The synthesis of four-bar mechanisms is a well-understood, classical design problem. The original systematic work in this field began in the late 1800s and continues to be an active area of research. Limitations to the classical theory of four-bar synthesis potentially limit its application to certain real-world problems by virtue of the small number of precision points and unspecified order. This paper presents a numerical technique for four-bar mechanism synthesis based on genetic algorithms that removes this limitation by relaxing the accuracy of the precision points.

A symbolic calculus for reasoning about process plans is proposed in this paper. The main focus of attention is the selection and sequencing of material removal operations for components in accordance with the design geometry. This is a central issue in automated process planning. The proposed symbolic calculus defines a computational formalism for symbolic manipulation of feature volumes, so that reasoning about volumetric removals can be treated in a logical manner by using well-defined procedures of algorithmic synthesis. This potentially encourages a more generic approach to the automation of outline and detailed process planning. The underlying philosophy is that a properly interpreted object topology upon the feature model allows the logical synthesis of volumetric removal sequences. The number of sequences is constrained by algorithms within the planning system that consider part geometry as expressed by features. This reduces the problem space associated with plan synthesis. Some of the geometrically viable sequences have the potential for further development to form viable machining removal sequences, or outline process plans.

This issue of AIEDAM is based on a workshop on Machine Learning in Design held at the 1994 Conference on Artificial Intelligence in Design, AID'94, (Gero & Sudweeks, 1994); the second of such workshops, with the first being held at AID'92 in 1992 (Gero, 1992). The first workshop also resulted in a special issue of AIEDAM (Maher et al., 1994).

This paper presents an approach to solving constraint satisfaction problems using Asynchronous Teams of autonomous agents (ATeams). The focus for the constraint satisfaction problem is derived from an effort to support spatial layout generation in a conceptual design framework. The constraint specification allows a high-level representation and manipulation of qualitative geometric information. We present a computational technique based on ATeams to instantiate solutions to the constraint satisfaction problem. The technique uses a search for a solution in numerical space. This permits us to handle both qualitative relationships and numerical constraints in a unified framework. We show that simple knowledge, about human spatial reasoning and about the nature of arithmetic operators can be hierarchically encapsulated and exploited efficiently in the search. An example illustrates the generality of the approach for conceptual design. We also present empirical studies that contrast the efficiency of ATeams with a search based on genetic algorithms. Based on these preliminary results, we argue that the ATeams approach elegantly handles arbitrary sets of constraints, is computationally efficient, and hence merits further investigation.

As designs exist to satisfy some purpose or function, knowledge of functionality is essential in a wide variety of designrelated activities, including generation and modification of designs, comparison, evaluation and selection of designs, and explanation, diagnosis or repair of designs. Functional modelling refers to a wide variety of approaches to model a design and its requirements from its functional aspects so as to allow reasoning about its functionality for various activities such as the above.

Function logic methods have been successfully used in Value Analysis (VA) and Value Engineering (VE) for several decades. This functional approach attempts to provide a common language for specialists in multiple domains. This paper describes an extension of function logic that assists in systematic identification of design functions, allocations, and their interrelations. Our approach identifies a three-level function/allocation/component information structure to represent the state of the design. We illustrate new types of links that exist between functions and the effect of these on the representation of the interrelated functions. These linkages provide new pathways for design information and function evaluation through allocation arithmetic and supported functions. A computational model of the conceptual design process is proposed based on the extended function logic design representation. An outline of the inputs, outputs and operations on form and function variables is given as a step prior to the synthesis process. We illustrate, by example, the process of translating functional representations across specialist domains. Finally, a computer-based aid to developing functional models is described.

This paper presents a novel formulation of the configuration-design problem that achieves the benefits of the concurrent engineering (CE) design paradigm. In CE, all design concerns (manufacturability, testability, etc.) are applied to an evolving design throughout the design cycle. CE identifies conflicts early on, avoids costly redesign, and leads to better products. Our formulation is based on a distributed dynamic interval constraint-satisfaction problem (DDICSP) model. Persistent catalog agents map onto DDICSP variables and constraint agents map onto DDICSP constraints. These agents use a set of operations and heuristics to navigate the design space to eliminate sets of designs until a solution is found. Experimental results show that an architecture where each catalog agent resides on a separate computer has performance advantages over nondistributed approaches.

In this paper the need for Intelligent Computer Aided Design (Int.CAD) to jointly support design and learning assistance is introduced. The paper focuses on presenting and exploring the possibility of realizing “learning” assistance in Int.CAD by introducing a new concept called Shared Learning. Shared Learning is proposed to empower CAD tools with more useful learning capabilities than that currently available and thereby provide a stronger interaction of learning between a designer and a computer. “Controlled” computational learning is proposed as a means whereby the Shared Learning concept can be realized. The viability of this new concept is explored by using a system called PERSPECT. PERSPECT is a preliminary numerical design tool aimed at supporting the effective utilization of numerical experiential knowledge in design. After a detailed discussion of PERSPECT's numerical design support, the paper presents the results of an evaluation that focuses on PERSPECT's implementation of “controlled” computational learning and ability to support a designer's need to learn. The paper then discusses PERSPECT's potential as a tool for supporting the Shared Learning concept by explaining how a designer and PERSPECT can jointly learn. There is still much work to be done before the full potential of Shared Learning can be realized. However, the authors do believe that the concept of Shared Learning may hold the key to truly empowering learning in Int.CAD.

