Trust-driven Human-Autonomy/Robot Interaction
The use of autonomous agents to assist human performance is growing unprecedentedly. As the capabilities of autonomy advance, it would function as a full-fledged team member to perceive and analyze information, make decisions, and execute actions. For the human-autonomy/robot team to perform optimally, appropriate trust should be established between the two partners. This research aims to model trust between humans and autonomous agents and facilitate human-autonomy interaction.
Human Factors in Healthcare
The objective of this research is to optimize the interaction between healthcare providers, patients, and technology to improve patient safety and to maximize the well-being of patients and healthcare providers. We use knowledge from psychology and engineering to identify potential sources of error and to develop interventions. In this line of research, we collaborate closely with healthcare providers from the U-M Medical School and the VA Ann Arbor Healthcare System.
Human Performance in Teleoperation of Unmanned Vehicles
Robots have been deployed for applications such as urban search and rescue (USAR), border patrol, space exploration, and military service operations. While robots become increasingly more autonomous and intelligent, oftentimes human teleoperation is still needed, especially in dynamic and fast-changing environments. The goal of this research is to investigate human performance in teleoperation and propose design solutions that foster effective human-robot interaction.
Takeover Transitions in Highly Automated Driving
Highly automated driving (HAD) is becoming an engineering reality. In HAD, the driver’s role will transform from an operator to a system supervisor making it possible for him to engage in non-driving-related activities. If the automated vehicle reaches its system limit, the driver will be required to resume control of the vehicle in a certain amount of time. Despite the promising safety benefits of HAD, the concern from a human factors perspective is that drivers become increasingly out-of-the-loop (OOTL) once they start to engage in non-driving-related tasks. Drivers decoupled from the operational level of control have difficulty taking over in any situation, and particularly in situations that the automation is not able to handle. In this project, we aim to develop computational models capable of predicting drivers’ takeover readiness by analyzing both the driver’s physiological data and data from the current driving scenario in real time, and to design an adaptive in-vehicle alert system in response to drivers’ takeover readiness.
User Experience Design and Sentiment Analysis
Good user experience of a product or a service is a necessary component to ensure a product/service success. In this line of research, we develop quantitative methods to uncover users’ sentiment toward products/services and latent customer needs by mining online review data.