SBIR-STTR Award

TeamVision proposes that we research the feasibility of incorporating an adaptive object based optimization system into an existing multi-user object oriented application integration framework. Successful implementation of such a system would save NASA and industry billions of dollars annually by reducing significantly the false starts, redundant effort, long lead times and globally non-optimal solutions characteristic of the early decision process of complex systems. Fortunately, recent breakthroughs in software engineering have demonstrating a practical and proven strategy to this seemingly intractable problem. The wide spread adoption of Object Oriented languages like C++ in the last ten years has created a revolution in modern software development. This move away from serial and procedurally based languages to Object Oriented languages has allowed teams of software programmers to collaborate on large and complex software development projects unthinkable under the old language paradigms. TeamVision is the first company, through a previous NASA-SBIR's and private funding, to develop and commercialize a software solution that utilizes this approach in the early decision process. The proposed innovation builds on this successful track record and would significantly expand the field of optimization into new areas.

The proposed innovation is to combine traditional and cutting edge optimization techniques into an existing powerful object based organic enterprise decision network called FrameworkCT. This combination would represent not only a leap in the field of optimization itself but also in the methods by which these models are produced for optimization process in the first place. This will be the first time ever that (i) multidisciplinary model integration, (ii) distributed parallel processing, (iii) statistical simulation, and (iv) optimization algorithms will be put together to solve complex systems engineering optimization problems. In our Phase I research we have not only demonstrate the feasibility of the proposed innovation but in our preliminary tests we uncovered important synergies between the existing neural network database foundation of FrameworkCT and neural network optimization as well as between the current massive parallel processing capability and genetic optimization algorithms. Or overall effort is closely aligned with NASA's new initiative of going back to the moon, mars and beyond. This alignment with a current enterprise decision system will help insure the ultimate usefulness of the proposed innovation to NASA as well as industry and result in a tangible benefit during and immediately following the conclusion of the project.