information (such as history and intent), there is a need for these representations to be integrated into some form of conceptual design system that can effectively aid designers in creating and managing the ideas that will eventually lead to a shape to be manufactured.

4.1.2 Question 1.2

What levels of detail in VM are necessary to support making these manufacturability assessments?

Definition.

Given a computerized representation of the design and a set of manufacturing resources, the automated manufacturability analysis problem can be defined as follows:

  1. Determine whether or not the design attributes (e.g., shape, dimensions, tolerances, surface finishes) can be achieved.
  2. If the design is found to be manufacturable, determine a manufacturability rating, to reflect the ease (or difficulty) with which the design can be manufactured.
  3. If the design is not manufacturable, then identify the design attributes that pose manufacturability problems.

Background.

Researchers have developed several different approaches to evaluate manufacturability of a given design [8,49,62,79,135,40]. Some of these have been developed for specific application domains, while others have been developed for general domains. Existing approaches can be classified roughly as follows:

Discussion.

In general, there is a tradeoff between computational requirements and accuracy. Direct approaches are less computationally intensive---but in some domains it is difficult for a direct approach to give good results. Plan-based approaches can produce accurate results in cases where direct approaches have problems---but they can require large amounts of computing time in order to generate and evaluate plans. As the cost of computing power continues to decrease, we anticipate that the computational requirements of plan-based approaches will become less of a limitation, and thus plan-based approaches will become increasingly attractive.

4.1.3 Question 1.3

To what extent can existing models of parts/processes be used to support the manufacturability assessments using VM? Are there any models that describe plant-specific process capabilities?

Definition.

Virtual manufacturing requires a robust information infrastructure that comprises rich information models for products, processes and production systems [85].

Background.

In the product information modeling area, substantial work has been done by the International Standards Organization (ISO) to develop models for the representation of product information, including product configuration (BOM) [108], and product shape (geometry and topology [109] and form features [111]). These generic resource models support the development of application protocols which, in turn, focus on a certain product application area of wide interest. Application protocols under development include: Mechanical design [114], electrical/electronics connectivity [112], electronic PC assembly, design and manufacturing [113], and ship structures [115]. It has to be emphasized, however, that despite the progress by ISO in developing standard (STEP) product models, there has been a substantial lag in the adoption of these models by CAD vendors and industry. Instead, custom product models are used internally by various CAD systems, and limited information can be exchanged in a standard format; i.e., product geometry using the IGES standard, and product topology using IGES 5.0 [107]. The former is supported by most CAD systems, while the latter is supported by only a few.

Manufacturing process models typically relate critical process parameters to those attributes of the product that are significant to the product's function. Process models assume many forms including physical or mathematical models derived from first principles, statistical models derived experimentally, computer process simulations, and simple tables and/or rules found in manufacturing handbooks. It is emphasized that the modeling of manufacturing processes has been the subject of vast research in the engineering and material science fields. For example, a recent survey of work in machining dynamics, which is just one of the research areas in machining, reported more than 200 journal publications and several books describing various aspects of modeling the dynamics of turning, milling and grinding processes (see also [73]). Process models are invaluable for, and are routinely used in, process design and control. On the other hand, simple handbook tables, formulas and rules (e.g. [8,12]) have been used in Integrated Process and Product Development (IPPD) especially in the area of design critiquing and manufacturability assessment (see for example [149,41,127,57]). However, there has been little work to use more sophisticated process models in an integrated fashion with product design. An example of such work is the integration of a process planning system with a mathematical model that predicts the surface roughness of a machined component [102,59].

Production system models comprise the third important component of the information infrastructure for virtual manufacturing. There are at least two views of the production system that are needed to support virtual manufacturing: i) Representation of the system's capabilities and performance (static view), and ii) representation of the system's dynamical behavior (dynamic view). The first is necessary in order to assess the feasibility of manufacturing a certain product design in a certain production system. In addition, knowing the system's capabilities and performance, it is possible to compare them with the production requirements of the design, and, thus, assess the manufacturability of the product design with respect to the production system. The second system view is necessary to predict production performance based on given demand characteristics and the system state. Performance attributes of interest include cycle time, inventory levels, and on-time delivery. Informal representations of both system views are currently available. For example, ISO 9000 [26] attempts to capture the capabilities of the quality management systems of an enterprise. Furthermore, all MRP II systems include information about the resources of a manufacturing system (availability, capacity), the products manufactured (item master records, bill-of-materials, routings), and the dynamic system state (shop floor control data) [143]. Finally, production simulation languages and software tools are widely available to model and analyze various production scenaria (see for example [121,15]). However, formal information models of production systems have yet to emerge. Research is currently under way in this area prompted by the recent interest in distributed manufacturing and virtual enterprises. The AIMS (Agile Infrastructure for Manufacturing Systems) program is creating a format for standard data sheets, similar to the entries in a Thomas Register, that describe the capabilities, costs and availability (CCA) of a manufacturing shop [60]. Partner CCAs will be available upon request in the Internet and will be reviewed by a designer to select qualified partners that satisfy critical processing and capacity requirements. References [17] present the development of an information model that captures the capabilities and performance of a manufacturing firm (static view). The model has a hierarchical form and represents critical financial attributes of the firm, the products manufactured, information on the firm's quality management systems (adopting ISO 9000), engineering systems, production management systems and manufacturing processes. For manufacturing processes, the model captures their availability, capabilities (e.g. workspace, accuracy, repeatability, number of axes, etc.) and performance (capacity, yield, cost rates). The model has been developed in EXPRESS and has been implemented in an object oriented database (Objectstore).

Discussion.

There are several advancements in product, process and system models that are needed to effectively support manufacturability assessments using virtual manufacturing.

In addition to shape information, product models should be able to capture data that are directly relevant to manufacturing, such as tolerances (dimensional and geometric), and form features. Tolerances are essential in evaluating the manufacturability of a design with respect to the capabilities of manufacturing processes. As a result, tolerance modeling has been the subject of both research (see e.g. [43]) and of standards development efforts [47,110]. However, no existing CAD system has full tolerancing capabilities. In the features area there has been considerable difficulty in developing standard representations [111] that may be used to provide a process view of the design and to support design by features. There is also considerable doubt among researchers whether such standard representations will be practical. In addition, despite extensive research [92,129,84,144,6,24,52,25,119,124,142], the practicality of feature extraction from boundary product representations is still limited. Further research is therefore necessary in the form features area to develop powerful methods for representing and reasoning about features, which are critical for both process selection and planning, and for manufacturability evaluations.

Although process modeling is a mature subject, it has yet to play a significant role in concurrent engineering practice. One of the major reasons for this is the lack of unified ways to deal with process models in an automated design environment. This also appears to be a major roadblock for virtual manufacturing; especially in assessing the feasibility of producing a design with a certain set of manufacturing processes and evaluating the ease of manufacture with these processes. Thus, novel representations are necessary to capture diverse process models (analytical, statistical and simulation-based models) and to provide unified interaction mechanisms with the virtual manufacturing environment.

Two major areas require further investigation to realize the production component of virtual manufacturing. First, as pointed out in the background section, new information models are necessary to capture the capabilities and performance of production systems, and thus provide plant-specific information to the virtual manufacturing system for design evaluation. Secondly, it is necessary to develop methods for integrating product design and process planning with production planning and scheduling. By doing so, the designer will be able to determine early in the design stage the impact of design decisions in production planning and scheduling. Process planning defines the sequence of operations for the realization of a design and, therefore, provides the feasible subspace for both planning and scheduling. However, product design and process planning are time-independent, while production planning and scheduling are dynamic activities. This complicates the integration and presents a challenging topic for further research.

4.1.4 Question 1.4

What are the alternative mechanisms for displaying manufacturability results?

Definition.

Manufacturability analysis can vary greatly in scope. On one end of the spectrum, manufacturability during and early-stage design has to consider different manufacturing process and resource scenarios. During a more detailed design phase, manufacturability might refer to an estimation of cost and quality for individual components.

Further complicating matters is the fact that components rarely exist in isolation---they must be first machined and then assembled. The pioneering work of Boothroyd and Dewhurst [11] in developing the design-for-assembly guidelines revealed that assembly, in fact, often drives the majority of the cost of a product. A fundamental tenet of their methodology is that design for assembly precedes design for manufacture. Their results show that if a product is designed to be easy to assemble it will likely also be easy to manufacture.

In analyzing the many factors impacting manufacturability, the question arises of how to handle the results produced by the many different forms of analysis. What kinds of measurements and assessment of manufacturability be made and how can these results be fed back to the user?

Background.

Existing commercial and academic systems have attempted several approaches to manufacturability assessment:

Discussion.

It is clear that none of the above rating schemes are completely satisfactory. Several observations and recommendations can be made:

  1. For the most part, these available metrics are for individual manufacturing processes. In reality most products will be manufactured using multiple processes (e.g., a part is first cast and then mating surfaces are machined). Further, a crucial element of the design phase is evaluating the tradeoffs between the alternative manufacturing processes that are available. For example, consider the case where a designer is developing a conceptual design for a rear stabilizer for a new airplane. Initially she may have a general specification for the shape of the structures and rough outlines for the various components. The detailed design will depend on the process chosen: the cost and performance of a wing made of graphite composites will be very different from one made from machined parts and sheet metal. Future systems must be able to address multiple processes and their trade-offs.

  2. Most of the methodologies for rating manufacturability are limited to the consideration of single components---with the notable exception of design for assembly software. In current DFA tools (such as those of Boothroyd and Dewhurst Inc.), design worksheets for describing the function of components in assemblies help designers identify and eliminate redundant parts.

