A Framework for Model-Based Adaptive Training

Concurrent Engineering

A project life-cycle approach that is a hot topic in manufacturing is called concurrent engineering [Nichols 1992]. Concurrent engineering is a project life-cycle approach to developing, simultaneously, different elements of the same system.

The Institute for Defense Analysis defines concurrent engineering as follows: “Concurrent Engineering is a systematic approach to the integrated, concurrent design of product and their related processes, including manufacture and support” [IDA Report R-338, 1988].

In a software project this means concurrent information engineering on distributed computer platforms. Concurrent information engineering is an alternative to using the waterfall approach in which all analysis is completed before performing design and all design is completed before implementation. Instead lots of mini-waterfalls could be used that are operating in parallel.

A structured break-up into independent concurrent projects still needs to have a meta-level basis. Although these parallel projects can still be small waterfalls, they are short term and each project is closely coordinated with other projects.

For a spiral approach to system development, concurrent information engineering has implications similar to that of the waterfall approach. Instead of one linear, ever unwinding spiral of development, sub-projects of concurrent mini-spirals can be propagated as concurrent projects [Odell 1992].

Can concurrent information engineering be used to define solutions with the MOBAT specification framework? The proposed methodology for the MOBAT framework is based on ITSIE [Sime & Leitch 1993] which is compatible with the KADS modelling and spiral life-cycle approach. The ITSIE approach is to be refined and tested for MOBAT applications.

The MOBAT methodology should be a complete solution for building ITS applications. However, to allow adapting the MOBAT specification approach to various industrial environments, no major restrictions should be placed on the life-cycle technique. ITSIE has less complexity than KADS due to the bounded domain expertise representation in 3 perspectives: procedure-based, rule-based and principle-based. Each of the 3 perspectives stands on its own. There is no particular perspective that needs to be developed first. This means concurrent engineering can be used to design different training agents in parallel.

Flower Show
© | | Sitemap