Chris Tong

A Framework for Circuit Design

COMPCON, 1984

Models of innovative design, reasoning about physical systems, and reasoning about geometry

Academic Press eBooks, 1992

Part 1 Models of innovative design: introduction, C. Tong and D. Sriram automated reuse of design... more Part 1 Models of innovative design: introduction, C. Tong and D. Sriram automated reuse of design plans in BOGART ARGO - an analogical reasoning system for solving design problems, M. Huhns and R. Acosta retrieval strategies in a case-based design system, K. Sycara and D. Navinchandra case-based design - a task analysis, A. Goel and B. Chandrasekaran function sharing in mechanical design, K. Ulrich and W. Seering ALADIN - an innovative materials design system, M. Rychener, et al the ADAM design planning engine, D. Knapp and A. Parker using exploratory design to cope with problem complexity, C. Tong. Part 2 Reasoning about physical systems: temporal qualitative analysis - explaining how physical systems work, B. Williams. Part 3 Reasoning about geometry: studies of heuristic knowledge-based approaches for automated configuration generation and innovation, G. Nevill a case-based approach to the design of mechanical linkages, G. Kramer and H. Barrow mechanism comparison and classification for design, L. Koskowicz how and why to get an AI or CAD/CAM system to look at your objects, L. Leff and D. Yun.

Working Notes from the 1992 AAAI Workshop on Automating Software Design. Theme: Domain Specific Software Design

In recent years, there has been an increase in research and development effort aimed at the produ... more In recent years, there has been an increase in research and development effort aimed at the production of domain-specific software design systems-knowledge-intensive systems that aid in the design of software for specific classes of problems in science, engineering, telecommunications, manufacturing, business, education, and other areas. Despite substantive progress in developing general-purpose software design techniques, the application of these techniques to large, real-world software design tasks has proven di_cult. As a result, there is a growing realization that specialpurpose, domain-specific techniques will play a critical role in moving research into practice. When restricted to a specific domain, software design systems can avail themselves of additional sources of knowledge and constraints that simplify the overall design process. The goal of this workshop is to identify different architectural approaches to building domainspecific software ge%ign systems and to explore issues unique to domain-speclfic (vs. generalpurpose) software design. Some general issues that cut across the particular software design domain include: Knowledge representation, acquisition, and maintenance: In building a domain-specific software design system, decisions must-be made about what domain knowledge isnecessary to support the design task and what formalism to use for the representation. In addition, knowledge-intensive design systems cannot be deployed without seriously addressing the extra burden that comes along with acquiring an-cT_ta_u_nga significant body of domain knowledge. Specialized software design techniques¢_ By restricting both the domain and the class of software design tasks to be addressed by a sTstem, it becomes poss_le to utilize specialized design techniques that may simplify and.speed up the overall design process. _ _' _ User interaction J_ user interface{The typical end-user of a domain-specific software design system is an application specialist, not a progr_er with special analytic skills. As a result, domain-specific systems need to pay Special attention to the interaction b_een the user and the system. C.) ..... We hope that you find the workshop and the papers in this collection both stimulating and informative.

Download

Design representation and models of routine design

Academic Press eBooks, 1992

Part 1 Design representation: representation and control for the evolution of VLSI designs - an o... more Part 1 Design representation: representation and control for the evolution of VLSI designs - an object-orientated approach, J. Gabbe and P. Subrahmanyam tools and techniques for conceptual design, D. Serrano and D. Gossard automobile transmission design as a constraint satisfaction problem, B. Nadel and J. Lin design compilers and the labelled interval calculus, D. Boettner and A. Ward knowledge representation for design creativity, M. Flowers and M. Dyer. Part 2 Models of routine design: investigating routine design problem solving, D. Brown and B. Chandrasekaran design as top-down refinement plus constraint propagation, L. Steinberg a knowledge-based framework for design, S. Mittal and A. Araya BioSep designer - a process synthesizer for bioseparations, C. Siletti and G. Stephanopoulos VT - an Expert elevator designer that uses knowledge-based backtracking, S. Marcus et al a design process model, F. Brewer and D. Gajski WRIGHT - a constraint-based spatial layout system, C. Baykan and M. Fox designer - a knowledge-based graphic design assistant, L. Weitzman.