    In future, however, such techniques must be tightly integrated with CAD. Further, problems discovered by DFA analysis are closely related to manufacturability analysis: modifications to the assembly provides an opportunity to redesign parts to be more manufacturable. Part of the manufacturability rating of an individual part must include its relationship to the product assembly. With current technology, this two-way many-to-many communication between assembly-manufacturability is largely unautomated and thus highly labor intensive. Automation of this communication loop and its tight integration with CAD are ripe areas for research.

  3. Ratings schemes should enable the designers in identifying and eliminating the manufacturing bottlenecks. If a particular component is estimated to have ``poor manufacturability'' or ``high production cost'' it is important to be able to represent the cause of this rating and feed the reason back to the designer. For example, if high production cost is due to an excessively tight tolerance or difficult geometric configuration, it will be necessary for future systems to identify the specific design attributes causing the problem. With current technology, ratings are often not tied to any specific aspect of the part being rated and hence it can be difficult for designers to determine exactly what is causing manufacturing problems.

  4. To move beyond rudimentary cost estimation and rating schemes it is necessary to develop algorithms that help to automate the redesign task. In the context of existing systems, redesign is a whole unautomated task, leaving it to the designer to decide how to interpret the design's problem and how they can be fixed while continuing to adhere to the design's functional intent and design constraints.

    Some methods for automated redesign have been developed, such as algorithms for modifying a design in order to reduce the number of machining setups [30]. But, in isolation, such algorithms are of limited benefit. It is unclear how these methods will scale to the problems posed by redesign of a multiple-component assembly or changes to manufacturing and production resources.

4.1.5 Question 1.5

What is the reliability of cost, schedule, and quality estimates made using VM during the conceptual and detailed design stages?

Definition.

Reliability of an estimated value of a criterion (such as cost, schedule, or quality) is defined as the closeness of that estimate to the average value of the criterion resulting from actual manufacturing.

Background.

Estimating the cost and processing time (set-up and run) of a given design is a traditional activity performed in industry when responding to request for quotes. There exist numerous industrial handbooks that support this activity. References [34,147,148] are typical cost and time estimation handbooks for mechanical parts, while [66,28] focus on printed circuits and boards. The typical use of these handbooks requires a detailed description of the product design. Design and manufacturing parameters (such as shape, dimensions, tolerances, material and production quantity) are then used as inputs to rules, formulae, tables and graphs in the handbook to derive the cost and time estimates. There also exist computerized versions of such handbooks that require user input on the product's design and use similar formulas and tables to determine the cost and time estimates and provide them to the user (see e.g. [32,91]).

The research being conducted over the last two decade on estimating the manufacturability of a design is also relevant. Various measures of manufacturability have been defined by several authors. In general, these measures have an indirect (at best) correlation with cost, manufacturing cycle time and even quality. A good survey of the literature in this area is given in [150]. A quantitative scheme for rating the manufacturability of turned parts is also given in [150]. Reference [57] presents a method to estimate the cost and machining time of prismatic milled parts using a CAD part representation. From a solid model of the part the method extracts alternative machining features (that closely related to machining operations), determines the most suitable ones with respect to manufacturing, and derives the corresponding milling operations to minimize the cost and processing time. The expert system described in [64] also uses process planning for design evaluation.

Discussion.

The manual (or semi-manual) estimation techniques described above require a detailed description of the design, and knowledge of the processes to be used in production. Since they are based on empirical knowledge, which has been derived from years of experience, they typically provide reliable estimates for both the cost and the processing time.

Manufacturability-related studies have automated the design critiquing process to a certain extend. The product and process information used in such studies may vary greatly in detail. Some methods assess the manufacturability based on information that is known at the initial design stages (see e.g. [48]). Other methods (e.g. [57,64]) require a fully developed design. As discussed above, however, most studies use indirect metrics for design critiquing, which quantify the relative and not the absolute difficulty of manufacture. Thus, it is difficult to assess the reliability of the manufacturability estimates. Even these methods that do estimate processing times, do not account for the dynamics of the production system, and therefore they cannot estimate the product's lead (or cycle) time which contains queuing time. (Note that the latter may range from 50 to 95% of the cycle time). Similarly, although it may be possible to estimate material and labor costs, it is not feasible to estimate inventory costs without considering the dynamics of the production system. Finally, the ability to estimate product quality is minimal since there manufacturability studies do not generally use sophisticated process models.

Virtual manufacturing, when mature, is expected to be able to provide accurate estimates for processing times, cycle times and costs (including inventory), as well as product quality. This is because VM can model both the processes employed for the product's manufacture and the production system dynamics. By employing comprehensive models of manufacturing processes, VM will be able to accurately predict set-up times and run times, and, consequently, labor costs. Furthermore, if these process models are able to predict the variance of key product attributes, then process yields or the values of quality ratios (such as C pk) may be obtained by comparing the process capability with the corresponding design tolerances. On the other hand, modeling the production process will yield queue times, as well as Work-In-Process and finished goods inventory. Consequently, accurate estimates of overall cycle times and overall costs may be obtained. The potential of VM to provide accurate cost, lead time and quality estimates is a major motivation for further research and development in this area.

4.1.6 Question 1.6

Are/should different types/levels of models necessary for using VM to make manufacturability estimates versus cost/schedule/quality estimates?

Definition.

Existing technology is based largely around geometric representations of designs. Such models of design information are severely restrictive, for in most cases creation of the design's geometry is the final step in the process. Currently essential information about the conceptual design, the design's history, and the functionality are represented in a largely ad hoc manner (if at all). Further complicating matters is the fact that individual analysis applications, such as finite element analysis or a physically-based simulation, often require their own special product models. The results of their analyses are specific views onto the particular aspect of the design. For example, a table of critical stress points might be the product of the finite element system. Each individual system has is own sets of results, thus making it difficult to unify them into an broader assessment of manufacturability.

Further complicating matters is the fact that information about manufacturing resources (such as data regarding materials, tools, processes, etc.) is represented in a largely ad hoc manner. This makes it difficult to share resource information across applications---often forcing users to key in the data for their specific manufacturing facility into each separate software package. Sharing and reasoning with resource data through the software tools in a Virtual Manufacturing environment will be critical.

Background.

We have already presented an overview of models for parts and processes, for Question 1.3 above. We now augment this overview by brief discussing models of design functionality. A number of researchers have investigated how to represent the functionality of a part in its CAD model. In most cases the goal of the research was to find general ways to represent the functionality of designs, and so the scope of the work was very broad. In others, the research focused on a specific class of products where the features and functionality are directly coupled. Below are some examples of both:

Discussion.

Designers rarely start conceptualizing a product as geometry. Needed are models that enable designers to make ``back of the napkin'' sketches and calculations. These new representations will include several layers of abstraction; as the design progresses these layers will become increasingly detailed.

Currently, resource models and manufacturing practice information are, for the most part, stored in handbooks. In order to exploit this handbook information, software tools require highly detailed description of the design. However, analysis and interaction with other virtual manufacturing tools will occur at every point and time and at each level of abstraction. We need to be able to analyze these back-of-the-napkin drawings and provide guides to the designer on how to turn then into final drawing that are manufactured with the appropriate processes, easy to assemble, and surpass performance expectations.

For VM, it is evident that we need to move product representations beyond geometry. Further, new representations must be developed that can accommodate the heterogeneous manufacturability-relevant data produced different types of analysis. These representations must provide a means of accommodating the inclusion of as-yet unforeseen new software tools for engineering analysis. Lastly, tools for handling and analyzing this meta-level engineering information must be developed. In this way, VM can begin to break down the barriers between what are currently isolated software tools and enable truly collaborative and concurrent engineering.

Lastly, a true assessment of manufacturability includes some aggregation of the ratings for cost, schedule and quality over the entire product and at each level of abstraction. It is difficult to imagine how existing technology will scale in its attempt to manage the new magnitudes of manufacturing information. Creating tools that can intelligently assess the trade offs between conflicting design requirements and manufacturing constraints and produce an informed and useful rating of manufacturability is a great challenge in bringing about the Virtual Manufacturing workplace.

4.1.7 Question 1.7

Can VM effectively support process planning?

Definition.

Process planning is the task of planning what manufacturing steps to take in order to manufacture a desired product.

The two primary existing approaches to automated process planning are the variant and generative approaches. In variant process planning, the process engineer uses a Group Technology (GT) coding scheme to map a proposed design D into an alphanumeric code, uses this code as an index into a database to retrieve a process plan P0 for a CAD design D0 similar to D, and then modifies this process plan manually to produce a plan P for the design D. In generative process planning, the computer system attempts to synthesize the process plan P directly. For further information on variant and generative approaches, see [61,20,21,134].

Background.

A great deal of research has been done to try to develop practical techniques for generative process planning (e.g., [20,100,22,81,142,14,106,59,30,63,13], but very few generative systems are in industrial use [61,101], because of the extreme difficulty of devising computer systems capable of reasoning about the subtle problems that can arise in complicated process plans.

In process planning practice, variant techniques have been the tools of choice: they currently support almost all practical implementations of Computer Aided process Planning (CAPP) [70,90,130,139]. However, variant techniques also have limitations [99]. If the part mix varies over time, then for a new proposed design it may be difficult to find existing designs in the database that satisfy similar design specifications or require similar manufacturing processes. Furthermore, if new manufacturing processes or machines are made available in a plant and none of the process plans stored in the variant database use them, then the new processes and machines may be under-utilized unless the stored process plans are modified to use them.

Because of the limitations that the variant and generative techniques have individually, there is increasing interest in developing hybrid approaches that combine elements of variant process planning with generative process planning [99]. One approach is to use variant techniques to retrieve process plans for existing designs that are similar to the new one, and use these plans as a starting point for synthesizing a final plan for the new design. Another is to retrieve process plans for components of a complex assembly, and use generative planning techniques to combine these sub-plans a plan for the overall product.

