The task of solving optimal engineering design problems is considered as a demanding decision-making process where the real-life industrial problems typically have to be considered from very different perspectives. In this context the most logical approach to achieving the best solution, at the presence of multiple design criteria and numerous design variables, has been the task of performing scientific optimization to produce potential solutions for further decision-making. Accordingly multiple criteria decision-making approaches to optimal engineering design problems, via employing efficient, robust, global and multi-objective optimization algorithms, have brought a significant and competitive advantage to the optimal design. However most of these approaches, due to the characteristics of the real-life problems, often associated with the usage, dimensionality and high computational cost of the objective evaluations, have not been practical and widely acceptable in engineering design community. Here the difficulties and further requirements of utilizing the optimization approaches in optimal engineering design are discussed with a more emphasis on challenges to complex geometries, dimensionality, and multiple criteria nature of the real-life engineering design problems. As a response to the considered challenges, performing the optimizations approaches in the framework of an integrated design environment is proposed as the key success to win industry. Further this research the metamodels in general approaches to optimal engineering design, are seen as the essential but not sufficient tools to enhance creating the efficient global optimization approaches in dealing with dimensionality. In fact by extension the dimension of multiple criteria decision-making problems which has been mostly due to the increasing number of variables, optimization objectives, and decision criteria, presenting a decision-maker with numerous representative solutions on a multidimensional Pareto-optimal set can not be practical in engineering applications. Accordingly for better dealing with the ever increasing dimensionality a supplementary decision-support system to enhance the metamodels is proposed. As the result an improved decision procedure is formed according to the limited human memory and his data processing capabilities. In this context the research further contributes in shifting from generating the Pareto-optimal solutions, to the reactive and interactive construction of a sequence of solutions, where the decision-maker is the learning component in the decision-making loop. To doing so the conventional evolutionary and interactive optimization and decision-making algorithms are updated by reactive search methodology, empowered with the advanced visualization techniques, in the framework of an integrated design environment.