Knowledge-based circuit design

Google, Inc. (search), Subscribe (Full Service), Register (Limited Service, Free), Login. Search:... more

Toward an engineering science of knowledge-based design

Artificial Intelligence in Engineering, Jul 1, 1987

In this paper, we present a framework for organizing, evaluating, and developing knowledgebased m... more In this paper, we present a framework for organizing, evaluating, and developing knowledgebased models of the design process. We argue that evaluation of a design process model can be carried out from three usefully distinguished perspectives: the knowledge it embodies; the functionality of the design process, from a problem-solving viewpoint; and the implementation of the design process as an actual program. This paper focuses on the first two perspectives. We systematically introduce a set of basic functional components, and show how existing approaches or systems can be viewed as configurations of these components, in which domain knowledge has been incorporated. As we lay out this framework, we illustrate it in a simple way by using it to describe knowledge-based house floorplanners. We then complete our presentation by analysing a more complex knowledge-based system (DONTE) that designs circuits.

format_quoteConstraint reasoning introduces explicit representations of interactions in design, enhancing decision-making efficacy.format_quote

Download

Knowledge acquisition, commercial systems, and integrated environments

Academic Press eBooks, 1992

Contributors. Preface. Introduction. Part IV: Knowledge Acquisition. Part VII: Commercial Applica... more

Artificial intelligence in engineering design (Volume III): knowledge acquisition, commercial systems, and integrated environments

Academic Press Professional, Inc. eBooks, 1992

Very few expert system applications have been distributed widely beyond the boundaries of the org... more Very few expert system applications have been distributed widely beyond the boundaries of the organizations within which they were developed. Instead, ex pert systems typically address problems for which local expertise dominates the methods for solutions to problems, and thus these expert systems represent idiosyncratic solutions that are applicable only within the organization that developed them. The obvious, though not easily attainable, solution to this problem is to provide to the end user of such an expert system the ability to cus tomize its knowledge base. In this paper, we present the Expert Cost and Manufacturability Guide (ECMG) from this perspective. ECM G is an expert system designed to provide mechanical engineers with first-order manufacturing cost estimates and manufacturability feedback very early in the design process, during preliminary design. We describe the architecture of ECMG, particularly those aspects of its design that accommodate the need for customizability. This is followed by a description of the expert systems design methodology we employed to permit us to construct a customizable expert system application.

Download

Artificial intelligence in engineering design (Volume I): design representation and models of routine design

Academic Press Professional, Inc. eBooks, 1992

Very few expert system applications have been distributed widely beyond the boundaries of the org... more Very few expert system applications have been distributed widely beyond the boundaries of the organizations within which they were developed. Instead, ex pert systems typically address problems for which local expertise dominates the methods for solutions to problems, and thus these expert systems represent idiosyncratic solutions that are applicable only within the organization that developed them. The obvious, though not easily attainable, solution to this problem is to provide to the end user of such an expert system the ability to cus tomize its knowledge base. In this paper, we present the Expert Cost and Manufacturability Guide (ECMG) from this perspective. ECM G is an expert system designed to provide mechanical engineers with first-order manufacturing cost estimates and manufacturability feedback very early in the design process, during preliminary design. We describe the architecture of ECMG, particularly those aspects of its design that accommodate the need for customizability. This is followed by a description of the expert systems design methodology we employed to permit us to construct a customizable expert system application.

format_quoteECMG is a generic expert system providing cost estimates and manufacturability feedback for mechanical designs in early design stages.format_quote

Download

Knowledge Compilation: A Symposium

IEEE Expert, 1991

ABSTRACT The question of why some Wall Street AI applications are successful while others fail is... more

Automated Synthesis of Constrained Generators

International Joint Conference on Artificial Intelligence, 1989

Knowledge compilation is an emerging research area that focuses on "compiling" a problem solver's... more Knowledge compilation is an emerging research area that focuses on "compiling" a problem solver's inefficient, explicit knowledge representation into more efficient, implicit forms. This paper presents a technique that transforms a declarative problem description (specifying the problem but not how to solve it) into a reasonably efficient, generate-and-test problem solver. Our technique performs constraint incorporation, modifying the parameter generators so they only generate values that satisfy the problem constraints. Successful constraint incorporation depends upon choosing the right solution representation (i.e., the set of parameters). Having expressed a constraint in terms of a particular set of parameters, incorporation fails if the constraint is not factorable into constraints on the individual parameter generators. RICK, a Refinement-based constraint Incorporator for Compiling Knowledge, is a prototype program that compiles a problem specification into a problem solver using least commitment, topdown refinement to achieve constraint incorporation. RICK refines an abstract solution representation to avoid premature commitment to representations that hinder constraint incorporation. RICK is able to incorporate local constraints that constrain relatively small portions of the entire solution. We have tested these ideas by having RICK automatically construct a house floor planning problem solver.

