Manufacturing process planning is a difficult problem with a prohibitively large search space. It is normally tackled by decomposing goal objects into features, and then sequencing features to obtain a plan. This paper investigates an alternative approach. The capabilities of a manufacturing process are represented by a formal language of shape, in which sentences correspond to manufacturable objects. The language is interpreted to describe process plans corresponding to the shape generation, complete with cost estimates. A macro layer that describes single operations of the machine is implemented on top of the formal language. The space it describes is searched by the generative simulated annealing algorithm, a stochastic search technique based on simulated annealing. Plans that are close to the optimum are generated in reasonable time.

Few existing Computer Aided Design (CAD) systems provide assistance to designers in developing geometric concepts at the early design stages. Instead they require a high level of precision and detail suited to detail design. To support the early geometric design, a CAD system should provide utilities for the rapid capture and iterative development of vague geometric models. This paper presents a pilot system that is being developed based on such a vision. The system has adopted minimum commitment modelling and incremental refinement as the guiding principles. The representation of geometric configuration is based on a parametric and constraint-based geometric design model, and provides a uniform representation of the approximate and precise size and location parameters. A constraint-based mechanism has been developed for processing geometric information. The use of the system in assisting the development of a geometric configuration is also demonstrated. Finally, features and limitations of the system as well as relations to relevant works are discussed, and based on this a number of key research directions are established.

This paper defines, for use in design, rules for propagating “distribution constraints” through relationships such as algebraic or vector equations. Distribution constraints are predicate logic statements about the values that physical system parameters may assume. The propagation rules take into account “variation source causality”: information about when and how the values are assigned during the design, manufacturing, and operation of the system.

This paper examines the architecture of back-propagation neural networks used as approximators by addressing the interrelationship between the number of training pairs and the number of input, output, and hidden layer nodes required for a good approximation. It concentrates on nets with an input layer, one hidden layer, and one output layer. It shows that many of the currently proposed schemes for selecting network architecture for such nets are deficient. It demonstrates in numerous examples that overdetermined neural networks tend to give good approximations over a region of interest, while underdetermined networks give approximations which can satisfy the training pairs but may give poor approximations over that region of interest. A scheme is presented that adjusts the number of hidden layer nodes in a neural network so as to give an overdetermined approximation. The advantages and disadvantages of using multiple output nodes are discussed. Guidelines for selecting the number of output nodes are presented.

Design rationale is a topic that implies different things to different people. To some it implies argumentation and frameworks for argumentation. To others it implies the documentation of design, like that required for many types of industrial or government work. Still others describe design rationale as the capture and potential reuse of normal communication about design. These perspectives of design rationale use different representations, which influence their ability to capture and to retrieve and use information. We propose an integrated approach to design rationale where design communication is captured and, over time, incrementally structured into argumentation and other formalisms to enable improved retrieval and use of this information. Two systems, PHIDIAS and the Hyper-Object Substrate, are used to demonstrate: (1) how to capture and integrate a variety of design information, (2) how to support the structuring of unstructured information, and (3) how to use design information to actively support design.

The “sketch” drawn by a human designer represents a shape class of wider variability than can be captured by traditional CAD models; these typically work with parametrizations based on a nearly finished shape. Traditional Qualitative Reasoning is also unable to model this degree of ambiguity in shape. Cognitively, shapes are often represented in terms of an axial model. In defining 2D contours, such an axial representation is called the Medial Axis Transform or MAT. By perturbing the parameters of the MAT—length, link angle, and the node radius—one can define a shape class. Unlike the contour-to-MAT transform, which is well-known to be unstable, the MAT-to-contour process is an integrative process and is very stable. The variation in these parameters can be controlled by defining a suitable discretization over the parameter space. This leads to a broad class of similar shapes from which an optimized shape can be obtained for a given set of criteria. The optimizing criteria may involve the boundary description for each shape; the axial model is only used for generating the shape class. This Qualitative MAT model has been tested in several design optimization contexts, using Genetic Algorithms, and we show results for Automobile contours, IC engine parts, building profiles, etc.

