Collaborative Prototyping and Product Development on the Web M i s
by user
Comments
Transcript
Collaborative Prototyping and Product Development on the Web M i s
M Crriittiiccaall Miissssiioonn C EEnntteerrpprriissee SSyysstteem mss SSyym m 22000000 mppoossiiuum LLeevveerraaggiinngg EEnntteerrpprriissee IInnffoorrm maattiioonn PPoorrttaallss Collaborative Prototyping and Product Development on the Web Jeffrey P. Stavash Senior Member Lockheed Martin ATL (856) 338-4031 [email protected] John Welsh Manager Lockheed Martin ATL (856) 338-4223 [email protected] Bipin Chadha Principal Member Lockheed Martin ATL (856) 338-3865 [email protected] Abstract Collaborative environments for product development have become the new design paradigm for today’s engineering organizations. Collaboration permits greater information sharing, concurrent engineering, virtual prototyping and testing, and total quality management. The results are an increase in product quality and a decrease in lifecycle cost. One key ingredient missing in collaborations is the ability to capture the dynamic causal relationships between components of the system being designed. How the components relate to each other is either captured at a very basic level (physical interfaces), or is left for the designers to understand and keep in mind as the design evolves. The COTS based collaboration environment has been extended to capture the dynamic causal behavior and is used to develop a tradeoff analysis tool where designers can evaluate multiple configurations and options to achieve higher quality and cost effective designs. This paper describes the foundational technologies that are required to implement a virtual team environment for collaborative product development. It presents the Collaborative Enterprise Environment (CEE) that Lockheed Martin Advanced Technology Laboratories (ATL) developed for the Air Force Research Laboratories (AFRL). It discusses how issues of heterogeneous applications, access control, security, and integration across disparate tool sets were resolved to permit AFRL to collaborate with Lockheed Martin and industry on technology development programs over the complete lifecycle. Keywords Collaborative enterprise environment, dual-use application program, product data management, simulation-based acquisition, smart enterprise model, system dynamics, virtual team, virtual prototyping. 1 . Introduction Many of today’s large organizations, such as Lockheed Martin, are relying heavily on collaborative engineering and virtual prototyping approaches to product design and development. Increasing product complexity is forcing multiple organizations to share in the design and development of a product. The continuing evolution of the Internet and new information management technologies are also making this possible. Process modeling and workflow tools identify bottlenecks and inefficiencies of design processes. Modeling and simulation tools prove out a product’s design before it’s produced. Cost estimating tools permit cost analyses to be performed throughout the lifecycle. Requirements management tools ensure that the product is compliant with all customer requirements. And Product Data Management (PDM) tools allow product data to be configuration managed and shared with the distributed design team, including the customer and the supply chain vendors. A collaborative environment for virtual prototyping and product development must therefore provide the proper infrastructure for using these tools, as well as support the product’s development over the total lifecycle. Examples of lifecycle phases include system definition, detailed design, manufacturing, and operational support. Each lifecycle phase will involve different users who are dealing with a subset of the virtual prototype and potentially adding information to it. In a virtual team environment, the users will be from different organizations in geographically distinct locations. The collaborative environment must therefore permit the virtual prototype to be shared among the virtual team, provide multiple domain-specific views of the product data to the proper users, and manage product configurations and baselines. 2 . Infrastructure Components 2.1. Internet Architecture The Internet is highly scalable. It derives these properties via simple mechanisms, such as hyperlinking information across web servers. Nodes on networks can be clients and/or servers, and each can perform the role that is necessary for operation. The ability of enterprise agents to hyperlink to each other across the network and to travel across clients and servers will provide essentially unlimited scalability for collaborative environments and virtual prototyping. 2.2. Product Model An important infrastructure component for collaborative environments is a product model. The product model should capture the product’s structure and behavior. Since the virtual prototype will be shared among the virtual team, the product model should store only the shared data. The product model should configuration manage the product data over the entire lifecycle and permit users to recall earlier versions of the data whenever necessary. The product model should be modular and component based so as the virtual prototype evolves, so does the product model. 2.3. Business Systems Business systems are the tools an organization uses throughout the lifecycle. Examples of business systems include CAD/CAM tools, requirements tools, PDM tools, Enterprise Resource Planning (ERP) tools, cost estimating tools, legacy and heritage tools, and project management tools. These business systems should be web-enabled or provide an Application Programming Interface (API) that allows easy integration without a lot of customer software development. 