Using knowledge-based systems with CAD

The use of `Knowledge based systems´ (KBS) in CAD has been talked about and a few example systems produced over the last 5 years. These examples are nearly all small University produced systems rather than industrial applications. The CAD industry has perhaps correctly recognised but `hype´ and reality are often very different. The majority of successful KBS have been very limited in the domain to which they can be applied, have utilized only shallow knowledge and have been expensive to construct and maintain. However, KBS have a series of useful attributes which the CAD industry should be attempting to exploit. Examples are: increased intelligence; adaptable behaviour; high quality human computer interfaces (HCI); explicit representation of knowledge; and can be applied to problems which traditional computing techniques find hard to tackle. The problem is then one of finding efficient ways in which traditional CAD techniques can be merged with newer techniques in order to improve CAD system performance and profitability