We are researching technologies related to interaction and user interface (UI) design to support people's behavior change and idea transformation. We measure various human-related indicators, such as work performance, cognitive load, memory, eye movement, etc., as objective or subjective data and apply them to interaction and UI design.

Interaction Design to Improve Resilience Capacity

 In socio-technical systems that support social infrastructure, such as airlines, railroads, medical care, and nuclear power plants, the goal has been to achieve service safety by minimizing human error as much as possible (Safety-I). However, as systems and environments have become more complex and constantly changing in recent years, emergency events that workers have never experienced are increasingly occurring. Conventional safety management methods for analyzing known accidents are insufficient to deal with these unexpected events. Therefore, plans based on people’s resilience are attracting attention (Safety-II). Resilience refers to a person's ability to respond flexibly. When encountering strange events, people can avoid failures and worst-case scenarios by exercising resilience and responding autonomously and adaptively. Resilience is often described as a positive human capacity compared to human error.

In our laboratory, we are engaged in research on interaction design to improve people's resilience capabilities, using work in the air traffic control and medical fields as subjects. We are experimentally testing the hypothesis that user interfaces designed based on some guidelines can elicit and strengthen people's resilience and positive abilities. In particular, we have recently been researching methods for extracting generic resilience abilities independent of the work's expertise and content. We set up an experimental task that generalizes the cognitive characteristics of air traffic control work, embeds scenarios likely to elicit resilient abilities, and extracts features of resilient behaviors by measuring and analyzing participants' work performance, eye movements, thought processes, and strategies. We then design interactions that promote and reinforce these behaviors.
 

Development of visual salience model and application to UI design

 We are developing an attractiveness model that quantifies the attractiveness of GUI screens and applies it to UI design. We are developing an attractiveness model specialized for GUI screen features by extending the model for extracting attractiveness regions of pictures and photos that have been conventionally grown in image processing. In this study, we clarified that color, size, and layout are the design elements that significantly affect the prominence of objects (UI components, icons, etc.) on GUI screens. We then measured these influences through cognitive psychology experiments and developed a numerical model. Furthermore, we used this model to design a color scheme for air traffic control radar screens. We demonstrated that differences in conspicuousness by color affect worker performance and that design based on this model effectively improves worker performance and reduces cognitive load.