3 . System Dynamics Information Technologies are trending toward common systems, tools, and processes. While that is a trend away from disjointed systems, it is unrealistic to assume that there will ever be a single process or system that satisfies every need in an organization with multiple businesses. Figure 1 shows diminishing returns as the degree of commonality increases beyond a limit. Figure 2 shows that hidden or ignored qualitative factors increase costs while easily quantifiable and visible metrics decrease costs. In this example, maintenance, deployment, and integration represent quantifiable costs, while coordination and changes and implementation delays represent qualitative factors that cause significant cost penalties in overall solutions. Most approaches ignore qualitative factors that become important as commonality approaches 100 percent. Existing architectures for collaborative environments and virtual prototyping only account for technical aspects and ignore cultural and organization aspects [1]. Better architectures are needed to address this dichotomy among business needs and commercial, off-the-shelf (COTS) capabilities. The federated approach [2] to the architecture is more suited to current business trends: It provides a low-risk alternative to existing technology investment strategies and it enables organizations to bring new technologies in a modular fashion as opposed to the big bang approach. A federated architecture enables programs to implement a total-systems view and to organize their supply chains at a global level. 4 . Collaborative Enterprise Environment The goal of the Air Force’s Collaborative Enterprise Environment (CEE) dual-use application program (DUAP) was to provide the capability for AFRL to collaborate with industry and academia on technology development programs. CEE needed to support simulation-based acquisition (SBA), virtual prototyping and product development in AFRL’s core technology areas of space vehicles, air vehicles, information technology, conventional weapons, directed energy weapons, materials and manufacturing, sensors, propulsion, and human effectiveness [3]. Page 2 of 6 Disjointed Architecture Federated Architecture Centralized Architecture Cost of coordination, inflexibility, suboptimization (qualitative, hidden) Cost ¥ ¥ ¥ ¥ Common Systems Common Tools Common Processes Common Best Practices Cost of deployment, maintenance, integration (quantitative, visible) 0% 100% Degree of Commonality Figure 1. Federated architecture provides an optimal solution. Cost of Deployment - + Cost of Maintenance Degree of Commonality Cost + + Cost of Change + + Cost of Coordination Implementation Delay Business Fit - + + ≈ Inflexibility - + B1 + Cost of Integration - + + Quantitative Cost + Qualitative Cost + Cost of SubOptimal Solution Figure 2. Architecture trade-off model. Air Force goals for improved collaboration included: • Team interaction through electronic modeling and data interchange during early design phases • Improved capability for teams to conceptualize, develop, mature, and transition new concepts • Improved insight into lifecycle concerns early in the development process • Early validation of system concepts on virtual test ranges • Maximizing synergy among various AF laboratory disciplines 4.1. CEE Design Process A majority of AFRL’s design processes are distributed modeling and simulation tasks, which involve multiple USAF organizations and legacy systems. These tasks often produce large amounts of design data that needs to be managed throughout the product’s lifecycle. Figure 3 shows a typical CEE design process as a workflow. The Technology Alternative Assessment node can have multiple subflows attached to it. The number of subflows depends on the number of Page 3 of 6 alternative assessments needed to determine which approach produces the optimal solution. These subflows are executed concurrently by different CEE users. The Lethality Analysis process is common to all subflows (Figure 3) and consists of executing the SAC (Suppressor/ALARM Context) toolset. Suppressor is a mission-level simulation model used to analyze military operations. A typical Suppressor simulation represents a military operation, such as aircraft attacking defended targets. Input files from the scenario, environment and type databases define a Suppressor simulation. Output from Suppressor is a time-ordered list of events [4]. The Advanced Low Altitude Radar Model (ALARM) evaluates the performance of a groundbased radar system attempting to detect low-altitude aircraft. Its primary mission is to provide areas of detectability by a single radar and to help the analyst understand the detectability phenomenon. ALARM input data is several data blocks that correspond to the components being modeled. Output from ALARM is a flight-path sequence file that specifies whether or not the target aircraft was detected [5]. F2T2E Tech Analysis Requirements Technology Alternative Assessment Cost Analysis Lethality Analysis Configure SAC Execute SAC Post Output Off Board Sensor Design Hyper-Spectral Data Processing Target Area Assessment Lethality Analysis Off Board Sensor Design Improved HyperSpectral Data Processing Target Area Assessment Lethality Analysis Off Board C4I Improved Off Board C4I Figure 3. CEE design process. 