The impact of visual computer displays on different types of scientific problem solving and the improvement of human performance in these tasks form the unifying theme of the research to be presented. One goal of this research is to design a computer system which allows us to explore human performance in a visual reasoning task, and ultimately, which facilitates and stimulates human reasoning capabilities by providing intelligent assistance. Such intelligence consists of knowing what type of assistance is needed, and when it may be cognitively effective to afford it. Since it is important to develop an understanding of human cognitive processes such as perception and problem solving before we can hope to assist those processes, our general approach has been the following: 1) use cognitive science techniques to collect and analyze experimental data in a particular task domain; 2) utilize experimental results in the development of a cognitive model of the process under consideration; and 3) embed this model in a computer system designed to probe and test the model in order to refine it. This paper is meant to provide an overview of the research progress to date on this topic. The collection and analysis of experimental data have been completed, and a first version of the Cognitive Model of Visual Interaction is presented. The details of this work can be found in [11, 12, 13]. Current work in progress is the design and implementation of the intelligent computer system, and our final effort will involve the testing and evaluation of the system, once again using subjects from our task domain.