U.S. Army Missile Command, Research, Development, and Engineering Center, RDEC Procurement Office, R&D Contracts Division, Redstone Arsenal, AL 35898-5275

A -- THE ARMY IS SEEKING IDEAS FROM INDUSTRY AND ACADEMIA ON RESEARCH, DEVELOPMENT AND ENGINEERING EFFORTS FOR JOINT SOFTWARE TECHNOLOGY INVESTMENT: A PILOT PROGRAM FOR THE TRI-SERVICES OF DOD SOL 95X196 DUE 092695 POC (RDPC) Natalynn Weddle, Contract Specialist, AMSMI-RD-PC-HA, (205) 876-4900 or Harold Smith, Contracting Officer. Synopsis No. R196-95. Research and Development Sources Sought. Joint Software Technology Investment: A Pilot Program for the Tri-Services of DoD. This is a request for technical information in the form of white papers. The government does not intend to imply that future contracting opportunities will be available, although the possibility exists that Bailment Agreements and CRDA's may be utilized. The Army is seeking ideas from industry and academia on research, development and engineering efforts for a pilot technical agenda to address the systematic lowering of software development costs and maintenance for DoD's Tri-Services. The essence of the effort would be to establish the automatic production of efficient, reliable software for advanced weapon systems - beginning with conceptual mathematics or physical systems models and continuing to executable and maintainable system and simulation code. Automatic programming would be used for rapid prototyping and for the evaluation of hardware-software interactions to determine those combinations which create the best systems solutions and the most efficient code executions. Realization of joint software-technology investment would meet an express need of the Tri-Services, with the work conducted by the U.S. Army Missile Command (MICOM) and the U.S. Naval Air Warfare Center (NAWC). Work would be defined for joint efforts, within an implementation structure, for co-developing domain-specific software environments in a manner which would evolutionize the processing of designing, developing and maintaining software for current and future missile systems. Strategies for this Joint Software Technology Investment and the technical approach would include the implementation of a concept that would automate a significant portion of the software life-cycle, and in return, improve software quality and reduce the costs of developing modern weapon systems. This plan is a result of joint efforts between the Army and the Navy in the Joint Services Guidance and Control Committee (JSGCC) and Joint Directors of Laboratories (JDL). Efficient, high-quality automatic programming to product software would be the main objective of this research, development, and engineering effort. Mathematical models for weapon systems involve processing matrix operations for the real-time solution of integral-differential equations, using solution algorithms and numerical methodology, either computationally synchronous or asynchronous. Work would include the development and utilization of software tools that use graph-theoretic methodology to optimize operational, simulation and prototyping systems software. The tools would acomplish software-component allocation to available hardware resources so the real-time execution is minimal. Because of microprocessor improvements in relation to execution speed, memory, and communications, it is increasingly more possible to enlarge the granularity of the mathematical model components and the software representing them and their computation parts. Partitioning the mathematical model into optimum granularities provides representation options, where partitioned components can execute and communicate efficiently within appropriate hardware structures to meet real-time embedded requirements. The mathematical models would be partitioned with task-level parallelism, and the partitions meeting the precise requirements of object orientation for software developement and applications use. Effective and efficient software reuse is achievable more readily when these partitioned software elements are formed with an object-orientation. Present and future processor technologies include highly-parallel execution which leads to much higher efficiency. Coupling this execution parallelism at the processor level with object-oriented processing and caching has potential for single-processor real-time execution, as well as, for highly-paralled and distributed processing for systems environment simulations and flexible rapid prototyping. Task parallelism and variations in computational granularity are able partners in the development of efficient software for guidance and control systems design. Task parallelism must begin at the operations level to exploit the inherit execution parallemism used by current-technology and future processors. Computational granularity is a variable quantity of execution code which can exploit the entire clock-cycle computation capabilities. It can vary in the amount of execution code; depending on the execution graph, the communication data-dependent paths, and the inherit paralleslim of the target processor(s). The object-oriented, mathematics-based software depends on task parallelism and comutational granularity to tailor the systems computation--described by the execution graph--to achieve the minimal execution for the flight systems software, rapid prototyping and the simulation software. This adds significantly to the importance of the designers extended involvement in the software-hardware development process. The partitioning and tearing apart of the mathematics of the models and the environment simulations, must match the task parallelism, computational granularity and the communications topology of the target flight hardware. The effort to realize the utmost in design and performance is carried-out by knowledgeable design engineers who are qualified to ensure system performance according to requirements. The deliverables from this work are intended to increase the system life-cycle involvement of design engineers by developing a design software environment to extend their effective range. The design graphical-user-interfaces, including software tools and repositories, would be designed to accomodate and extend the design engineers involvement and effectiveness from system conception through the theoretical and practical design phases and through the software development and maintenance stages. Applications-specific hardware and software variation and flexibility can produce realizations that are finely tuned to the applications, when systems design engineers are involved and knowledgeable about these implementation environments and how the executions of their designs are carried out. In addition to the quality of execution and realization, the development costs are lowered by the available design expertise in that it makes the entire process more efficient. A part of this work would be focused on the development of software tools to facilitate the systems design process as it is extended further into the development of the flight hardware and software. (0233)

Loren Data Corp. http://www.ld.com (SYN# 0005 19950822\A-0005.SOL)