This paper proposes and evaluates an artificial neural network based method of modeling the dynamic behavior of spatially distributed continuous engineering processes. The technique is applicable to situations where the differential equations governing the behavior of a system are nonlinear and poorly understood, such as is the case for frost-heave and thaw-settlement processes in soils. A description is first provided of a means of modeling the unknown component of governing differential equations. A range of levels of understanding of these equations is considered. A method of discretizing the resultant neural models of these equations is then illustrated, and the way in which these can be used to simulate the behavior of a process is described. The performance of the proposed neural network approach is then assessed in a series of experiments simulating the nonlinear thermal behavior of translucent solid materials. The system is proven capable of providing highly accurate simulations of system behavior sustained over many thousands of simulation time steps. The paper concludes with an identification of several ongoing areas of further development and application of the proposed tool.

A method is proposed for creative innovative design that is in concordance with the act of knowledge integration in learning; creative innovative design is defined as a guided (creative) process for arriving at an artifact that is socially valuable (practical and needed) and original (innovative). Within the context of models of reasoning, the process of design is interpreted and analyzed with a goal of extracting the stages at which it can be consciously improved by mindful control. A language is proposed for team-oriented intra- or interdisciplinary collaborative, as well as individual design that facilitates communication, mutual understanding, and makes explicit alternative nonverbal, nonquantitative thinking processes that otherwise may remain latent. The role of motivation in innovation is briefly discussed, as well as the role of artifact valuation in a societal context. Although not central to the present discussion, computer models of design are presented as an instance of design practice captured for computer-based design support. A brief discussion highlights the application and implications of the proposed method to education and research.

Numerical simulation of partial differential equations (PDEs) plays a crucial role in predicting the behavior-of physical systems and in modern engineering design. However, to produce reliable results with a PDE simulator, a human expert must typically expend considerable time and effort in setting up the simulation. Most of this effort is spent in generating the grid, the discretization of the spatial domain that the PDE simulator requires as input. To properly design a grid, the gridder must not only consider the characteristics of the spatial domain, but also the physics of the situation and the peculiarities of the numerical simulator. This article describes an intelligent gridder that is capable of analyzing the topology of the spatial domain and of predicting approximate physical behaviors based on the geometry of the spatial domain to automatically generate grids for computational fluid dynamics simulators. Typically, gridding programs are given a partitioning of the spatial domain to assist the gridder. Our gridder is capable of performing this partitioning. This enables the gridder to automatically grid spatial domains with a wide range of configurations.

The relative significance of conceptual design to basic design or detail design is widely recognized, due to its influential roles in determining the product's fundamental features and development costs. Although there are some general methodologies dealing with functions in design, virtually no commercial CAD systems can support functional design, in particular so-called synthetic phase of design. Supporting the synthetic phase of conceptual design is one of the crucial issues of CAD systems with function modeling capabilities. In this paper, we propose a computer tool called a Function-Behavior-State (FBS) Modeler to support functional design not only in the analytical phase but also in the synthetic phase. To do so, the functional decomposition knowledge and physical features in the knowledge base of the modeler, and a subsystem Qualitative Process Abduction System (QPAS) play crucial roles. Modeling scheme of function in relation with behavior and structure and design process for conceptual design in the FBS Modeler are described. The advantages of the FBS Modeler are demonstrated by presenting two examples; namely, an experiment in which designers used this tool and the design of functionally redundant machines, which is a new design methodology for highly reliable machines, as its application.

New methods of configuration analysis have recently emerged that are based on development trends characteristic of many technical systems. It has been found that though the development of any system aims to increase a combination of the performance, reliability and economy, actual design changes are frequently kept to a minimum to reduce the risk of failure. However, a strategy of risk reduction commits the designer to an existing configuration and an approved set of components and materials. Therefore, it is important to analyze the configurations, components, and materials of past designs so that good aspects may be reused and poor ones changed. A good configuration produces the required performance and reliability with maximum economy. These three evaluation criteria form the core of a configuration optimization tool called KATE, where known configurations are optimized producing a set of ranked trial solutions. The authors suggest that this solution set contains valuable design knowledge that can be reused. This paper briefly introduces a generic method of configuration evaluation and then describes the use of a self-organizing neural network, the Kohonen Feature Map, to analyze solution sets by performing an initial data reduction step, producing archetype solutions, and supporting qualitative clustering.