Artificial neural networks are finding wide application to a variety of problems in civil engineering. This paper describes how artificial neural networks can be applied in the area of construction project control. A project control system capable of predicting and monitoring project performance (e.g., cost variance and schedule variance) based on observations made from the project environment is described. This project control system has five neural network modules that allow a project manager to automatically generate revised project plans at regular intervals during the progress of the project. These five modules are similar in design and implementation. Therefore, this paper will present the main issues involved in the development of one of these five neural network modules, that is, the module for identifying schedule variance. A description of a graphical user interface integrating the neural network modules developed with project management software, and a discussion on the power and limitations of the overall system conclude the paper.

During the course of the design of a complex artifact many thousands of objects will be created that will refer to many different types of information. Thus, an effective design support system must be able to store various types of data and allow easy navigation of the resulting extensive network of objects. In addition, the quality of the design records (both the record of the design artifact and rationale) is increased by the quality and quantity of information which the designer is able to record. In this paper we describe two new developments to Kbds a design support system for chemical process design, which enable easier navigation of the design network, and a fuller representation of the design process. In addition, we show how these extensions may be used together to improve the quality of the design information recorded—both for the evolution of the design artifact and the supporting rationale. The first development enables the designer to record a variety of complementary types of document within the process design history, fulfilling the task of improving the design information recorded. The second enhancement to Kbds eases the navigation of the design history by categorizing design objects according to a user-defined set of keywords. The categorization of design objects is carried out semiautomatically using, in part, the various design representations (documents) enabled through the first extension and enables rapid navigation of the design network. Finally, a method of checking the consistency of design rationale structures using keywords is also described. Thus, various representations may be used to generate keywords that in turn may be used to improve the quality of design rationale records. An example showing this situation is described.

This paper presents a review of CBD and its application to building design in particular. Case-based design is the application of case-based reasoning to the design process. Design maps well to case-based reasoning because designers use parts of previous design solutions in developing new design solutions. This paper identifies problems of case representation, retrieval, adaptation, presentation, and case-based maintenance along with creativity, legal, and ethical issues that need to be addressed by CBD systems. It provides a comprehensive review of CBD systems developed for building design and provides a detailed comparison of the CBD systems reviewed.

The geometric dimensioning and tolerancing (GD&T) specifications of a design are directly associated with its performance and functional requirements. They also govern the manufacturing and quality control processes needed to achieve those requirements. This paper reviews recent work in geometric tolerance representation and reasoning and presents a generic and uniform graph-based representation scheme, called the Tolerance Network, to represent GD&T specifications across a part or assembly. The network can accommodate GD&T specifications related to the function, behavior, manufacturing, and inspection requirements embedded in design specifications and supports the use of different types of tolerances. The network also accommodates common design practices such as the specification of overconstrained features and parts. The necessary properties of such a network are discussed that allow under- and overconstrained design specifications to be detected and analyzed.

The following erratum is to correct an omission of a table during the publication phase of the article entitled: “User interface for specification language for case-based mechanical design” by Abhay Dandekar, Ibrahim Zeid, and Theodore Bardasz (AI EDAM, 11(1), p. 18).

The table referred to reads:

The Editor-In-Chief and Publisher regret the inadvertent mistake.

This paper proposes a new method for dealing with geometrical layout constraints. Geometrical layout constraints are classified into three classes of dimensional, regional, and interference constraints. Dimensional constraints are handled by using an existing methodology. A method is proposed to translate the other two classes of constraints into dimensional constraints. Thus, it is possible to uniformly deal with all of those geometrical layout constraints. The method is twofold. First, it converts regional, interference constraints into a set of simple inequalities. Then each inequality is solved by a geometric gadget, which is a structured set of dimensional constraints. A prototype system is developed and applied to some layout design examples.