Application of the synthesis process to flight trainer simulation software development and reuse - Lessons learned and technology challenges

Abstract

As operationing budgets become more austere, military and commercial industries are reducing live training and becoming increasingly more reliant on modeling and simulation. hi turn, engineers are challenged to provide more sophisticated simulation fidelity for highly complex weapon systems prior to initial operational capability of the actual system, and within the very tight budgets traditionally afforded simulation-based trainers. Furthennore, the problem is exacerbated because life cycle budgets are being significantly constrained even though the parent system evolves through multiple changes and up-grades over its life cycle. and improved software reliability. Achievement of this goal will require fundamental changes in software development methodologies, CASE tool utilization, technical management strategies, and current DO11 acquisition practices. The objective of this paper is to foster the continued development of megaprogramming technologies through an interchange of lessons learned and technology challenges obtained thus far in the ongoing Navy research project designated STARS. The Navy STARS (Software Technology for Adaptable, Reliable Systems) effort is currently investigating software process improvement and reuse Currently, researchers hypothesize that the development of standardized, reusable simulation software assets could potentially effect significant flight simulation life-cycle cost and risk reductions. This hypothesis presupposes tlie availability of a well defined, repeatable, automated software process. The cost and risk reduction savings would be realized in the areas of development, verification and validation, hardwarelsoftware integration, documentation, configuration management, portability and support. A shift toward automated software reuse is creating new technological challenges for software engineering, particularly in the highly complex flight si~nulation community. Synthesis (developed by the Software Product Consortium under tlie Virginia Center of Excellence) is a candidate software engineering development process addressing software reuse through the use of well defined processes which can be implemented within an automated software engineering environment. 1Jnlike the classical single life-cycle acquisition model, the Synthesis two-life cycle approach, supports development of a software engine encompassing reusable assets, and a user interface which facilitates simulation model development, implen-mtation, documentation, verification and validation, portability and support. During the second life cycle, an application engineer utilizes the software engine to generate the specific software products needed for a trainer (instance) utilizing only the necessary software code required to support the given simulation. When combined with process automation, this product-line based, two life-cycle approach to software development is termed megaprogra~nmi~ig. The goal of lnegaprogranuning is cost saving *** Aerospace Engineer, AlAA Senior Member * * Aerospace Engineer * Electrical Engineer This paper is declared a work of the I1.S. Governlent and is not subject to copyright protection in the United States. techniques within the air vehicle training systems (AWS) domain. This paper addresses the difficulties experienced in defining standardized and reusable soflware assets within the A W S software domain. Discussion of technical lessons learned will include use of the Synthesis process and associated reuse techniques, including interplay with standards and guidelines. Additionally, lessons learned in the development and use of CASE tools required to facilitate software domain development and growth, reuse process automation, application generation and project management will also be addressed. Furthermore, a discussion addressing the requirement for, and difficulty in defining, estimating and measuring return on investment will be presented with examples drawn from the Navy STARS program. Recommendations on various reuse concepts best suited for the A W S domain will be provided.