This paper describes decisions made during the development of an expert system for advising on problems that arise in the use of cutting fluids in engineering. It covers the problems of knowledge acquisition and of knowledge representation and of the relationship between them. The need for iterative prototyping is noted, and the choice between a database and a rule-based approach is discussed. The paper model and the machine model may not be isomorphic, although both are useful, albeit for different purposes.

This paper presents an approach for the case adaptation, especially case repairing, in a case-based process planning system: PROCASE, for machining of rotational parts. In PROCASE, a new process plan is generated by adapting an existing similar process plan from its case library. Case adaptation is a crucial issue in achieving an automated case-based process planning system. This is because, usually, an existing process plan cannot necessarily produce an exact identical part as of the desired part. Adaptation is essential to tailor this existing plan to generate a new process plan for the new part. The case adaptation in this paper comprises case modification, case simulation, and case repairing. The modifier uses the knowledge extracted from case library to edit the retrieved similar plan. The simulator plays an important role in verifying the adapted plan as well as in directing the plan repairing. The repairing rules are indexed by the error messages obtained from the simulation. With the proposed case adaptation, the system will have the capability to repair the erroneous plans to achieve an automated and intelligent process planning system. This paper will first briefly introduce the case representation and case retrieval in PROCASE. Then the rest of the paper is dedicated to the case adaptation.

This paper describes current research toward automating the redesign process. In redesign, a working design is altered to meet new problem specifications. This process is complicated by interactions between different parts of the design, and many researchers have addressed these issues. An overview is given of a large design tool under development, the Circuit Designer's Apprentice. This tool integrates various techniques for reengineering existing circuits so that they meet new circuit requirements. The primary focus of the paper is one particular technique being used to reengineer circuits when they cannot be transformed to meet the new problem requirements. In these cases, a design plan is automatically generated for the circuit, and then replayed to solve all or part of the new problem. This technique is based upon the derivational analogy approach to design reuse. Derivational Analogy is a machine learning algorithm in which a design plan is saved at the time of design so that it can be replayed on a new design problem. Because design plans were not saved for the circuits available to the Circuit Designer's Apprentice, an algorithm was developed that automatically reconstructs a design plan for any circuit. This algorithm, Reconstructive Derivational Analogy, is described in detail, including a quantitative analysis of the implementation of this algorithm.

This paper discusses the issues related to retrieval of cases in a case-based design system. The limitations of the conventional approaches are brought out, and this is followed by a description of a new approach to indexing and retrieval. This approach aims at capturing control knowledge required for retrieval from individual retrieval examples. Since this approach is based on past retrieval examples, we call it RBEX (Retrieval Based on EXamples). The proposed approach is implemented in a prototype system called CADREM. CADREM uses design methods used in solving past design problems to solve new problems. This is closer to the derivational analogy approach rather than the more common transformational analogy. CADREM has been tried in the domain of conceptual structural design of buildings. Sample outputs from CADREM are also included.

The ability to reason spatially is an important skill required for engineers, particularly in engineering design and construction. One aspect of spatial reasoning is visualizing and constructing three-dimensional (3D) solid objects from two-dimensional (2D) projections. To assist in teaching this to engineering students, an instructional software system is being developed at the University of Illinois. This instructional software system is comprised of the Visual Sweeper and the Visual Teacher. The Visual Sweeper is a geometric framework for solving missing view problems. In missing view problems, students create 3D solid objects from two 2D projections by applying operations inverse to orthographic projection. The Visual Teacher, which is the focus of this article, is an intelligent critiquing and tutoring module that gives feedback to the student regarding partial solutions to missing view problems. The Visual Teacher is comprised of a Recognizer and a Critiquer. The Recognizer identifies which solution solid the student's partial solution is closest to. Based on the solution solid and a student's partial solution, the Criti-quer gives critique and advice to the student. The Recognizer is based on an algorithm for bipartite graph matching, while the Critiquer uses a rule-based approach. This paper describes the Visual Teacher, gives examples of how it can be used, presents preliminary evaluation results, and discusses the system's assumptions and limitations.

In many creative design processes, cross-domain knowledge is required to inspire the new design result. Thus, in knowledge-based design, how we represent the cross-domain knowledge becomes a key issue. In this paper, we present a formalism for design knowledge representation. By analyzing function representation in different design domains, from graphic design and industrial design to architectural and engineering device designs, we find that although the focus of each kind of design is different, the function representation can be generalized into a small number of categories. This formalism can be used in an explorative model of design by analogy, where designs from different design domains are sources to help produce a new design.

This paper gives a theoretical foundation and implementation details of data and control structures in a diagnostic expert system with adaptive uncertainty management. The considered problem covers a class of expert systems and the application domain in which the operating personnel actively monitor and control the outcome of an industrial manufacturing process. Based on subjective ranking of the latent fault in the final output product, the operator queries the expert system for a recommended corrective action. Subsequent successful or unsuccessful corrective actions consistently modify confidence measures of all corrective recommendations from the same fault set, according to the embedded analytical model. The described approach is based on the relational data representation and procedural control structures, as opposed to a more traditional declarative, production-based system.