CHI.2020

The quality of air in office spaces can have far-reaching impacts on the well-being and productivity of office workers. We present a system, called Hilo, that capitalizes on machine learning methods to forecast the level of carbon dioxide (CO2) in shared office spaces. Experimenting with Human-AI Interaction techniques, our main objective is to engage users in taking preventive actions when a harmful level of CO2 is predicted. We elicited three main elements of such prediction–Risk, Temporal Proximity, and Certainty–  and explored alternative ways of displaying indoor CO2 forecast through these elements. Three interfaces on Apple Watch were tested by 12 participants (within-subjects, a total of 36 sessions). We describe the results of this study and discuss implications for future work on how to create an engaging interaction with the users about the quality of air in offices and particularly its forecast. 

Hilo-Wear: A Preliminary Study to Explore the Air Quality Forecast Interfaces in Office Space