Inference processes in incomplete knowledge systems in engineering projects

A substantial number of engineering problems can be qualified as heuristic ones. Inference processes, allowing for their solution, are most frequently performed with rule-based expert systems. Such systems, however, are significantly limited by their possibility of inference only if complete knowledge is possessed, which is not always the case in practice (in particular in crisis situations). It is then indispensable to employ other methods and tools. Given the specificity of engineering projects, it seems that case-based reasoning (CBR) is an appropriate solution in this respect. In conjunction with rule-based inference mechanisms, it may constitute a complementary solution, allowing for efficient inference in problem situations in which the level of knowledge completeness varies.