Research

Our research goal is to improve human performance, safety, and well-being by applying human factors, data analytics, and cognitive psychology to the analysis, design, and evaluation of the intelligent systems. Research topics include:

Human Factors in Intelligent Transportation Systems

Human Behavior Modeling

Trust in Human-Automation Teaming

Universal Design

 

Human Factors in Intelligent Transportation Systems

Intelligent transportation systems with connected and automated vehicles (CAVs) have the potential to provide our society with more fuel-efficient driving, reduce driving-related injuries and deaths, and reshape transportation and logistics. While CAVs are not completely ready on the road, it is necessary to study human factors in intelligent transportation systems to improve human performance, safety, and well-being. Our studies investigate drivers’ behavioral and physiological responses in various driving environments. We also design and evaluate human-machine interfaces (HMIs) to improve user experience and driving safety in intelligent transportation systems.

  • Du et al. (2020). Physiological responses to takeover requests in conditionally automated driving. Accident & Analysis Prevention. [PDF]
  • Du et al. (2020). Examining the effects of emotional valence and arousal on takeover performance in conditionally automated driving. Transportation Research Part C: Emerging Technologies. [PDF]
  • Du et al. (2020). Evaluating Effects of Cognitive Load, Takeover Request Lead Time, and Traffic Density on Drivers' Takeover Performance in Conditionally Automated Driving. Automotive UI 2020. [PDF] [Video]

 

 

Human Behavior Modeling

We use both data-driven methods and cognitive architecture to model dynamic human behaviors and mental states in cyber-physical systems. By leveraging models and methods from both human factors and machine learning, we develop computational models that are capable of predicting or inferring human behaviors when they interact with technologies. The inputs of models can be text, video, physiological data, etc. For example, we develop computational models to predict driver takeover performance in conditionally automated driving and situational awareness in highly automated driving. We use large-scale naturalistic driving datasets and crash databases to model real-world driver bahaviors and improve driving safety.

  • Ayoub, J.*, Du, N.*, Yang, X. J., & Zhou, F. (2022). Predicting driver takeover time in conditionally automated driving. IEEE Transactions on Intelligent Transportation Systems. (*Equal contribution). [PDF]
  • Du, N., Zhou, F., Pulver E., Tilbury, D. M., Robert, L. P., Pradhan, A. K., & Yang, X. J. (2020). Predicting driver takeover performance in conditionally automated driving. Accident & Analysis Prevention. [PDF]
  • Du, N., Zhou, F., Pulver E., Tilbury, D. M., Robert, L. P., Pradhan, A. K., & Yang, X. J. (2020). Predicting Takeover Performance in Conditionally Automated Driving. CHI 2020. [PDF]

 

 

Trust in Human-Automation Teaming

The advances in artificial intelligence (AI) and machine learning empower a new generation of intelligent systems. However, human operators increasingly treat automation as a mysterious black box and lack appropriate trust in and dependence on the intelligent systems. To tackle this challenge, our studies design and evaluate interfaces with different information types, timing, and modality to increase system transparency. This work helps human operators develop appropriate trust in automation, increase technology acceptance, and collaborate with intelligent systems more effectively. The application domains include automated vehicles, automated decision aids (military operations and medical diagnosis), and general algorithms.

  • Du, N., Robert, L., & Yang, X. J. (2022). A Cross-cultural Investigation of the Effects of Explanations on Drivers’ Trust, Preference, and Anxiety in Highly Automated Vehicles. Transportation Research Record. [PDF]
  • Du, N., Haspiel, J., Zhang, Q., Tilbury, D., Pradhan, A. K., Yang, X. J., & Robert Jr, L. P. (2019). Look who’s talking now: Implications of AV’s explanations on driver’s trust, AV preference, anxiety and mental workload. Transportation Research Part C. [PDF]
  • Zhang, Q., Du, N., Yang, X. J., & Robert, L. (2018). Trust in AVs: The Impact of Expectations and Individual Differences. The Conference on Autonomous Vehicles in Society [PDF]

 

 

  • Du, N., Huang, K., Yang, X. J. (2019). Not all information is equal: Effects of disclosing likelihood information on trust, compliance and reliance, and task performance in human-automation teaming. Human Factors. [PDF]
  • Du N., Zhang Q., & Yang, X. J. (2018). Effects of automation reliability and reliability information on trust, dependence and dual-task performance. HFES 2018. [PDF]

 

 

  • Luo, R., Du, N., & Yang, X. J. (2022). Enhancing autonomy transparency: an option-centric rationale approach. International Journal of Human-Computer Interaction. [PDF]

 

 

Universal Design

We design assistive technologies for People with Disabilities (PwD), such as People who are Blind or Visually Impaired and Wheelchair Users. The assistive technologies could be navigation apps, exoskeleton, robotic wheelchair, etc. Following human-centered design principles, we employ an iterative design process to understand user needs, and design and evaluate assitive technologies. Our research could help PwD improve their quality of life and social independence by being able to perform mobility activities easier and more frequently. For example, we design the interface of Mebot with the collaboration of Human Engineering Research Laboratories.