Just as in variant process planning, the plans produced by such approaches may still need to be improved by the process engineer. However, by making generative modifications to the plan before presenting it to the user, such an approach should often be able to overcome some of the major drawbacks of variant process planning. Furthermore, since the generative component involves making modifications to pieces of the plan rather than constructing the plan from scratch, the plan-generation problem will be significantly more straightforward than full-fledged generative process planning. Thus, this approach should be able to combine the best features of variant and generative process planning, while avoiding the worst of the problems that they have individually.

Discussion.

Process planning is a very complex task which requires considerable experiential knowledge---and thus for the foreseeable future, most computer aided process planning systems will require a significant amount of supervision by an experienced human user.

Since variant process planning in the traditional sense is basically a scheme for retrieving existing process plans from a database, the role of VM in supporting traditional approaches to variant process planning appears to be rather limited. However, in the case of generative and hybrid approaches, where almost all implemented systems work by reasoning about the processes to be used, the story is much different. In existing systems, the rules for reasoning about the processes are in effect models of the processes---not necessarily simulation models, but models nonetheless. VM will have a clear and critical role in supporting generative and hybrid approaches to process planning planning: it will enable such approaches to include more sophisticated process models, so that they may produce more realistic plan evaluations.

4.2 Production-Centered VM

4.2.1 Question 2.1

Will VM enable a stronger movement toward distributed manufacturing? What advantages accrue from distributed manufacturing? Is it any different from current supplier relations? Does the availability of an ``information highway'' offer opportunities to significantly change the way manufacturing is accomplished? Does VM have a role in facilitating distributed design?

Definition.

Distributed manufacturing is performed by virtual enterprises. A virtual enterprise is a partnership of companies that forms in response to a certain market opportunity. The partners, who may geographically distributed and of various sizes and technical sophistication, contribute their core competence to the enterprise, enhancing its ability to deliver high quality, cost effective products to the market in a timely fashion.

Distributed design is performed by multiple designers who may be distributed geographically and who employ hetereogeneous design support systems.

Background.

The recent strong interest in virtual enterprises was sparked by the position paper of Nagel and Dove [97], which defined the concept of agile manufacturing. Distributed manufacturing is a key component of agility defined in [104], an all encopassing, enterprise-wide strategy that targets the ability to remain competitive in an environment of continuous change.

The realization of the vision of virtual enterprises depends upon the resolution of several technical, business, legal and cultural issues. Important technical challenges include:

  1. Development of an information infrastructure that enables manufacturers to advertise and exchange data on products and services sought and offered. After the formation of the enterprise, the partners should be able to use this infrastructure to share seemlessly engineering, manufacturing, and business information.
  2. Enhancement of design and planning methods and tools. Cooperative distributed design requires advancements in standards and goupware, as well as in design critiquing methods. In a distributed manufacturing environment the design should be critiqued and optimized with respect to the manufacturing capabilities of potential partners. Process and production planning should also account for the capabilities, capacity, availability and cost of the members of the virtual enterprise.
  3. Development of decision support systems for partner selection. These systems should be able to seek, find, assess, and select partners that are most suitable to meet the design and manufacturing requirements of a candidate product.

  
Table 1: Initial set of topics selected for awards under ARPA's agile manufacturing program.

These issues have been the subject of recent forums (see e.g. [104]), and are being addressed by several projects under way in both academia and industry. Table 1 lists twenty projects sponsored by ARPA under its Agile Manufacturing Program.

The AIMS (Agile Infrastructure for Manufacturing Systems) program, led by Lockheed focuses on issues 1 and 3 above, as well as business issues [60]. It is developing an information infrastucture for agile manufacturing that will use the Internet and will encompass business processes and information technology. Business processes under study include standardized trade agreements, sourcing prequalification, predefined protocols, and standard cost libraries. Information technology under development includes interfaces, protocols for manufacturing services, brokers and shop floor applications. The AIMS information infrastructure builds upon the MADE reference architecture, which uses emerging Internet standards for hypertext, multimedia, and information retrieval, and provides services that support engineering teams and electronic commerce.

The NIIIP (National Industrial Information Infrastructure Protocols) consortium also focuses on issue 1 above [103]. Its primary objective is to develop and demonstrate technology that enables distributed manufacturing. This technology comprises common communication protocols, an object technology base for system and application interoperability, specification and exchange of standard information models, and cooperative management of processes within virtual enterprises. NIIIP also recognizes the key role of the Internet to the implementation of industrial information infrastructure.

The Virtual Factories project at the University of Maryland is addressing issues 2 and 3 above [42]. It extends process and production planning to the aggregate, virtual enterprise level, which encompasses multiple partners; it is developing design critiquing tools that consider multiple domains and the capabilities of multiple partners; it studies the optimal selection of partners for the formation of a virtual enterprise based on multi-criteria trade-off analysis. The results of this work will be integrated into a toolkit for design, planning and partner selection in an agile manufacturing environment.

Discussion.

Distributed manufacturing, as described above, is radically different from current interactions between companies.Today, the process of establishing such interactions is long (in the order of months) and tedious. Potential suppliers are first qualified following certification of their capabilities, which in many cases involves site visits, and laborious record checking. Actual suppliers are selected from the small pool of qualified companies based on a largely ad hoc and highly subjective process, which is performed typically by the purchasing department with little input from the technical staff. The agreement for cooperation involves lengthy negotiations of terms and conditions. After establishment of such an agreement, communication between engineering and manufacturing personnel across companies is rare and slow; drawings are typically mailed or faxed, and, in many cases, must be redrawn in the supplier's CAD system; clarifications of drawings and instructions, as well as notification of engineering changes may take days or weeks. Delays in production are typically communicated after the fact and, thus, may cause severe planning problems in the supply chain. Finally, the true production quality of the supplier is rarely known, since non-conforming product is inspected out of the lots produced.

The advantages of distributed manufacturing are obvious when contrasting the current practice with the functionality of the virual enterprise. They include: i) Determining the most qualified supplier(s) to provide the products and services sought, in almost real time. ii) Developing better designs by taking into account the capabilities of potential partners. iii) Converting partner selection to a rigorous process that is centered on the product to be manufactured, not on purchasing concerns of little or no relevance. iv) Simplifying the partnering process and reducing the time required to negotiate terms and conditions. iv) Facilitating the exchange of engineering, manufacturing, and business information. v) Converting the partnership to a truly cooperative and open relationship that supports better planning, problem handling and resolution. These advantages are highlighted in the UltraComm scenario of [99], which describes the environment, practices and operation of a virtual company of the 21st century.

Virtual manufacturing (VM) may play a significant role in distributed manufacturing, since it may improve design critiquing and process planning. Consider the case of assessing the manufacturability of a design with respect to the capabilities of a potential partner. Given a model of the manufacturing process(es) to be employed and using the partner-specific process parameters, the feasibility of manufacture is first assessed. Subsequently the set-up, and run times may be determined, as well as the corresponding processing cost. This is combined with the material cost and the partner's overhead rate to determine the product's standard cost. The expected quality may also be assessed by comparing the design tolerances with the process variability predicted by the process model. Improved design critiquing through VM will result in better designs and more informed partner selection.

The VM environment shares basic similarities with the distributed design environment described in [42]. First it provides for intelligent decision making, that characterizes cooperative concurrent engineering. Secondly, it simplifies the sharing of knowledge through standard product, process and production models. Given these similarities, VM is expected to support distributed design if it provides protocols and computer aids for negotiation.

4.3 Generic VM Issues

4.3.1 Question 3.1

What are the relationships between VM, Virtual Prototyping, the Virtual Enterprise, etc.?

Definitions.

  1. Virtual Manufacturing (VM): is an integrated, synthetic manufacturing environment exercised to enhance all levels of decision and control in product and process design, process planning, production planning, and shop floor control [36].

  2. Virtual Enterprise (VE): A rapidly configured multi-disciplinary network of small, process-specific firms configured to meet a window of opportunity to design and produce a specific product [104,117].

  3. Virtual environment (VE), alternatively known as virtual reality (VR), or synthetic environment, is a collection of technologies, which offer the opportunity to integrate the human into a computing system [125].

  4. Virtual prototype: A computer-based simulation of systems and subsystems with a degree of functional realism comparable to a physical prototype. Virtual prototypes are used for test and evaluation of specific characteristics of a candidate design [50].

  5. Virtual prototyping (VP): The process of using a virtual prototype, in lieu of physical prototype, for test and evaluation of specific characteristics of a candidate design [50].

Background

It is well recognized that over 70% of the cost of a product is committed at the product's design stage [46]. From a product-life cycle viewpoint, virtual manufacturing provides design, process and production engineers with the ability to validate their designs, associated process plans, and operational plans with respect to technical feasibility and cost. This is done early in the design phase before committing to real production. The fundamental notion of VM is that it is a computer-based, simulated environment for product development that enables us to ``make it virtually'' before we ``make it for real'' [36,37]. Simulations over the life cycle of parts will provide accurate data that precludes the development of a design that is difficult, or impossible, to manufacture [31]. Thus, VM supplements the IPPD process since it provides a pathway for manufacturing knowledge to be migrated to early phases in the product life cycle [36,37].

The functional areas of the product life cycle in VM include the man-machine interface (visualization), product design, process development, prototyping, production, and shop floor control. In addition to projects in Table 1 funded by ARPA, NSF also supports many projects in different areas of virtual manufacturing product life cycle [82,23,83,132,94,53,29,93,118,87,95,7,71,141]

Both virtual reality and virtual prototyping are relevant to VM. Virtual reality may support the development of enhanced graphic user interfaces (GUIs) for VM, and, thus, enhance the integration of the human user into a computer system.

