In the early history of engineering, the engineer was an integrator of many technologies. With the advent of the digital computer in the mid 1950's this began to subtly change as the demands of implementing various analytical methods on digital computers tended to focus engineers on the analysis of a single discipline such as structural mechanics. This revolution in engineering practice is reaching maturity after almost 40 years of intensive research. For such fields as rigid body mechanics, structures, and fluid mechanics, computational methods exist which span a range from simple techniques to full implementations of the underlying field equations. This paper will argue that the remaining part of the revolution in engineering should refocus on the engineer as a generalist by using the computational techniques in a more integrated fashion. To accomplish this, work that needs to be carried out in several areas will be discussed. These include modelling and data representation strategies, automated discretiztion methods, directional information generation and decision methods.