Download

Reformulating Constraints for Compilability and Efficiency

KBSDE is a knowledge compiler that uses a classification-based approach to map solution constrain... more KBSDE is a knowledge compiler that uses a classification-based approach to map solution constraints in a task specification onto particular search algorithm components that will be responsible for satisfying those constraints (e.g., local constraints are incorporated in generators; global constraints are incorporated in either testers or hillclimbing patchers). Associated with each type of search algorithm component is a subcompiler that specializes in mapping constraints into components of that type. Each of these subcompilers in turn uses a classification-based approach, matching a constraint passed to it against one of several schemas, and applying a compilation technique associated with that schema. While much progress has occurred in our research since we first laid out our classification-based approach [Ton91], we focus in this paper on our reformulation research. Two important reformulation issues that arise out of the choice of a schema-based approach are: (1) compilability-...

format_quoteRICK generates constrained generators that ensure local constraints are satisfied by adjusting range values or pruning unnecessary instances.format_quote

Download

Reformulating Constraints for Cornpilability and Efflciency

KBSDE is a knowledge compiler that usee a classification-based approach to map solution constrain... more KBSDE is a knowledge compiler that usee a classification-based approach to map solution constraints in a taek specification onto particular search algorithm components that will be responsible for satisfying those constraints ‘-,y., local constraints are incorporated in generat: 14’ global constraints are incorporated in either .ba*ters or hillclimbing patchen). Associated with each type of search algorithm component is a subcompiler that specialisea in mapping constraints into components of that type. Each of these subcompilem in turn una e classification-baaed approach, matching a constraint p d to it against one of several schemes, .and applying a compilation technique associated with that schema. While much progene h a occurred in our research since we first laid out our classification-bad a p p r o d [TonSl], we focus in this paper on our reformulation research. Two important reformula. tion issuer that arise out of the choice of a schemabased approach are: Compilobility. Can e...

Stability of Dynamic Systems: A Constructive Approach

Automating the Construction of Patchers That Satisfy Global Constraints

Download

Automated Synthesis of Constrained Generators

Knowledge compilation is an emerging research area that focuses on "compiling" a problem solver's... more Knowledge compilation is an emerging research area that focuses on "compiling" a problem solver's inefficient, explicit knowledge representation into more efficient, implicit forms. This paper presents a technique that transforms a declarative problem description (specifying the problem but not how to solve it) into a reasonably efficient, generate-and-test problem solver. Our technique performs constraint incorporation, modifying the parameter generators so they only generate values that satisfy the problem constraints. Successful constraint incorporation depends upon choosing the right solution representation (i.e., the set of parameters). Having expressed a constraint in terms of a particular set of parameters, incorporation fails if the constraint is not factorable into constraints on the individual parameter generators. RICK, a Refinement-based constraint Incorporator for Compiling Knowledge, is a prototype program that compiles a problem specification into a problem solver using least commitment, topdown refinement to achieve constraint incorporation. RICK refines an abstract solution representation to avoid premature commitment to representations that hinder constraint incorporation. RICK is able to incorporate local constraints that constrain relatively small portions of the entire solution. We have tested these ideas by having RICK automatically construct a house floor planning problem solver.

Download

Use of Feedback to Control Redesign

Using Exploratory Design to Cope with Design Problem Complexity

Artificial Intelligence in Engineering Design, 1992

Automatic Construction of a Hierarchical Generate-and-Test Algorithm

International Conference on Machine Learning, 1989

Publisher Summary This chapter describes research on automating the construction of such hierarch... more Publisher Summary This chapter describes research on automating the construction of such hierarchical solvers. Hierarchical problem solvers prune out large chunks of the search space by first constructing an abstract solution and then refining this into a concrete solution to the given problem. The chapter presents a procedure called HiT that builds a hierarchical Generate and Test (G&T) problem solver by constructing an abstract search space and problem space, related to the corresponding spaces of the given problem domain through an abstraction function. The resulting system is guaranteed to be complete and satisfies Tenenberg's upward solution property. HiT builds a new abstract CSP domain on top of the provided (base level) problem domain. The hierarchical system constructed by HiT achieves improved performance by reducing the cost of solving some of the tests of the base level problem. HiT creates an abstract space by clustering the base search space objects.

Toward Automated Rational Reconstruction: A Case Study

International Conference on Machine Learning, 1989

Publisher Summary This chapter reviews a sequence of rational reconstructions of a particular des... more Publisher Summary This chapter reviews a sequence of rational reconstructions of a particular design/learning system called CPS/SCALE that characterize that system's development. CPS is a design system for implementing Boolean functions as gate networks, while SCALE is a learning system that adds refinement rules to CPS's knowledge base. Knowledge-based systems can be represented so their development over time can be expressed as a rational reconstruction process CPS is a search-based design program that performs decomposition, implementation, patching, and backtracking. It contains a knowledge base of implementation rules that map Boolean functions onto gate network implementations. CPS performs (recursive) decomposition of the Boolean function design problem and direct implementation of primitive subproblems, accompanied by patching and chronological backtracking. CPS also contains a knowledge base of patching rules that map gate networks into functionally equivalent gate networks. CPS is the problem solver of a simplified, LEX-like architecture whereas SCALE carries out the functions of critic, generalizer, and problem generator. CPS/SCALE began its life as a perturbation of the DONTE circuit design system.

Uploads

Papers by Chris Tong

Log In