One of my most recent projects was the localization of an intelligent assistant for use at airports in Sweden and Japan. My team did a questionnaire survey, as well as interviews with digital strategy experts. They were asked to evaluate the prototypes that we created.
I wanted to focus on the area of the internet of things and interface localization/standardization. The study used software from Stanford University and was inspired by the InfoGate project at the Munich Airport, which tests the idea of virtual airport assistants already.
Face to face interaction with intelligent assistants is a very personal experience, similar to human-to-human interaction. Cultural norms have to be followed in addition to making sure the agent gives the right answer. It is an emerging area of research to provide hints on the localization of the digital assistants.
The study consisted of two parts, an online crowd-sourced survey of user attitudes and preferences when interacting with the virtual airport assistants. The survey was accompanied by a live demo, which we prepared so that the users could try themselves interacting with different kinds of service agents.
The prototype of an airport service agent
In the second part, we did semi-structured interviews with digital strategy experts in both countries to shed some light on the quantitative survey results. We discussed issues like trust and privacy, and overall attitudes to the new technology.
At the end, we obtained an ideal prototype of the virtual airport agent for both countries and a set of recommendations for designing such information systems in the future.
PLS Structural Model for Sweden
PLS Structural Model for Japan
- Web app prototype: https://www.petegerhat.com/kiosksurvey/
- Document: http://lup.lub.lu.se/student-papers/record/8567357/file/8567377.pdf