Virtual prototyping (VP) has been recently used in an effort to reduce the product development time and cost [78,122,125,50,27,136,4,3,1,2]. In VP, computer-aided design information is directly transformed to products/models without building physical prototypes to validate/optimize the design. If the virtual prototype were ``constructed'' using simulations of the planned production processes, then one could say that the virtual prototype was created using virtual manufacturing [36]. There are cases that virtual prototypes have advantages over physical prototypes [122]. However, some substantial work (for example, validating the virtual product computer models and establishing process models and material databases) needs to be done before VP may offer a viable alternative to physical prototyping [122,78].

Virtual enterprises focus on sharing of partner expertise, technologies, resources, and profits to achieve agility, i.e., ``the ability to thrive in a continuously changing, unpredictable environment'' [33,35,68,151,45]. Exchange of information among partners across the internet becomes a critical issue to the success of a virtual enterprise. The partners in such a virtual enterprise may share information about products, processes, and production. This information could be represented in the form of data, knowledge and/or models and may be distributed to different computing environments. The issues associated with the virtual enterprise have attracted considerable interest in both industry and academia. For example, the ARPA funded project, ``Agile Manufacturing Information Infrastructure (AMII),'' focuses on providing the required network tools and utilities to support manufacturing enterprises [88]. The NIIIP consortium will incorporate the standards that government and industry have developed for object technology, product data, and computer communications to create a seamless environment for the virtual enterprise [98,103]. There appears not to be strong interdependence between virtual enterprise and virtual manufacturing. In the area of information integration and sharing, the emphasis of VE is in the network while the emphasis of VM is in the product development. Furthermore, the focus of VM is the simulation of the product life-cycle in the design stage.

Discussion.

A comparison of the focus areas Virtual Manufacturing, Virtual Environment, Virtual Prototype, and Virtual Enterprise is shown in Tables 2 and 3.

  
Table 2: Comparison -- Product Life Cycle

  
Table 3: Comparison -- Information Integration

Virtual manufacturing focuses on all the activities of the product life cycle. Virtual environments will provide visualization technology for virtual manufacturing. The Virtual Prototype is an essential component in the virtual product life cycle. Therefore, the developments in the area of virtual prototyping will enhance the capabilities of virtual manufacturing. The virtual enterprise focuses on partner interactions, whereas virtual manufacturing focuses on the product/process development. However, the technologies developed in virtual enterprises will be useful to virtual manufacturing once the VM centers are distributed over networks (distributed design).

4.3.2 Question 4.2

What is the role of object oriented technology in VM?

Definition.

Virtual Manufacturing requires a robust infrustructure to capture the diverse information associated with manufacturing, and to exchange, incrementally build, and reuse information, knowledge and models.

Background.

Object oriented technology has been playing an important role in software engineering. The primary reason is its ability to deal with the complexity of information about systems. The object oriented approach has brought together the principles of abstraction, encapsulation, inheritance, modularity, and hierarchy.

An object oriented system can be described as a collection of objects and their relationships. The data in an object are hidden from other objects. The mechanism to share information between objects is by means of message passing. An object sends a message to another object to request the service of the object. To accomodate object interfaces, the Object Management Group (OMG) includes the Object Request Broker (ORB) which provides the mechanisms by which objects transparently make requests and receive responses [54].

As object-oriented applications become larger and are distributed over networks, the need to persistently store and share the objects among applications becomes more important. For example, a STEP database which is created by an application may need to be accessed by another application at a different location [44]. One of efforts of the National Industrial Information Infrastructure Protocols (NIIIP) Consortium has been toward defining a series of protocols that will enable STEP databases to be shared over a network based on the OMG's Common Object Request Broker Architecture (CORBA) [103]. Object-oriented STEP applications have also appeared in product/process models [55,18,16,17,57,76,74].

Finally, Object Oriented technology has been adopted in the simulation of manufacturing systems [5,9] and has been applied to the modeling and analysis of communication and control of manufacturing systems [9,89].

Discussion.

Object oriented technology provides a richer way to classify the components, and their relationships, in manufacturing systems. By using this approach, the characteristics of manufacturing systems can be identified and described by classes of objects and their relationships. Currently, for example, STEP databases generated from one application may not be able to be shared with another application. Methodologies will be required to develop a generic application framework, which defines a commonly accepted object classes, that will provide the foundation for the sharing of information/knowledge/models within a VM center or between VM centers.

By building a library of these classes, a product model, a process model, or a manufacturing system model can be described as a combination of some object classes and be stored in a database. The model itself will become part of the object classes library. By developing the schemas for classes of models, we can build agile models for Virtual Manufacturing. The important functionality of this library is the capability of reuse of existing models. A well-defined classification structure for characteristics of manufactruing systems is the essential to design of reusable models [89,120].

5 Conclusions

 

This section summarizes the findings of our background research in certain aspects of Virtual Manufacturing and identifies critical topics that, if addressed successfully, will contribute significantly to the realization of VM. The discussion focuses on three areas: VM and manufacturability; VM and virtual X; the VM environment. The section concludes with recommendations on using the World Wide Web site to support the research in these areas.

5.1 VM and Manufacturability

Virtual manufacturing, when mature, is expected to greatly support assessing the manufacturability of a candidate design and to provide accurate estimates for processing times, cycle times and costs (including inventory), as well as product quality. This is because VM will be able to model both the processes employed for the product's manufacture and the production process. By employing comprehensive models of manufacturing processes, VM will be able to accurately predict set-up times and run times, and, consequently, labor costs. Furthermore, if these process models are able to predict the variance of key product attributes, then process yields or the values of quality ratios (such as C pk) may be obtained by comparing the process capability with the corresponding design tolerances. On the other hand, modeling the production process will yield queue times, as well as Work-In-Process and finished goods inventory. Consequently, accurate estimates of overall cycle times and overall costs may be obtained. Tools that assess manufacturability by generating and evaluating manufacturing plans require more computing time than approaches that try to analyze the design directly, but they also offer more accurate results. As the cost of computing power continues to decrease, we anticipate that such approaches will become increasingly widespread.

The potential of VM to support manufacturabilty assessments and provide accurate cost, lead time and quality estimates is a major motivation for further research and development in this area. There are several advancements, however, that are needed to effectively support manufacturability assessments using virtual manufacturing. These include:

  1. Support for computer-aided conceptual design. Software tools should be developed to support the evolution of function and ideas prior to their realization as geometry (the latter being the domain of tradition CAD).
  2. Integration beyond single applications and single manufacturing domains. Integration must be enabled across the entire spectrum of design and manufacturing systems and over all partners in the larger enterprise. Such as, planning for manufacture using multiple processes, selection of optimal processes, integration of process planning with scheduling and shop floor control. Furthermore, it will be crucial in next-generation systems to seamlessly integrate conceptual-stage design analysis tools with some form of conceptual design system.
  3. The role of VM in supporting variant process planning appears to be rather limited, but VM will have a clear and critical role in supporting generative and hybrid approaches to process planning: it will enable such approaches to include more sophisticated process models, so that they may produce more realistic plan evaluations.

  4. It is necessary to develop methods for integrating product design and process planning with production planning and scheduling. By doing so, the designer will be able to determine early in the design stage the impact of design decisions in production planning and scheduling. Process planning defines the sequence of operations for the realization of a design and, therefore, provides the feasible subspace for both planning and scheduling. However, product design and process planning are time-independent, while production planning and scheduling are dynamic activities. This complicates the integration and presents a challenging topic for further research.
  5. Movement toward common representation for manufacturing resource information and common interfaces among manufacturing software systems. This is very much in line with the ``open architecture'' model which, in providing complete and free access to internal data structures and representations, can liberate manufacturing software systems from their rigidly defined ``black boxes'' and enable integration and innovation.
  6. Product information models need to be further developed. In addition to shape information, product models should be able to capture data that are directly relevant to manufacturing, such as tolerances (dimensional and geometric), and form features. Although tolerances are essential in evaluating the manufacturability of a design with respect to the capabilities of manufacturing processes, no existing CAD system has full tolerancing capabilities. In the features area there has been considerable difficulty in developing standard representations that may be used to provide a process view of the design and to support design by features. There is also considerable doubt among researchers whether such standard representations will be practical. In addition, despite extensive research, the practicality of feature extraction from boundary product representations is still limited. Further research is therefore necessary in the form features area to develop powerful methods for representing and reasoning about features, which are critical for both process selection and planning, and for manufacturability evaluations.

  7. Process models need to be integrated into concurrent engineering and Virtual Manufacturing systems. Although process modeling is a mature subject, it has yet to play a significant role in concurrent engineering practice. One of the major reasons for this is the lack of unified ways to deal with process models in an automated design environment. This also appears to be a major roadblock for virtual manufacturing; especially in assessing the feasibility of producing a design with a certain set of manufacturing processes and evaluating the ease of manufacture with these processes. Thus, novel representations are necessary to capture diverse process models (analytical, statistical and simulation-based models) and to provide unified interaction mechanisms with the virtual manufacturing environment.

  8. New information models are necessary to capture the capabilities and performance of production systems, and thus provide plant-specific information to the virtual manufacturing system for design evaluation.

5.2 VM and Virtual X

Virtual manufacturing may play a significant role in distributed manufacturing accomplished by virtual enterprises, since it may improve design critiquing and process planning. These improvements are expected to result in better designs and more informed partner selection. Furthermore, VM is expected to support distributed design, which is another aspect of virtual enterprises. For this purpose, however, VM needs to provide protocols and computer aids for negotiation. On the other hand, the technologies developed from virtual enterprise, such as information exchange protocols and standards, can benefit VM. Therefore, the technology progress in these areas should be closely observed and identified for use in VM.

