A new solid modeling scheme is proposed and developed as the core of an integrated approach to the computer aided design and manufacture of mechanical parts. The benefits of this methodology, which considers the manufacturing process during the design phase, are discussed in the context of an integrated CAD/CAM system.

Colleges and universities are not meeting industry needs for graduates trained in the use or implementation of computer-aided design and manufacturing systems since few schools have experience in teaching CAD/CAM. Furthermore, many Bachelor's-level graduates are going directly into industry, rather than pursuing graduate degrees, thereby compounding this problem. Members of this panel represent both schools and industry and together attempt to outline approaches to developing more extensive CAD/CAM emphasis in education, including- the role colleges and universities can or should play in addressing industry needs in this area- user- and implementor-oriented CAD/CAM education- CAD/CAM education as an integral part of university engineering curricula- problems encountered in organizing and implementing CAD/CAM curriculum- universities' relationships with industry.It is our hope that panelists' experiences will help guide schools in structuring CAD/CAM programs to best insure that the United States remain at the forefront of technology and compete industrially world-wide.Frank PuhlVictor LangerDonald P. GreenbergMark S. ShepardHerb VoelckerUniversities' Relationships With IndustryFrank PuhlAmerican industry needs engineers who have been brought up with the realization that computers are intimately involved in all aspects of design, analysis, and manufacturing. Industry, of course, needs engineers who are already trained in fundamental CAD/CAM principles and in solving real world problems using computers, as well as those who have been taught to be productivity- and cost-conscious.To help meet these needs, we must now provide answers to several significant questions:What part should industry play in encouraging and sponsoring research and education in CAD/CAM?How can we resolve conflicts between industries' “trade secrets” and universities' “open research”?How can industry and educational institutions work together to increase U.S. productivity so as to regain our competitive edge?Lockheed Corporation and CADAM, Inc. are addressing these issues by supporting selected schools, installing CADAM systems, and providing fellowships for education and research at upperclass and graduate levels. In addition, we are involved in various joint research projects with universities, and we look optimistically toward continuing these projects.CADD CAM User EducationCADD CAM User EducationVictor LangerGeneral Electric Medical Systems Division in Milwaukee experienced a severe shortage of CADD CAM operators and encouraged MATC to develop a training program which started in 1980. A three-year NSF- CAUSE grant and a partial donation of a Computervision CADDS 3 (now CADD 4 Designer V six-station) system resulted in a program for upgrading employed designers and for two-year associate degree full-time students. The program enrolls 200 students per semester, with 70% continuing education students and 30% full-time students. Before students can effectively apply CADD CAM education, they must have drafting and design experience or at least a year of engineering training, plus ability to use spatial relationships and a course in descriptive geometry. With this background, students learn to create geometry for a 2-D drawing database and gain sufficient experience in the first course to be as productive as employees with six months' full-time experience. In the final semester or in the second advanced course, students create 3-D geometry and apply analytical computer capabilities for design, specializing for uses in mechanical, electrical, structural, architectural, and graphics arts applications. The defined geometry is also used in CAM for generating numerical control machining and flame-cutting paths, and for robotic control.Each course has been evaluated, follow-up studies have been completed, and an advisory committee has guided development. Employees have verified success and ease in transferring geometric skills to many different CADD CAM systems in the market. Beginning in the Fall of 1982, Apple microcomputers will be used to teach all 2-D computer graphics skills previously taught on the Computervision system, and CADD CAM education is now becoming available on an economical basis to all users.Some Problems in CAD EducationDonald GreenbergIt is imperative that universities educating the next generation of engineers introduce computer-aided design courses into their curricula. There are many obstacles in accomplishing this within a university structure. This presentation describes the facilities and operation at Cornell University and discusses the potential benefits and difficulties.CAD/CAM Education in an Engineering CurriculumMark S. ShepardToday there is a large industrial demand for engineering graduates that understand CAD/CAM techniques and computer graphics. Therefore, many colleges and universities already have or are planning to introduce computer graphics and CAD/CAM concepts into their curriculum. The major questions to be addressed in integrating these techniques into the curriculum include type and amount of hardware, development and maintenance of software and method of introduction into the curriculum.In 1977, RPI's school of engineering established the Center for Interactive Computer Graphics which is charged with integrating interactive computer graphics into the entire undergraduate engineering curriculum and providing a facility for graduate instruction and research. With heavy industrial support, the Center has also developed a research program in computer graphics and CAD/CAM. This presentation will discuss RPI's overall approach to integrating interactive computer graphics into the engineering curriculum.A Postgraduate Program in “Programmable Automation”Ari RequichaHerb Voelcker“Programmable Automation” designates the emerging body of knowledge surrounding CAD/CAM and industrial robotics. Graduate study in the field is aimed at (1) understanding the informational aspects of design and production in the discrete goods industries, and (2) developing new technologies for producing goods automatically with programmable, general-purpose tools. Some of the knowledge and techniques used in Programmable Automation are drawn from established fields (computer science, material science, control theory, ...), but the distinctive character of Programmable Automation is set mainly by the pervasive roles played by geometry and computation.A postgraduate program in Programmable Automation is being launched at the University of Rochester to train MS/level systems engineers for industry, and Ph.D-level researchers and teachers. The program's evolution reflects a “trickle-down” philosophy of education, wherein major new fields enter engineering education through on-going research; research begets seminars, seminars sometimes evolve into graduate courses, and graduate courses sometimes spawn undergraduate courses. (Put differently, the process starts with mature minds grappling with poorly understood concepts and ends with immature minds assimilating tightly codified concepts.)The Rochester program is sited in Electrical Engineering and draws heavily on the staff and facilities of the Production Automation Project; it also has strong links with Mechanical Engineering and Computer Science. The initial curriculum is based on two core courses in computational geometry, a graphics lab, and a systems seminar; these are supplemented with established courses in computer science, digital systems, finite-element analysis, control theory, and so forth. An NC Systems course and lab will be introduced a year hence. Plans for linking the program with Rochester's VLSI program, and for launching robotics research and teaching, are still in an embryonic stage.

