Publications & Presentations
Larry L. Johnson

Organizations often resist standards in order to protect their proprietary information and processes.  Protecting proprietary information is essential. However, enterprises often have a misplaced idea of what information is truly proprietary to their business. The protection of information has cost associated with it. While it is important to protect intellectual property, it is also important not to protect that which is part of the broad state of the art. Furthermore, some information increases dramatically in value when shared in a collaborative environment. This article explores the concept of a misplaced sense of what is proprietary, the business case for separating proprietary and nonproprietary information, and the value of standardizing the latter.

Keynote Presentation delivered to the iViP-RoaM Workshop. Addresses the application of standards and OMG-MDA to preserve investment in business process applications in a changing execution infrastructure

[Excerpt from article's introduction]

Simulation is the act of predicting product performance prior to a product’s manufacture. Most often, analysts, those doing CAE (Computer-Aided Engineering), perform simulation, and in too many cases only as an adjunct – not as an integral part – of the product development process. Analysts often consider CAE “outside the process,” because of the difficulty and time involved in obtaining accurate inputs.

PDM (Product Data Management) would seem to be a natural complement to CAE, by providing a repository of data that is accessed by CAE. Historically, though, that has not been the case. The CAE community generally still views incorporation of PDM as an added burden, with benefits yet unknown.

Despite the issues inherent in integrating CAE and PDM, a subset of “Simulation Data Management,” the benefits include enabling the access, verification, and control of CAE input and output data. Simulation Data Management is the next frontier in making simulation such an integral part of the product development process that it will lead and dictate the overall design (“form follows function”).

The presentation addresses aspects of interoperability technologies such as OMG, STEP and XML using the vehicle of ISO's RM/ODP (Reference Model / On-line Distributed Processing)

A concept of "Architectural Aspects" is discussed with a short treatment on interoperability for various aspects. A "Blend and Build" approach to evolving a systems architecture is introduced that enables the migration of legacy systems over time.

Traditional Product Information Management (PIM) focuses on "Form" and "Fit". Vast amounts of data are generated and consumed in the course of analyzing products. These data allow us to manage the "Function" of the product as well as its "Form and Fit". Effective management of this data can allow analysis to precede detailed design, minimizing the false steps of designers and managing trade-off studies.

This paper contains two different perspectives that arrive at the same conclusion. "PDM" is an outmoded term and is actually counterproductive. One perspective is oriented toward simply communicating more concretely, and not forgetting what we are about as IT providers. The other perspective is oriented toward a better way to design and build better solutions, leveraging modern technology. Over time, some people have included such a broad range of applications under the banner of "PDM" that the term has lost whatever meaning it once had. In stating, "We don't need no stinkin' PDM!", we mean the term and the mind set it tends to encourage, as well as the confusion it creates for the users and IT providers. Existing and emerging technologies are examined in the context of designing and providing IT Infrastructure services in the terminology of the business.

An Infrastructure Services Architecture is presented for the Enterprise Integration Architecture framework of the Profile for Enterprise Integration. The architecture enables flexible reengineering of the business process, and the construction of product-centered virtual enterprises. The architecture accommodates distributed, heterogeneous systems, providing protocols of service invocation and cooperation. Emphasis is placed on following the current trends and directions and industry at large, and on taking advantage of current and emerging commercial-off-the-shelf software.

This paper was based in a great part on the "Profile for Enterprise Integration", CALS Industry Working Group, June 1994 (below).

Abstract Coming Soon

An Infrastructure Services Architecture is presented for the Enterprise Integration Architecture framework of the Profile for Enterprise Integration. The architecture enables flexible re-engineering of the business process, and the construction of product-centered virtual enterprises. The architecture accommodates distributed, heterogeneous systems, providing protocols of service invocation and cooperation. Emphasis is placed on following the current trends and directions and industry at large, and on taking advantage of current and emerging commercial-off-the-shelf software.

The Rapid Response Manufacturing Consortium was formed to develop, implement, and validate technologies required to fundamentally change the development and planning of mechanical components from a serial process to an integrated set of concurrent processes. To facilitate this goal a reference architecture will be developed to guide the evolution of the systems of member companies toward a common target architecture tailored for each company's specific needs. This paper documents the context and concept of operation of the reference architecture, per se, within the program, and introduces candidate technical approaches.

Abstract Unavailable

A data flow model of a distributed, loosely-coupled, time-sharing computer network is discussed. This model has been used to initiate an evolving implementation of an automated software distribution/installation and configuration management system capable of handling software distribution and management for a system involving many machines. The model defines a generic software product, independent of its function, in the context computer services shop.

We have constructed a dataflow model of a generic software product in the context of a general tine-sharing environment. In this model, points of control for each product are defined for each "system-event". One of those events is the deletion of a process (specifically, LOGOUT). The current implementation of the activation of our logout driver (which calls per-product logout functions) is weak. An unprivileged user can exit the system in a manner which will bypass the activation of the logout driver. By the definition of our model, logout operations arc not optional. A weak activation of the logout driver is intolerable. This study was undertaken to determine the feasibility of implementing a robust logout driver activation.

Transition probabilities for the collision induced dissociation reaction, A + BC -> A + B + C, are calculated for a harmonic oscillator model using a semiclassical approximation. One of the unique features of this investigation is the appearance of quantum oscillations in the dissociation probability as a function of collision energy. The number of oscillations is correlated to the number of bound states of the BC molecule.