4.2. CEE Architecture A Smart Enterprise Model (SEM) was developed using the Windchill Collaborative Product Commerce tool. Windchill was chosen as the foundational infrastructure tool for CEE based on its following capabilities: • Web-centric tool • Federated data model • Lightweight web browser based interface • Web-enabled legacy applications and databases • CORBA support • Rapid application development environment The SEM is a product model that captures the structure and behavior of the virtual prototype. It’s complete in the sense that it captures only the product data that needs to be shared among the users of the collaboration. Data that’s proprietary to an organization is not stored in the SEM. Logically the SEM is a single entity, but since it’s implemented using a relational database, it can physically reside in 1..N locations. The physical partitioning of the SEM is usually governed by disk space, number of users, and database performance issues. To ensure data integrity, once product data is stored in the SEM, it Page 4 of 6 is configuration managed throughout the product’s lifecycle. Changes to the data are stored and tracked allowing users to recall earlier versions at any time. Configuration management is a fundamental capability of a PDM system. For CEE, the Windchill tool provides the configuration management for the SEM data, as well as providing access control and security capabilities. Figure 4, shown below, illustrates the SEM based CEE architecture. Requirements Server Target Area Computation Tools Cost Analysis Tool Smart Enterprise Model Sensor Design Tools Hyper-Spectral Data Processing Tools Lethality Analysis Tollset (Suppressor/ALARM Context) Figure 4. SEM based CEE architecture. The design tools (e.g. Sensor Design Tools) were integrated into CEE using CORBA (Common Object Request Broker Architecture) to interface with the SEM. The user interface is a workflow implementation of the CEE design process displayed in a web browser on the client machine(s). When the user executes a workflow task, by clicking on it, the appropriate resource agents are executed to retrieve the appropriate dataset from the SEM. The design tool is then executed. When the workflow task is complete, another resource agent is executed to store the modified dataset back in the SEM with a new configuration. Should the user wish to compare data from a previous workflow task, or if system administration is required, the Windchill client can be executed to provide that capability. 5 . Summary The paradigm for virtual prototyping and product development continues to move towards a web-based collaborative computing environment. Instead of the traditional heavyweight client/server approach to enterprise systems, collaborative environments inspire a Web-Centric federated approach. Many vendors have recognized this trend and are making their products more web-friendly by incorporating a lightweight client into their product that can be executed from within a web browser. These lightweight browser clients eliminate support and maintenance costs as well as software compatibility issues. This offers power with simplicity as mission-critical information applications are disguised as simple web pages. There are limits to benefits achieved through commonality across distributed collaborative environment architectures. Qualitative factors also play an important role in these trade-offs. Lockheed Martin ATL continues to research, explore, and apply new and emerging technologies in virtual prototyping and collaborative product development. 6 . References [1] Chadha, B., Welsh, J., “Next-Generation Architecture to Support Simulation-Based Acquisition”, American Society of Naval Engineers (ASNE) Day 2000, May 2000. [2] Chadha, B., “A Federated PIM for Supply Chains”, DH Brown Symposium, 1997. [3] Stavash, J., Welsh, J., “Integrating Product Data Management in a Collaborative Engineering and Virtual Prototyping Environment”, MCES, 1999. [4] Science Applications International Corporation. 1996. Suppressor Release 5.4 Analyst’s Manual. [5] Science Applications International Corporation. 1997. Operational Concepts Document (Analyst’s Manual) for the Advanced Low Altitude Radar Model (ALARM 3.2). Page 5 of 6 7 . Author Biography Jeffrey Stavash is a Senior Member of the Engineering Staff. He was the lead engineer for the USAF’s CEE program, in which he developed a collaborative prototyping environment using the Windchill PDM system. Prior to that, Mr. Stavash developed an enterprise infrastructure for the RASSP (Rapid Prototyping of Application-Specific Signal Processors) program to support an Integrated Product Development Environment for Digital Signal Processor systems. Mr. Stavash has a Masters degree in computer science from New Jersey Institute of Technology and a Bachelors degree in computer science from Seton Hall University. He is also a member of the Association for Computing Machinery (ACM). Mr. Welsh manages the enterprise technology organization, leading collaborative engineering projects for ship systems, mission-level analyses, and logistics supply chain improvement. He has over 20 years technical and managerial experience on enterprise software, systems engineering, and electrical engineering. He has a Bachelors degree in electrical engineering from Villanova University and a Masters degree in systems engineering from the University of Pennsylvania. Dr. Chadha is the principal investigator on enterprise engineering, supply-chain integration, and process improvement initiatives within Lockheed Martin. He leads the SPM architecture team for Lockheed Martin’s DD 21 program. Dr. Chadha is a member of the ASME Engineering Information Management Committee, the Supply Chain Council, and chairs Lockheed Martin’s PDM Interfaces Working Group. He received his Ph.D. in Mechanical Engineering from Georgia Institute of Technology. Page 6 of 6