Design and evaluation of smart home user interface: effects of age, tasks and intelligence level

Smart homes are expected to promote productivity and enhance living experience, especially for old adults. To achieve this, the level of user interface intelligence should be designed to meet the needs of users and tasks. The aim of this study was to investigate the impact of user interface intelligence level on user performances for different tasks and different users. Two objectives were pursued. The first was to investigate how the intelligence of the user interface affects user performance when different cognitive tasks are performed. The second was to determine the differences between young and senior users when interacting with smart homes. A two-dimension conceptual model is proposed to describe the impact of these two factors, i.e., user interface intelligence and cognitive tasks, on smart home user performance. A 3-by-3 experiment was designed and three prototypes were developed. Thirty-six young and 36 senior people were recruited as participants. The results showed that different levels of intelligence of user interfaces facilitated user performance with different cognitive tasks. Users completed skill-based tasks using the least time and committing the least errors using the low-intelligence-level interface; when completing rule-based tasks, users exhibited best performance in terms of less error with the high intelligence interface. However, for senior people, because of the decline in cognitive abilities, their performance was not as clearly differentiated as that of younger people when using different smart home interfaces, especially for highly cognitively demanding tasks. Moreover, when senior people completed skill-based tasks, the task load they perceived increased together with the intelligence level of the smart home interface.