The University of Michigan’s Heejin Jeong discusses computation human performance models and data analytics alternative to costly human-subject research.
Seminar: Research on Human Interaction with Autonomous Systems: Integrating Human Performance Modeling, Data Analytics and Behavioral Experiments
Tuesday, March 20, 2018
10–11 a.m.
Santan (SANTN) 220, Polytechnic campus [map]
This seminar is free and available to view online via Adobe Connect.
Abstract
Human-systems engineering research traditionally relies on human-subject experiments, which are costly and time-consuming. To minimize these limitations, computational human performance models and data analytics can be used as alternatives. Human performance models generate a digital simulation of human performance and examine underlying psychological and physiological mechanisms to help understand and predict human performances. Data analytics, such as machine learning algorithms, help analyze and predict human behavior patterns from experimental and naturalistic data. This seminar describes Jeong’s work of integrating human performance modeling, data analytics and behavior experiments in two studies on transportation human factors. The studies’ research methods included human performance modeling, behavioral experiments used to predict human behavior in in-vehicle stimulus-response tasks, data analytics and Michigan crash report data to investigate risk factors associated with the severity of car crash injury. Plans for research, funding, collaboration and teaching will be presented.
About the speaker
Heejin Jeong recently completed his doctorate in Industrial and Operations Engineering from the University of Michigan-Ann Arbor with a focus on human factors and systems engineering. His primary research focuses on supporting human decision making in human interaction with autonomous systems. He has conducted many human-systems engineering research projects at the Center for Ergonomics at Michigan and at the University of Michigan Transportation Research Institute. He seeks to apply multiple methodologies including experimental, human performance modeling and data analytics to societal problems in multiple domains including transportation, robotics and healthcare.