Virtual Manufacturing may also provide an application environment for the technologies of virtual reality and virtual prototyping. Virtual environments may provide visualization technology for VM and the Virtual Prototyping may provide technology for making virtual prototypes, which is an essential component in the virtual product life cycle, for VM.

5.3 The VM Environment

In order to be able to ``manufacture in the computer,'' VM will provide a modeling and simulation environment from product design to production planning, and process optimization. Each of the three types of VM discussed in Section 1 (design-centered VM, production-centered VM, and control-centered VM) plays specific roles to accomplish its tasks. Models are the essential components that facilitate the performance of these tasks. The models for each type of VM are closely coupled. For example, a cost model used in design-centered VM can be an aggregated model derived from a more detailed cost model in production-centered VM. Since the development of these models is critical to the success of VM, a possible framework for the development of Virtual Manufacturing models is presented as shown in Figure 2.

  
Figure 2: Virtual Manufacturing Model Development Cycle [POSTSCRIPT IMAGE HERE]

The efforts to develop the VM environment should focus on the following:

  1. Development of generalized models, and an open system architecture.
  2. Development of a set of tools, such as model representation tools, model editors, and model management utilities.
  3. Identification of the characteristics of product/design/manufacturing and development of the required standards, modeling technologies, representation tools, and methodologies. Characterization of manufacturing may be accomplished by gathering data, developing models, etc. Object Oriented technology may provide a powerful representation and classification tools for manufacturing. The use of Object Oriented technology in virtual manufacturing needs to be seriously investigated.
  4. Development of ways to verify simulation models.
  5. Study of methodologies for storing/retrieving and dynamically modifying models. A new model can be generated from an old one by reusing of an existing model or part of a complex existing model, or by adding/deleting/modifying from an existing model. The new model needs also to be deposited into a model database.
  6. Establishment of a framework for developing object-oriented application schemas to provide a common communication platform for implementation of standards or protocols. CORBA has been becoming a standard for the communication between objects in industry. Applications of CORBA in virtual manufacturing need to be investigated.

5.4 VM World-Wide Web Repository

The large response to our email solicitation suggests that there is a great deal of interest in virtual manufacturing. For this reason, we feel it would be useful to maintain, on an ongoing basis, an internet repository of information on virtual manufacturing that could be made available to the global research community. This could be accomplished by taking the following steps:

As the information is archived, it would be widely available to the participants (and eventually the global research community). In this manner, the site could evolve into a virtual manufacturing testbed which would link to (and thus be leveraged from) other efforts in such directions, such as the National Testbed for Process Planning Research [123] being developed at the National Institute of Standards and Technology (NIST).

Such a web site could also leverage from our own nascent project [101] to build a manufacturing planning testbed for use in the ARPA/Rome Laboratory Planning Initiative (ARPI). We are expecting ARPA funding for this project in the near future. The principal investigators are Jim Hendler (hendler@cs.umd.edu) and Dana Nau (nau@cs.umd.edu), both of whom are University of Maryland faculty with joint appointments between the Institute for Systems Research and the Computer Science Department. The work will be carried out jointly with Steve Ray (ray@cme.nist.gov) at NIST.

A References

References

1
Achieving agile manufacturing in the automotive industry: Part I. Automotive Engineering, 102(11):19, November 1994.

2
Achieving agile manufacturing in the automotive industry: Part II. Automotive Engineering, 102(12):13, December 1994.

3
Changing the system. Computer Integrated Manufacturing and Engineering, pages 28--29, December/January 1995.

4
The very real benefits of virtual prototyping. Computer Integrated Manufacturing and Engineering, pages 23--26, December/January 1995.

5
S. Adiga and C. R. Glassey. Object-oriented simulation to support research in manufacturing systems. International Journal of Production Research, 29(12):2529--2542, December 1991.

6
David C. Anderson and Mark R. Henderson. Computer recognition and extraction of form features: A CAD/CAM link. Computers in Industry, 5:329--339, 1984.

7
Ruzena K. Bajcsy and Jonathan Smith. ANVIL: A national virtual laboratory. Technical Report NSF a9412913, Office of Research Administration, U. of Pennsylvania, 133 South 36th Street, Suite 300, Philadelphia PA 19104, July 1994.

8
R. Bakerjian, editor. Design for Manufacturability, volume 6 of Tool and Manufacturing Engineers Handbook. Society of Manufacturing Engineers, 1992.

9
D. A. Bodner, S. Narayanan, U. Sreekanth, T. Govindaraj, L. F. McGinnis, and C. M. Mitchell. Analysis of discrete manufacturing systems for developing object-oriented simulation models. 3rd Industrial Engineering Research Conference Proceedings, May 1994.

10
Roger W. Bolz. Production Processes: Their Influence on Design. The Penton Publishing Company, Cleveland, 1949.

11
G. Boothroyd and P. Dewhurst. Design for Assembly -A Designer's Handbook. Department of Mechanical Engineering, University of Massachusetts at Amherst, 1983.

12
J. Bralla, editor. Handbook of Product Design for Manufacturing. McGraw Hill, 1986.

13
J. Britanik and M. Marefat. Hierarchical plan merging with applications to process planning. In IJCAI-95, 1995.

14
S. Brooks and M. Wolf. An overview of Allied Signal's XCUT system. In Jami Shah, Martti Mantyla, and Dana Nau, editors, Advances in Feature Based Manufacturing, pages 399--422. Elsevier/North Holland, 1994.

15
CACI Products Company. SIMFACTORY II.5, 1991.

16
Arun Candadai, Supratik Champati, Jeffrey W. Herrmann, Ioannis Minis, and Vijaysundar Ramachandran. Information needs in agile manufacturing. International Computers in Engineering and Database Symposium Conference, Sept. 1994.

17
Arun Candadai, Jeffrey W. Herrmann, Ioannis Minis, and Vijaysundar Ramachandran. Product and process information models for microwave modules. ASME Winter Annual Meeting, Dec. 1994.

18
S. Chan. An object-oriented implementation of step. In ACM/SIGGRAPH Symposium on Solid Modeling Foundations and CAD/CAM Applications, 1991.

19
B. Chandrasekharan, Ashok K. Goel, and Yumi Iwasaki. Functional representation as design rationale. Computer, 26(1):48--56, 1993.

20
T. C. Chang and R. A. Wysk. An Introduction to Automated Process Planning Systems. Prentice-Hall, Englewood Cliffs, NJ, 1985.

21
T.C. Chang. Expert Process Planning Systems. Addison-Wesley, 1989.

22
Tien-Chien Chang. Expert Process Planning for Manufacturing. Addison-Wesley, 1990.

23
Jerome Chen. Virtual model interlocking. Technical Report NSF a9460368, Software Magic, 1505 S Purdue Avenue Suite 105, Los Angeles, CA 90025-3126, Jan. 1995.

24
S. H. Chuang and M. R. Henderson. Three-dimensional shape pattern recognition using vertex classification and the vertex-edge graph. Computer Aided Design, 22(6):377--387, June 1990.

25
S. H. Chuang and M. R. Henderson. Compound feature recognition by web grammar parsing. Research in Engineering Design, 2(3):147--158, 1991.

26
R. Clements. Quality Manager's Complete Guide to ISO 9000. Prentice Hall, Englewood Cliffs, N.J., 1993.

27
Will Cochran. Steroscopic images in aerospace design deployment. Computer Integrated Manufacturing and Engineering, pages 15--18, December/January 1995.

28
C. F. Cooms. Printed Circuits Handbook. McGraw Hill, 1979.

29
Oscar D. Crisalle and Haniph A. Latchman. Virtual control-laboratory for multidisciplinary engineering education. Technical Report NSF a9352523, U. of Florida, 219 Grinter Hall, Gainesville, FL 32611, September 1993.

30
D. Das, S. K. Gupta, and D. Nau. Reducing setup cost by automated generation of redesign suggestions. In Kosuke Ishii, editor, Proc. ASME Computers in Engineering Conference, pages 159--170, 1994. Best-paper award winner.

31
DDR&E. Defense science and technology strategy. Technical report, DDR&E, July 1992.

32
Defense Technical Information Center, Alexandria, VA. ICAM Manufacturing Cost/Design Guide, 1982.

33
R. Dove. Beginning the Agile Journey: A Guidebook. 1993.

34
T. J. Drodza. Machining. In Tool and Manufacturing Engineers Handbook, volume 1. SME, 1983.

35
Gary S. Vasilash (ed.). Manufacturing agility: The conference. Production, 105(4):58--63, 1993.

36
Lawrence Associates Inc. (ed.). Virtual Manufacturing User Workshop, August 1994. Technical Report.

37
Lawrence Associates Inc. (ed.). Virtual Manufacturing Technical Workshop, January 1995. Technical Report.

38
General Electric. Manufacturing Producibility Handbook. Manufacturing Services, Schenectady, NY, 1960.

39
H. A. ElMaraghy, K. F. Zhang, and H. Chu. A function-oriented modeler prototype. In P. J. Guichelaar, editor, Design for Manufacturability, ASME Winter Annual Meeting, pages 57--62. ASME, 1993.

40
H.A. ELMaraghy, K.F. Zhang, and H. Chu. A function-orinted modeler prototype. In ASME National Design Engineering Conference, volume DE-Vol. 52, Chicago, IL, March 1993.

41
A. S. Lazaro D. T. Engquist and D. B. Edwards. An intelligent design and manufacturability system for sheet-metal parts. Concurrent Engineering: Research and Application, 1(2):117--123, 1993.

42
Mark R. Cutkosky et al. Pact: An experiment in integrating concurrent engineering systems. Computer, January 1993.

43
Wozny et al. A unified representation to support evaluation of designs for manufacturability: Phase ii. Proceedings of The 1992 NSF Design and Manufacturing Systems Conference, 10(4):469--486, January 1992.

