This thesis presents a taxonomy of expert elevator and amusement device inspector knowledge that was developed using task and cognitive task analysis. While literature concerning research into quality control inspection exists, very little research has been performed into safety inspection. A qualitative study captured the knowledge used by elevator and amusement device inspection. The existence of expert performance in the elevator and amusement device inspection domains was identified and a taxonomy of expert inspector knowledge was created. This taxonomy was based on a model of knowledge that distinguishes between three types of knowledge - declarative, procedural, and strategic. Further development of this taxonomy, along with an effort to perform expert inspector knowledge capture, is expected to lead to improved inspector training and performance, and an increase in consistency between the inspections performed by all inspectors.
Optimization and Performance Modeling of Stencil Computations on Modern Microprocessors