The isiCAD dimensions

    The key goal of isiCAD-2004 International Workshop is to present the approaches and results in research and technology that improve intelligence of market-oriented solutions and products in PLM (CAD/CAM/CAE/PDM), computer graphics, virtual reality, simulation, computer games, multimedia, and other adjacent domains. There will be a special emphasis on development and application of constraint-based techniques and geometrical solving. The workshop is generally aimed at outlining relationship between two dimensions: the above-mentioned domains and, on the other hand, the methods that are soundly expected to support and improve intelligence for those domains. These methods include constraint-based solving (including geometric constraint solving, interval constraints, and methods on finite domains), rule-based reasoning, optimization methods, and some others.

    Bringing together the domains in question and relevant methods for improving the domain's underlying intelligence seem to be fruitful and should enable future establishing of the isiCAD community.

    With the intention of facilitating a discussion before, during, and after isiCAD-2004, the organizing committee will outline building a matrix of "domains* methods" that are relevant to the scope of interest for isiCAD-2004 and for isiCAD community. At present, the matrix is open to many and varying ideas presented by world-class experts. Through the contributions of these experts, the isiCAD matrix will be defined to be of great value to attendees of the workshop, current and future, as well as the PLM and graphics industry in general.

    We begin with publication of a short overview related to the PLM intelligence.


    Product Life-cycle Management (PLM):
    Intelligence and mathware

    The meaning and future development of the field of computerized design are related to making CAD/CAM systems "intelligent"able to "understand" the intentions of designers.

    Michel Theron, president of the Mithec consulting firm, noted in 1997 that "in more than 25 years that industrial CAD systems have been available, with rare exceptions they offered only drawing tools and accelerators for geometric calculations". Their users' main goal is the ability to modify a design easily and quickly without having to redesign other components and assemblies. An equally important problem is related to adding intelligence to CAD/CAM systems, enabling them to "understand" the intentions of designers. Raymond Kurland, president of another consulting firm, Technicom, at the same time said that future design technologies will be entirely based on models that rely on properties encapsulating intelligence.

    In the six years that have passed, "intelligence" of industrial CAD systems improved dramatically. Moreover, the field itself has changed: there is currently a huge demand for PLM systems (where PLM stands for Product Life-cycle Management). PLM is the generic name referring to a class of software applications used in the creation of a product (including market studies, product engineering, and manufacturing plans), organization of its sales and consumer service. A clearly visible recent trend is to develop and market PLM packages made up entirely by solutions from a single, large developer (like PTC, EDS, IBM/Dassault Systemes, SAP, Oracle, etc.).

    The market of PLM products is steadily expanding: different estimates forecast 20% of annual growth for the coming years. Therefore, investment in the development of PLM solutions is growing as well. The leading suppliers of PLM solutions, such as Dassault Systemes, the developer of well-known industrial systems like CATIA/DELMIA/ENOVIA/SmarTeam, are actively adding intelligence to their products. These intelligent components are often built by smaller companies that grew out of research teams (such as D-Cubed and Telekinesys in Great Britain or Ilog in France) and then purchased by the international giants like Dassault Systemes or IBM.

      Intelligent PLM Tools: What Are They?

      In our discussions of intelligent components, we mean first of all the ability to handle a declarative statement of a problem (in design, planning, etc.) based on the fundamental concept of constraint. A constraint is a requirement to a product or plan, specified by the user in a standard way. For example, in conceptual design the engineering requirements have the form of equations and inequalities that related its geometric, physical, and economic parameters. These equations can be of a complex nature: they can be expressed by algebraic formulas or defined by functions without analytical representation. A user is working in terms of a drawing in which the design problem is described in terms of flat geometric primitives such as point, segment, arc, spline, etc., and can specify both logical constraints on its elements (when geometric elements are incident, tangent, parallel, etc.) and dimension constraints (distances and angles). In the case of three-dimensional detailed design, similar constraints are specified on 3D primitives. When designing a mechanism, assembly constraints can be specified. These apply to the dynamic simulation of the product's behavior in accordance with the laws of physics (specified as constraints too). Finally, the planning constraints determine the sequence of tasks, resource allocation priorities, durations, etc. Thus, almost any field of PLM has to solve constraint satisfaction problems. The methods used for solution of these problems combine algorithms from computational mathematics, graph theory, and artificial intelligence; some of the algorithms are quite nontrivial. State-of-the-art industrial PLM solutions rely on a foundation that remains unseen by the end user, a foundation that extensively uses mathematics and knowledge processing techniques. PLM is one of those hi-tech areas in the software market where it is appropriate to distinguish between software and mathware.

      Due to explosive growth of capabilities of modern computers, industrial problems can now benefit from sophisticated solution methods that previously had only been studied theoretically or applied only in pure research contexts and required the use of supercomputers. The development of new methods and optimization of the existing ones create the main challenges facing the present-day researchers in this field. The exchange of views and results between researchers and the industry within isiCAD-2004 will enhance the development and practical introduction of new computational and reasoning methods for PLM-solutions.

    General information about PLM can be found in the following Web pages:

    as well as in the following papers:

    • V.Koshelev and V.Molochnik, "What is PLM?"(In Russian), SAPR i Graphika (CAD & Graphics), 10, October 2003, pp.34-37,
    • PLM Market is Growing and Developing, (In Russian), CAD/CAM/CAE Observer, N2 (11), 2003, c.4-8.

Powered with SIB3