In this human-robot interaction project, we created a chatbot prototype designated for hotel quarantine life with RASA framework.
Quarantine is a relatively common thing during the COVID-19 pandemic, and a lot of people have been experienced a short period of quarantine life. I talked to some of my friends to learn about the problems occurred during quarantine, and try to think about what the bot could do to improve their quarantine experience.
New policies about the quarantine period, test requirements and hotel rules are often confusing, really need a better channel to know these messages. Also there are too few ways to entertain oneself, and often feel lonely due to lack of communication.— Chu (Translated)
During the brainstorm session, we’ve considered various demands of quarantine life, including food, information, entertainment and mental & physical health. Based my own experience, the entertainment demand is critical in days without talking to people, so I proposed that the bot could offer personalized recommendations on entertainment activities, such as movies, musics, etc.
Finally we’ve chosen 3 functions to implement into our demo: food ordering, quarantine information and emotional guiding. We consider these functions as essential and helpful during one’s quarantine.
Food ordering is one of the basic demand of human being, and the food condition in a quarantine hotel is often limited and not supplied in 24 hours. Therefore we designed a user-friendly catering process, with up-to-date menu and on-time order processing.
Quarantine information gives users a convenient, reliable and personalized way to know about the probably complex and rapid-changing policies. The bot can retrieve and process personal information, combined with the latest policy of local government to answer users’ questions. This could reduce the workload of public health departments and also help users to save time on looking up these information.
The bot also offers emotional guidance as part of its function. It contains professional suggestions and also practical way to get appropriate help. The bot try to detect user emotions from context and find the best match on its database.
We’ve also conducted usability testing and collected a lot of informative advices, which were used to further iterate our design.
I also edited the showcase video recorded by Simon.
The usability test reveals to me that it’s really important to look at the software we made from different perspectives: a perfect function in developer’s eyes could be full of problems in actual use.
Also we may need to bring forward the whole process of the project, as there appears to be a lot of time pressure by the deadline. All in all, a great teamwork of all of our group members, and the project itself is very helpful and educational, both in the designing works and programming works.