44
Elizabeth Fong et al. (Editors). X3/sparc/dbssg/oodbtg final report. Technical Report Final Report, NIST, Sept. 1991.

45
L. Scott Flaig. Alliances' future: The ``virtual enterprise:'' your new model for success. Electronic Business, 18(6):153--155, March 1992.

46
ManTech Task Force. DOD's manufacturing technology (ManTech) program. Technical report, ManTech, 1994.

47
Lowell W. Foster. GEO-METRICS II. Addison-Wesley, 1986.

48
W. A. Knight G. Boothroyd, P. Dewhurst. Selection of materials and processes for component parts. Proceedings of The 1992 NSF Design and Manufacturing Systems Conference, 10(4):255--263, January 1992.

49
Rajit Gadh, E. Levent Gursoz, Mark A. Hall, Friedrich B. Prinz, and Atul M.\ Sudhakar. Feature abstraction in a knowledge-based critique of designs. Manufacturing Review, 4(2):1--11, June 1991.

50
Albert B. Garcia, Jr. Robert P. Gocke, and Jr. Nelson P. Johnson. Virtual prototyping: Concept to production. Technical Report DSMC 1992-93, Defense Systems Management College, Fort Belvoir, VA 22060-5565, March 1994.

51
D. A. Gatenby and G. Foo. Designing for X: Key to competitive, profitable markets. AT&T Technical Journal, 69(3):2--13, 1990.

52
P. Gavankar and M. R. Henderson. Graph-based extraction of protrusions and depressions from boundary representations. Computer Aided Design, 22(7):442--450, September 1990.

53
Robert J. Graves. Agile electronics manufacturing research institute (AMRI). Technical Report NSF a9320955, Rensselaer Polytech Institute, 110 8th Street, Troy NY 12180-3522, April 1994.

54
Object Management Group. The common object request broker: Architecture and specifications. Object Management Group and X/Open, Dec. 1991.

55
P. Gu, Y. Zhang, and D. H. Norrie. Object-oriented product modelling for computer-integrated manufacturing. International Journal of Computer Integrated Manufacturing, 7(1):17--28, January 1994.

56
Jin-Kang Gui and Martti Mäntylä. Functional understanding of assembly modelling. Computer Aided Design, 26(6):435--451, 1994.

57
S. K. Gupta. Automated Manufacturability Analysis of Machined Parts. Ph.D. Dissertation, Mechanical Engineering Dept. University of Maryland, College Park, College Park, Maryland, 1994.

58
S. K. Gupta and D. S. Nau. A systematic approach for analyzing the manufacturability of machined parts. Computer Aided Design, 27(5), 1995. To appear.

59
S. K. Gupta, D. S. Nau, W. C. Regli, and G. Zhang. A methodology for systematic generation and evaluation of alternative operation plans. In Jami Shah, Martti Mantyla, and Dana Nau, editors, Advances in Feature Based Manufacturing, pages 161--184. Elsevier/North Holland, 1994.

60
R. Dove H. Park, J. M. Tenebaum. Agile infrastructure for manufacturing systems (aims). 1994.

61
Inyong Ham and Stephen C.-Y. Lu. Compute-aided process planning: The present and the future. Annals of the CIRP, 37(2):591, 1988.

62
G. Harhalakis, A. Kinsey, I. Minis, and H. Rathbun. Manufacturability evaluation of electronic products using group technology. In NSF design and manufacturing systems grantees conference, Charlotte, North Carolina, 1993.

63
C. C. Hayes. Using a manufacturing planner as a designer's tool: Identifying cost-critical areas of designs. In AAAI-95 Spring Symposium Series: Workshop on Integrated Planning Applications, pages 41--50, March 1995.

64
C. C. Hayes, S. Desa, and P. K. Wright. Using process planning knowledge to make design suggestions concurrently. In N. H. Chao and S. C. Y. Lu, editors, Concurrent Product and Process Design, ASME Winter Annual Meeting, pages 87--92. ASME, 1989.

65
C. C. Hayes and H. C. Sun. Plan-based generation of design suggestions. In R. Gadh, editor, Concurrent Product Design, ASME Winter Annual Meeting, pages 59--69, New York, 1994. Design Engineering Division, ASME.

66
R. Hedden. Cost Engineering in Printed Circuit Board Manufacturing. Marcel Dekker, 1987.

67
Mark R. Henderson and LeRoy E. Taylor. A meta-model for mechanical products based upon the mechanical design process. Research in Engineering Design, 5(4):140--160, 1993.

68
Peter D. Hilton and Geoffrey K. Gill. Achieving agility: Lessons from the leaders. Manufacturing Review, 7(2):172--179, June 1994.

69
D. J. Holt. Saturn: The manufacturing story. Automotive Engineering, 98, November 199o.

70
R. D. Holtz. Gt and capp cut work-in-progress time 80 Assembly Engineering, 21(7):16--19, July 1978.

71
Jiawei Hong. Direct rapid prototyping and process generation. Technical Report NSF a9461693, CIMplus Inc., 636 Broadway Suite 714, New York NY 10012-2623, February 1995.

72
Wynne Hsu, C. S. George Lee, and S. F. Su. Feedback approach to design for assembly by evaluation of assembly plan. Computer Aided Design, 25(7):395--410, July 1993.

73
I. Pandelidis I. Minis, E. B. Magrab. Improved methods for the prediction of chatter in turning: Part 1, 2, and 3. ASME Journal of Engineering for Industry, 115:12--35, 1990.

74
Industrial Technology Institute, http://www.iti.org. Industrial Technology Institute Home Page.

75
Kosuke Ishii. Modeling of concurrent engineering design. In Andrew Kusiak, editor, Concurrent Engineering: Automation, Tools and Techniques, pages 19--39. John Wiley & Sons, Inc., 1993.

76
ISO DIS 10303-11. Product Data Representation and Exchange - Part II: The EXPRESS Language Reference Manual, 1992.

77
M. Jakiela and P. Papalambros. Design and implementation of a prototype intelligent CAD system. ASME Journal of Mechanisms, Transmission, and Automation in Design, 111(2), June 1989.

78
David Japikse and Frank A. Olsofka. Agile engineering accelerates design. Mechanical Engineering, 115(11):60--62, 1993.

79
C.C. Jara-Almonte and S. Krishnamorrthi. Evaluation of the manufacturability of machined components using axiomatic design principles. In NSF design and manufacturing systems grantees conference, Austin, TX, 1991.

80
Ping Ji. A tree approach for tolerance charting. International Journal of Production Research, 31(5):1023--1033, 1993.

81
Subbarao Kambhampati, Mark Cutkosky, Jay Tenenbaum, and Soo Hong Lee. Integrating general purpose planners and specialized reasoners: Case study of a hybrid planning architecture. IEEE Transactions on Systems, Man and Cybernetics (Special Issue on Planning and Scheduling), 1992.

82
Homayoon Kazerooni. Small grants for exploratory research: Virtual design laboratory. Technical Report NSF a9222742, U of Cal Berleley, Berkeley, CA 94720, Nov. 1992.

83
Eun Sok Kim, Mehrdad G. Nejhad, Junku Yuh, and Stephen Y. Itoga. Rapid prototyping: Virtual rapid prototyping system for piezoelectric micro systems. Technical Report NSF a9420378, U. of Hawaii Manoa, 2540 Maile Way, Honolulu, HI 96822, Oct. 1994.

84
Y. S. Kim. Recognition of form features using convex decomposition. Computer Aided Design, 24(9):461--476, September 1992.

85
F. Kimura. Product and process modeling as a kernel for virtual manufacturing environment. CIRP Annals, 42:147--150, 1993.

86
Mark Klein. Capturing design rationale in concurrent engineering teams. Computer, 26(1):39--47, 1993.

87
Thomas J. LeBlanc, Randal C. Nelson, Michael L. Scott, Christopher M. Brown, and Dana H. Ballard. Rapid prototpying of parallel robot vision systems using virtual reality and systems simulation. Technical Report NSF a9401142, University of Rochester, Rochester, NY 14627, August 1994.

88
Nancy Lehrer and nlehrer@isx.com. What's current today is legacy tomorrow. ARPA 13 Initiative Home Page, http://isx.com/pub/I3.

89
Yi-Tzer Lin. Modeling and analysis of message reachability in distributed manufacturing systems. Ph.D. dissertation, School of Industrial and Systems Engineering, Georgia Institute of Technology, Aug. 1994.

90
C. H. Link. Capp--cam-i automated process planning system. In Proc. 13th Numerical Control Society Annual Meeting and Technical Conference, Cincinnati, March 1976.

91
Machinability Data Center. Machining Data Handbook. Metcut Research Associates, Cincinnati, Ohio, third edition, 1980.

92
M. Marefat and R. L. Kashyap. Geometric reasoning for recognition of three-dimensional object features. IEEE Transactions on Pattern Analysis and Machine Intelligence, 12(10):949--965, October 1990.

93
John D. Meyer. Agile manufacturing research institute (amri): National coalition for machine-tool technology. Technical Report NSF a9320944, U. of Ill Urbana - Champaign, 801 South Wright Street, Urbana, IL 61801, April 1994.

94
John J. Mills. Agile aerospace manufacturing research institute (AMRI). Technical Report NSF a9320949, U. of Texas Arlington, 800 South Cooper Street, Arlington TX 76010-1500, Arpil 1994.

95
Farrokh Mistree, David W. Rosen, Bert A. Bras, Ashok K. Goel, and Nelson Baker. Rapid prototyping: Virtual design, service and demanufacture studio. Technical Report NSF a9420405, GA Tech Research Corporation -- GIT, Administration Building, Atlanta GA 30332, October 1994.