A formal approach is proposed to the definition and the design of conceptual database diagrams to be used as conceptual schemata in a system featuring a multilevel schema architecture, and as an aid for the design of other forms of schemata. We consider E-R (entity-relationship) diagrams, and we introduce a new representation called CAZ -graphs. A rigorous connection is established between these diagrams and some formal constraints used to describe relationships in the framework of the relational data model. These include functional and multivalued dependencies of database relations. The basis for our schemata is a combined representation for two fundamental structures underlying every relation: the first defined by its minimal atomic decompositions, the second by its elementary functional dependencies. The interaction between these two structures is explored, and we show that, jointly, they can represent a wide spectrum of database relationships, of which the well-known one-to-one, one-to-many, and many-to-many associations constitute only a small subset. It is suggested that a main objective in conceptual schema design is to ensure a complete representation of these two structures. A procedure is presented to design schemata which obtain this objective while eliminating redundancy. A simple correspondence between the topological properties of these schemata and the structure of multivalued dependencies of the original relation is established. Various applications are discussed and a number of illustrative examples are given.

The purpose of this correspondence is to present an approach for verifying that explicitly stated integrity constraints are not violated by certain transactions. We utilize a relational model wherein constraints are given in a language based on the first-order predicate calculus. Transactions are written in terms of a Pascal-like host language with embedded first-order predicate calculus capabilities allowing queries and updates.

The purpose of this paper is to present a new approach to the conceptual design of relational databases based on the complete relatability conditions (CRCs). It is shown that current database design methodology based upon the elimination of anomalies is not adequate. In contradistinction, the CRCs are shown to provide a powerful criticism for decomposition. A decomposition algorithm is presented which (1) permits decomposition of complex relations into simple, well-defined primitives, (2) preserves all the original information, and (3) minimizes redundancy. The paper gives a complete derivation of the CRCs, beginning with a unified treatment of functional and multivalued dependencies, and introduces the concept of elementary functional dependencies and multiple elementary multivalued dependencies. Admissibility of covers and validation of results are also discussed, and it is shown how these concepts may be used to improve the design of 3NF schemata. Finally, a convenient graphical representation is proposed, and several examples are described in detail to illustrate the method.