Take a sneak peek at the new NIST.gov and let us know what you think!
(Please note: some content may not be complete on the beta site.).
NIST Authors in Bold
|Author(s):||Eric D. Simmon; John V. Messina; Arthur Griesser;|
|Title:||Simple UML Modeling to Improve the Development of Information Standards|
|Published:||September 24, 2005|
|Abstract:||Streamlining and Bulletproofing the Standards Development Process In the increasingly complex semiconductor fabrication environment capturing and transferring information throughout the product development cycle is critical. As the process of creating semiconductors becomes increasingly complex, the information associated with the process (on all levels from process control to product lifecycle information) is becoming massive. Building information standards to transfer data between different stakeholders is difficult. This difficult task is made even more difficult by the numerous players all working together, each one with their own set of needs and priorities. Traditional standards committees often follow an ad-hoc approach where the standard is developed piecemeal without following any overall structured development process. This often leads to standards that are either incomplete, cumbersome, or unworkable for their intended purpose. An alternative is to use structured development methods, currently in use by the software development industry, which are useful in both building solid information standards that capture all necessary data and in providing transport formats favorable to all players. While a full blown rigorous approach to information design might be difficult to implement in the standards arena (due to resource constraints) adopting even some of these tools can provide significant benefits to the to the standards development process. Problems seen in the present development paradigm One of the more common problems encountered by ad-hoc standards development is there is no method to insure all the information central to the standards is captured. Since the main purpose of most information standards is to gather and exchange electronic data, the failure to identify key information can potentially cripple or even render the standard useless. Often key information can be lost or misformed due to two reasons: confusion over the scope of the project or by a failure to keep the project's domain information (data to be exchanged) and the software transportation layer (transfer mechanism) separate. Both of these situations come about because the developers are looking at the standard with slightly different perspectives and expectations. During ad-hock standards development, these different concerns are often confused. When this happens, the model and the resultant standard is muddied. A structured development process will help those developers come together with a common understood view of the problem that the standard is being designed to address. Using basic UML to solve development problems The Unified Modeling Language (UML), a complex modeling language composed of many different types of diagrams used to understand and structure the software design process, can help mitigate some of the difficulties with information standards development. While the entire UML based Rational Unified Process is very complex, using portions of UML can easily and quickly provide benefit to standards development teams. The central UML model type, the class diagram is a great tool to help insure the domain information is understood and the scope of the standard is well defined to all parties. It can reduce the redundant effort when the standard must be revisited due to missing or conflicting data and help insure the capture of all necessary data in the standards. It will allow different stakeholders to develop software level implementations that meet their particular needs while interoperating with other stakeholders and containing data conforming to the information standard.|
|Proceedings:||AEC/APC Symposium XVII|
|Location:||Indian Wells, CA|
|Dates:||September 24-29, 2005|