96
R. O. Mittal, S. A. Irani, and E. A. Lehtihet. Tolerance control in the machining of discrete components. Journal of Manufacturing Systems, 9(3):233--246, 1990.

97
R. Nagel and R. Dove. 21st century manufacturing enterprise strategy, an industry-led view. Technical report, Iacocca Institute, Lehigh University, 1991.

98
National Industrial Information Infrastructure Protocols, http://www.niiip.org. NIIIP Consortium's Home Page.

99
D. Nau, M. Ball, S. Gupta, I. Minis, and G. Zhang. Design for manufacture by multi-enterprise partnerships: Current status and future directions. In ASME Winter Annual Meeting, 1994.

100
D. S. Nau. Automated process planning using hierarchical abstraction. TI Technical Journal, pages 39--46, Winter 1987. Award winner, Texas Instruments 1987 Call for Papers on AI for Industrial Automation.

101
D. S. Nau, W. C. Regli, and S. K. Gupta. AI planning versus manufacturing-operation planning: A case study. In IJCAI-95, 1995. to appear.

102
D. S. Nau, G. Zhang, and S. K. Gupta. Generation and evaluation of alternative operation sequences. In A. R. Thangaraj, A. Bagchi, M. Ajanappa, and D. K. Anand, editors, Quality Assurance through Integration of Manufacturing Processes and Systems, ASME Winter Annual Meeting, volume PED-Vol. 56, pages 93--108, November 1992.

103
NIIIP. NIIIP reference architecture concepts and guidelines. Technical Report NTR95-01, Jan. 1995.

104
NIST Advanced Technology Program and National Center for Manufacturing Science, Ann Arbor, Michigan. Workshop on Virtual Enterprise, March 1994.

105
Peter O'Grady, Robert E. Young, Arthur Greef, and Larry Smith. An advice system for concurrent engineering. International Journal of Computer Integrated Manufacturing, 4(2):63--70, 1991.

106
Jussi Opas and Martti Mäntylä. Feature-based part programming. In Jami Shah, Martti Mäntylä, and Dana Nau, editors, Advances in Feature-Based Manufacturing, pages 239--260. Elsevier/North Holland, 1994.

107
IGES/PDES Organization. The initial graphics exchange specification (iges). Technical Report Version 5.2, IGES/PDES Organization, 1993.

108
International Standards Organization. Configuration controlled design. Technical Report ISO 10303-203, International Standards Organization, 1993.

109
International Standards Organization. Integrated resources: Geometric and topological representation. Technical Report ISO 10303-42, International Standards Organization, 1993.

110
International Standards Organization. Product data representation exchange: Part 47, integrated generic resource: Tolerance. Technical Report Working Draft, ISO 10303-47, International Standards Organization, 1993.

111
International Standards Organization. Integrated resources: From features. Technical Report Working Draft, ISO 10303-48, International Standards Organization, 1994.

112
International Standards Organization. Electrical/electronics connectivity. Technical Report Working Draft, ISO 10303-103, International Standards Organization, 1995.

113
International Standards Organization. Electronic pc assembly, design and manufacturing. Technical Report Working Draft, ISO 10303-210, International Standards Organization, 1995.

114
International Standards Organization. Mechanical design using boundary/surface/wireframe representations. Technical Report Working Draft, ISO 10303-204/205/206, International Standards Organization, 1995.

115
International Standards Organization. Ship structures. Technical Report Working Draft, ISO 10303-218, International Standards Organization, 1995.

116
G. Pahl and W. Beitz. Engineering Design. Design Council, London, 1984.

117
H. Van Dyke Parunak. The virtual enterprise concept. In Workshop on Virtual Enterprise. NIST Advanced Technology Program, 1994.

118
Kerstin B. Pollack. Study of rapid prototyping facilities in the united states manufacturing research community. Technical Report NSF a8815774, National Academy of Sciences, 2101 Constitution Ave. NW, Washington, DC 20418, September 1988.

119
S. Prabhakar and M. R. Henderson. Automatic form-feature recognition using neural-network-based techniques on boundary representations of solid models. Computer Aided Design, 24(7):381--393, July 1992.

120
Ruben Prieto-Diaz and Peter Freeman. Classifying software for reusability. IEEE Software, pages 6--16, January 1987.

121
A. Alan B. Pritsker. Introduction to Simulation and SLAM II. John Wiley & Sons, 1986.

122
Michael Puttre. Virtual prototypes move alongside their physical counterparts. Mechanical Engineering, 114(8):59--61, 1992.

123
Steven R. Ray and Allison Barnard Feeney. A national testbed for process planning research. Technical Report NISTIR 5169, National Institute of Standards and Technology (NIST), April 1993.

124
William C. Regli, Satyandra K. Gupta, and Dana S. Nau. Extracting alternative machining features: An algorithmic approach. Research in Engineering Design, 1995. To appear.

125
Sandy Ressler. Applying virtual environments to manufacturing. Technical Report NISTIR 5343, U.S. Dept. of Commerce, National Institute of Standards and Technology, For sale by the National Technical Information Service, January 1994.

126
David Robertson, Karl T. Ulrich, and Marc Filerman. A structured design methodology and metadesigner: A system shell concept for computer aided creative design. In B. Ravani, editor, Advances in Design Automation, pages 308--313. ASME, 1989.

127
D. W. Rosen, R. J. Dixon, and C. Poli. Features and algorithms for tooling cost evaluation in injection molding and die casting. In Proc. ASME Computers in Engineering Conference, 1992.

128
David W. Rosen, John R. Dixon, Corrado Poli, and Xin Dong. Features and algorithms for tooling cost evaluation in injection molding and die casting. In Proceedings of the ASME International Computers in Engineering Conference, pages 1--8. ASME, 1992.

129
Hiroshi Sakurai and David C. Gossard. Recognizing shape features in solid models. IEEE Computer Graphics & Applications, September 1990.

130
G. Schaffer. Gt via automated process planning. American Machinist, pages 119--122, May 1980.

131
Reinhard Schiebeler and Klaus Ehrlenspiel. Reek: A knowledge-based system as an integrated design-assistant. In P. J. Guichelaar, editor, Design for Manufacturability, ASME Winter Annual Meeting, pages 69--74. ASME, 1993.

132
Terrence K. Schmoyer. Agile manufacturing enterprise forum (amef). Technical Report NSF a9320951, Lehigh University, Alumni Building 27, Bethlehem, PA 18015-3005, April 1994.

133
Michael Schulte, Christian Weber, and Rainer Stark. Functional features for design in mechanical engineering. Computers in Industry, 23(1):15--24, 1993.

134
Jami Shah, Martti Mantyla, and Dana Nau, editors. Advances in Feature Based Manufacturing. Elsevier/North Holland, 1994.

135
S. R. Shankar and D. G. Jansson. A generalized methodology for evaluating manufacturability. In W. G. Sullivan and H. R. Parsaei, editors, Concurrent Engineering, Contemporary Issues and Modern Design Tools, pages 248--263. Chapman and Hall, 1993.

136
Abdallah Shanti. Virtual robotics. Computer Integrated Manufacturing and Engineering, pages 68--69, December/January 1995.

137
Spatial Technology Inc., Boulder, CO. ACIS© Geometric Modeler, 1993. Version 1.4.1.

138
LeRoy E. Taylor. Meta-Physical Product Modelling. PhD thesis, Arizona State University, 1993.

139
TNO. Introduction to MIPLAN. Organization for Industrial Research, Inc., Waltham, MA, 1981.

140
H. E. Trucks. Designing for Economic Production. Society of Manufacturing Engineers, 1987.

141
Judy M. Vance. Research initiation award: A virtual environment for engineering design optimization. Technical Report NSF a9309663, Iowa State University, 213 Beardshear Hall, Ames IA 50011, July 1993.

142
J. H. Vandenbrande and A. A. G. Requicha. Spatial reasoning for the automatic recognition of machinable features in solid models. IEEE Transactions on Pattern Analysis and Machine Intelligence, 15(12):1269, December 1993.

143
T.E. Vollmann, W.L. Berry, and C. Whybark, D. Manufacturing planning and control systems. Business One Irwin, Homewood, Il, 1992.

144
Douglas L. Waco and Yong Se Kim. Geometric reasoning for machining features using convex decomposition. Computer Aided Design, 26(6):477--489, June 1994.

145
R. V. Welch and J. R. Dixon. Extending the iterative redesign model to configuration design: Sheet metal brackets as an example. In W. H. ElMaraghy, W. P. Seering, and D. G. Ullman, editors, Proceedings of the Design Theory and Methodology Conference, pages 81--88. ASME, 1989.

146
D. E. Whitney. Designing the design process. Research in Engineering Design, 2:3--13, 1990.

147
F.W. Wilson and P.D. Harvey. Manufacturing Planning and Estimating Handbook. McGraw Hill, 1963.

148
W. Winchell. Realistic Cost Estimating for Manufacturing. Society of Manufacturing Engineers, 1989.

149
N. Yannoulakis, S. B. Joshi, and R. A. Wysk. A manufacturability evaluation and improvement system. In L. A. Stauffer, editor, Design Theory and Methodology. ASME press, 1991.

150
N. J. Yannoulakis, S. B. Joshi, and R. A. Wysk. Quantitative measures of manufacturability for rotational parts. Journal of Engineering for Industry, 116(2):189, May 1994.

151
Mohamed A. Youssef. Agile manufacturing: A necessary condition for competing in global markets. Industrial Engineering, 24(12):18--20, December 1992.

