The Function-Human Error Design Method (FHEDM)
Candidate for Doctor of Philosophy in Mechanical Engineering
Major Professor: Dr. Irem Y. Tumer Co-Major Professor: Dr. H. Onan Demirel
This research presents a methodology to identify and analyze human-system interactions, and their potential failure modes using a functional decomposition approach during the early system design stage. The fundamental research contribution of this work is the development of a framework to identify associations between physical user-product interactions, generic human errors using a functional model framework. Providing designers with such associations will allow the integration of Human Factors Engineering (HFE) principles early in the design process to assess user-system interactions by identifying possible human error scenarios and component failures. This methodology can also be used to improve design engineering methods, by explicitly refining the user considerations and providing design guidelines to enhance the usability, safety, and comfort of the user while operating the engineered system. This research introduces the components of Function-Human Error Design Method (FHEDM). First, we build the human error database, which categorizes human errors associated with user-system physical interactions allowing designers to identify user errors while completing tasks within the system. Additionally, we present a method capable of mapping usersystem physical interactions and their possible human errors within a functional model framework. The preliminary results of the research suggest that design information from a rich dataset can be used for extracting association rules between functions, flows, components, and user interactions. This work contributes to the design community by automating the identification of user interactions and potential human errors from a single functional model.
Monday, June 3, 2019 at 2:00pm
Rogers Hall, 226
2000 SW Monroe Avenue, Corvallis, OR 97331
Nicolas F. Soria Zurita
Login to interact with events, personalize your calendar, and get recommendations.
No recent activity