John J. Mills

Agile manufacturing research : accomplishments and opportunities

Agile manufacturing is a new expression that is used to represent the ability of a producer of goods and services to thrive in the face of continuous change. These changes can occur in markets, in technologies, in business relationships and in all facets of the business enterprise. This paper discusses the genesis of several of the Agile Manufacturing Research Institutes (AMRIs) and their on-going activities and results to date. A vision for agile manufacturing research is articulated and initial accomplishments identified. Additional research needs are also discussed.

A conceptual approach managing design resource

This paper proposes a conceptual approach for managing design resources which supports continuous product development and collaborative design. Most of the current product design exists in the form of CAD files that contain geometry-oriented product data. This data is not sufficient for reuse in the design of a new product. This paper proposes an enhanced product model which is understandable, customizable and reusable. A model enhancer (ME) helps to transfer the CAD model into this enhanced model which is a reusable design resource. The ME is implemented as a software component, so that its exposed methods can be reused at the programming level. We believe that this approach to design resource management (DRM) will benefit modern product development.

AeroWEB: An Information Infrastructure for the Supply Chain

Information flow up and down the supply chain, particularly for complex products such as automobiles, aircraft, and weapons systems is complex, difficult, fraught with errors and time consuming. In the defense business, for example, just the engineering trade-off process in the conceptual design phase can take up to a year before all the information requested is available for the design engineering team to compare alternatives (Schwart97). Some of the problems encountered include the inability of small suppliers far down the chain to access technical data created with CAD systems to which they do not have access, inaccurate or wrong versions of technical data when the supplier eventually obtains it, and difficulty in contacting people to request clarification. EDI and geometric modeling standards (e.g. STEP) and Email have started to address these problems, but they do not address all of them. Email, for instance often results in huge CAD files bogging down a suppliers email system (often using a dial up modem) only to have it discarded because it was not appropriate to their business. EDI/EC and STEP standards focus on standards for the exchange and sharing of data, not on their smooth and efficient movement up and down the chain.

Enhancing Knowledge and Skill Chains in Manufacturing and Engineering

This introductory paper to the volume explains the DIISM problem statement and applies principles of architecture descriptions for evolutionary systems (IEEE 1471–2000) to the information infrastructure for engineering and manufacturing. In our vision, knowledge and skill chains depend on infrastructure systems fulfilling missions in three kinds of environments: the socio-industrial domain of society and its production systems as a whole, the knowledge domain for a scientific discipline, and the sectorial domain, which includes the operational entities (companies, organisational units, engineers, workers) in engineering and manufacturing. The relationships between these different domains are captured in a domain paradigm. For companies, the original scope for infrastructure systems was the factory floor and the engineering office. Recently the scopes of external collaboration and of mansystem collaboration have gained importance. Within each of the four identified scopes a system can offer services to different operational levels: operations, development or engineering, and research. The dimensions of scope and service level are briefly explained in relation to the architecting of an infrastructure. Papers are grouped according to their contribution to an infrastructure scenario or to an infrastructure component.

A taxonomy of the product realization process environment

The realization of concepts into products is an extremely complex and poorly understood process. Companies and researchers are developing software aids and tools in a somewhat ad hoc manner. Differing claims are made for such tools, without there being much underlying understanding about what they are supposed to do. As part of a project to develop a strategy for applying computer-based tools to the process of realizing products, this paper proposes a taxonomy which encompasses most of the factors involved and defines the environments in which computer-based tools operate. Several other uses for this taxonomy are suggested, including identifying how the different types of computer-based tools support or replace various human activities or tasks; outlining a strategy for automating the process, as part of a strategy for applying the right tool for the task; and identifying where new tools might be developed.

The Dynamically Reconfigurable Assembly System

Abstract An essential element of World Class Manufacturing is the existence of small, focused factories in which a variety of products can be made. Unfortunately, classical approaches to assembly automation result in systems which are difficult to reconfigure; i.e., to reprogram, refixture and retool. The Dynamically Reconfigurable Assembly System (DRAS) is a project at the Automation & Robotics Research Institute to address the issues involved in introducing more flexibility into assembly systems by making them able to be dynamically reconfigured for new products. Both hardware and software standards and architectures are being developed to facilitate the reconfiguration of DRAS in minutes instead of the usual weeks. The requirements for such a system and the hardware and software architecture will be described.

Using contexts in managing product knowledge

The appropriate use of context is a powerful tool for managing complexity asexemplified by the realization of a modern product in a globally distributed environment. We show that computational contexts can be created from the properties of entities such as the company, the project, the user, his organizational role, and the product itself. Our paper briefly presents the architecture of a product knowledge environment that is based on computational contexts. It then discusses several possible uses of computational contexts in the product realization process including management of product knowledge, managing the product realization project and supporting users while they perform the wide variety of tasks necessary to design, develop and produce a modern product.

The Systems Integration Architecture: an Agile Information Infrastructure

The Systems Integration Architecture project is taking a fresh approach to integration with special emphasis on supporting Agile, Virtual enterprises and their unique need for reconfigurable, de-centralized information systems. SIA is based on a different conceptual model of integration — called the TAR model — one which allows higher level integration than the traditional common, neutral format, shared database. The basic concept is that all information processing consists of a series of transformations of data sets called “Aspects” into other “Aspects”. The transformations are effected by what are called Functional Transformation Agents which provide a single function and have a well defined interface structure which allows them to be integrated in a variety of ways. The Systems Integration Architecture provides three high level services which allow FTA’s to be defined, combined into diverse networks which provide transformations of aspects in a manner transparent to the user and executed under a variety of control algorithms in a heterogeneous, de-centralized environment. These are: an Executive--to provide business control services; a Librarian--to act as somewhat of a meta data dictionary, keeping track of FTA’s and their input and output Aspects; and a Network--which is based on the Common Object Request Broker Architecture to facilitate communication between objects on a heterogeneous network. An extended entity relationship diagram is provided and each of the objects in the system is described. A simple example of its use is provided.

The Component-oriented Approach towards Complex Product Development

This paper proposes a new approach towards complex product development. The approach is based on the software component mechanism. A software component is a reusable software package which can run across heterogeneous platforms. This paper presents a Component-based Open System Architecture, in which the design of a product component is implemented into a software component. The product component data and operation methods are exposed and can be reused by others. The Collaborative Product Representation Model is proposed to represent the product component in this environment. It is a neutral, understandable, customizable and reusable model with self-management capability. The Product Model Processor helps to define this complex model data.

A Reference Model for Information Infrastructures

There are a number of different approaches to information infrastructures, all of which purport to be an answer to the problem of integrating diverse heterogeneous systems. Frameworks and Reference Models are also used for similar purposes. Each has a slightly different focus and approach, yet all are essentially addressing the same problem. Several reference models have been proposed and information infrastructure systems are under development for shop floor control in the discrete manufacturing and semiconductor industries, for high level distributed interactive simulations, software engineering and for collaborative design and manufacturing. In this paper we review several of the current approaches and develop a reference architecture which is then used to categorize and compare these same approaches. Approaches and models reviewed include the ADAPTIVE Communication Environment (ACE), the Common Object Request Broker Architecture/Common Object Services Specification (CORBA/COSS), the Distributed Computing Environment (DCE), the High Level Architecture (HLA), the Simulation-Based Design (SBD) architecture, the National Industrial Information Infrastructure Protocols (NIIIP), Shipbuilding Information Infrastructure Project (SHIIP), and the Systems Integration Architecture (SIA). A brief description of the proposed reference architecture and of each system is given and the services it provides are compared with those in the reference architecture.