152
M. C. Ziemke and Mary S. Spann. Concurrent engineering's roots in the world war II era. In W. G. Sullivan and H. R. Parsaei, editors, Concurrent Engineering, Contemporary Issues and Modern Design Tools, pages 24--41. Chapman and Hall, 1993.

B Brief Biographies of the Authors

 

B.1 Edward Lin

Edward Yi-Tzer Lin is a research engineer in the Institute for Systems Research at the University of Maryland. His research interests are in the areas of production/logistics/material handling systems, computer-integrated-manufacturing systems, factory automation, and object oriented technology. Dr. Lin received his undergradute degree in mechanical engineering and his M.S. in automatic control engineering, both from the Feng Chia University, Taiwan, and his another M.S. in operation research (1989) and his Ph.D. in Industrial and Systems Engineering (1994) from the Georgia Institute of Technology.

B.2 Ioannis Minis

Ioannis Minis is an assistant professor at the University of Maryland, in the Department of Mechianical and the Institute for Systems Research. His research interests are in the areas of production systems, distributed manufacturing, and machining dynamics and control. Dr. Minis received his undergraduate degree in mechanical engineering from the National Technical University of Athens, Greece (1982), his M.S. in mechanical engineering from Clarkson University (1983) and his Ph.D. in mechanical engineering from the University of Maryland. Dr. Minis is a recipient of the 1993 Earl E. Walker Outstanding Young Manufacturing Engineer Award of the Society of Manufacturing Engineers. He also received the best paper award in the area of Engineering Database Management: ``Use of PDES in Group Technology Applications for Electronics,'' at the 1992 ASME International Conference on Computers in Engineering.

B.3 Dana Nau

Dana S. Nau is a Professor at the University of Maryland, in the Department of Computer Science and the Institute for Systems Research. He is the leader of ISR's Virtual Factories project, and the co-leader of ISR Systems Integration research thrust. His research interests include AI planning and searching, and computer-integrated design and manufacturing. He received a B.S. in applied mathematics in 1974 from the University of Missouri---Rolla, and an A.M. (in 1976) and Ph.D. (in 1979) in Computer Science from Duke University, where he was an NSF graduate fellow. He has had summer and/or sabbatical appointments at IBM Research, NIST, the University of Rochester, and General Motors Research Laboratories. He has received an NSF Presidential Young Investigator award (1984-89), the ISR Outstanding Systems Engineering Faculty award (1993-94), a best-paper award at the ASME 1994 Computers in Engineering Conference, and various other awards. He has more than 150 technical publications; copies of recent papers are available at http://www.cs.umd.edu/~nau.

B.4 William Regli

William Regli holds an appointment as a Computer Scientist in the Manufacturing Engineering Laboratory of the National Institute of Standards and Technology (NIST). He is currently working in the Factory Automation Systems Division's Manufacturing Information Dissemination Technologies group, exploring the use of network information tools to support design, manufacturing, and process planning. This work is under the Systems Integration of Manufacturing Applications (SIMA) thrust of the High Performance Computing and Communications (HPCC) initiative. From 1996-1998 he will be a National Research Council Postdoctoral Fellow with NIST.

In addition, Mr. Regli is a Research Assistant at the University of Maryland in the Department of Computer Science and the Institute for Systems Research, where he is currently completing the Ph.D. in Computer Science. He recieved a B.S. in Mathematics from Saint Joseph's University in Philadelphia in 1989. His dissertation research involves solid modeling and its applications in computer-integrated manufacturing. He is a member Sigma Xi, ACM, AAAS, AAAI, and is the recipent of the Institute for Systems Research's 1995 George Harhalakis Outstanding Systems Engineering Graduate Student Award.

C Copy of Email Solicitation

 

We are doing a study of Virtual Manufacturing technologies.  Our
conclusions will appear in a report to the Air Force Mantech program.
We have the following goals:

- to assess what research and applications are relevant to key aspects
  of virtual manufacturing;

- to build an internet repository of virtual manufacturing information
  on the World-Wide Web;

- to identify gaps in these research and application efforts, and
  present our outlook for the future of virtual manufacturing
  technologies.

If any of your work is relevant to virtual manufacturing, then this is
an invitation to send us information about it, for possible inclusion
on the Web site and in the report.

At the end of this message is a list of 13 areas that are relevant to
our study.  If you are doing work on one of these areas, please send
email to the following address, before the end of March:

        virtual@frabjous.cs.umd.edu

In your email, include the following information:

- a 150- to 200-word abstract of your work and how it is relevant to
  the areas listed below;

- a list of relevant references;

- if possible, a URL for a world-wide-web or anonymous ftp site where
  interested parties can retrieve more detailed information about your
  work.

Also, please forward this message to anyone else whom you think might
be interested.

  Thanks!

        Dana S. Nau, nau@cs.umd.edu
        Computer Science Department and
        Institute for Systems Research
        University of Maryland

Here are the other members of the team that is doing this study:

        Thom Hodgson, North Carolina State University, hodgson@eos.ncsu.edu
        Hank Grant, University of Oklahoma, hgrant@mailhost.ecn.uoknor.edu
        Ioannis Minis, University of Maryland, minis@eng.umd.edu
        Radharamanan (Radha), Marquette University, 6233radharam@vms.csd.mu.edu


Here are the areas that are relevant for our study:
[ Authors' note: the following list of areas comes from [36,37]].
1.  VISUALIZATION:  The representation of information to the user in a way
that is meaningful and easily comprehensible.  In addition to graphical
user interfaces (GUIs) and virtual reality technologies, this technical
area includes information distillation, aggregation and autointerpretation.

2.  ENVIRONMENT CONSTRUCTION TECHNOLOGIES:  A computer based environment which
facilitates the construction and execution of VM systems.  The tools are used
to extract information, to create models supporting simulation, to properly
configure the virtual environment, to analyze the ``fit'' of the virtual
environment to the real production environment, to link real and virtual
processes, and to link to the manufacturing control systems.

3.  MODELING TECHNOLOGIES:  Since simulations are based on models, modeling
technologies are key technologies for VM.  Significant modeling issues are:
representation, representation languages, abstraction, federation, 
standardization, reuse, multi-use, and configuration control.

4.  REPRESENTATION:  The technologies, methods, semantics, grammars and
analytical constructs required to represent all of the types of information
associated with designing and manufacturing a product in such a way that the
information can be transparently shared between all software applications
that support the representation technologies, methods, semantics, etc.

5.  META-MODELING:  This area refers to modeling about modeling, in essence,
constructing, defining and developing models that accommodate inter-model
interaction.  The area involves standards and integration issues.

6.  INTEGRATING INFRASTRUCTURE & ARCHITECTURE:  The underlying infrastructure
(e.g. network, communications) that supports the ability to share models
and integrated product and process development across geographically 
distributed enterprises (e.g. global co-location).  The area also includes
creating a framework for the interoperation of all VM technologies.

7.  SIMULATION:  The ability to represent a physical system or environment
in a computer.  This area includes a wide range of computer software 
applications and, in the long term, links to real world systems that
enable simulation-based control.  Includes model optimization and validation.

8.  METHODOLOGY:  The methodology for developing, deploying and using VM
systems, including ``simulation-based reason.''  The latter refers to
``problems'' that are defined in such a way that ``simulation'' will generate
insights (i.e., alternatives, potential solutions, problem 
definition/refinement).  Problem solutions will likely require more than
just ``simulation''.  This methodology cannot be identical during the different
phases, however, it should be consistent across all phases.

9.  INTEGRATION OF LEGACY DATA:  This technical area primarily deals with
data and the many aspects of dealing data in general.  Also, corporate
culture and multiple platforms were identified.

10.  MANUFACTURING CHARACTERIZATION:  This ara involves the capture,
measurement and analysis of the variables that influence material
transformation during manufacturing.  It also involves the techniques
and methods for creating generic models of these processes based on actual
shop floor data.

11.  VERIFICATION, VALIDATION & MEASUREMENT:  For VM, this area refers to the
methodologies and tools to support the verification and validation (V&V) of
a VM system.  Making decisions on a VM ``simulation'' of manufacturing
demands a confidence that the impacts of those decisions on physical
manufacturing will be realized as predicted.  The methodologies and tools
are developed to provide the confidence.  Measurement is included in this
technical area because its central role in maintaining a mapping between
the physical and the virtual is necessary for the V&V methodologies.

12.  WORKFLOW:  The work of an organization follows a path called the
workflow.  This technical ara encompasses the capture, evaluation and 
continuous improvement of the processes that are associated with workflow.
The workflow area processes primarily involve information, whereas the
manufacturing characterization area primarily involves physical material
transformation processes.

13.  CROSS-FUNCTIONAL TRADES:  The essence is multi-discipline optimization
applied to large grain (specifically Life Cycle Cost disciplines) problems.
These trades will be general across organizations at a  high level, but will be
organization specific at a lower level as with factory floor operations, etc.
This has big technology transfer impacts.  Many people had a hard time dealing
with the specific labels of the underpinnings, however, they were adamant 
that it described what was really needed (e.g. requirement).  Figure 3-1
in the final report of the user's workshop (presented here as Figure 3-1)
provides the context of this issue.

D Responses to the Solicitation

 

The responses to the email solicitation of Section C appear in raw form here. These responses are also available in collated form at http://www.glue.umd.edu/~lin/vmproject.html

About this document ...

This document was generated using the LaTeX2HTML translator Version 95.1 (Fri Jan 20 1995) Copyright © 1993, 1994, Nikos Drakos, Computer Based Learning Unit, University of Leeds.

The command line arguments were:
latex2html -split 0 -no_navigation -show_section_numbers report.tex.

The translation was initiated by Generated by latex2html-95.1 on Wed May 17 15:14:41 EDT 1995


Generated by latex2html-95.1
Wed May 17 15:14:41 EDT 1995