Data Capture and Analyses from Conversational Devices in the Homes of the Elderly

Conversational devices such as Amazon Echo and Google Home represent more than a way to tap into the behavioral surplus of consumers. They provide an opportunity to address societal problems by examining data streams produced by these devices. In this paper, we describe usage patterns and problems related to the use of Amazon Echo devices at home by one specific demographic: the elderly. We rely on a pilot project to collect usage data over multiple months based on deployment of these devices in the homes of eight elderly individuals who either live alone or with a spouse. The paper describes methods used to ensure confidentiality, data collection and analysis procedures, and our findings. We find that the use of conversational devices remains restricted to single commands instead of conversations, making yourself understood remains a problem, sustained use remains a challenge, and the interaction rarely goes beyond simple commands. We interpret the results, and point to the potential for such devices in the lives of the elderly, specifically for health-related problems. The paper also describes lessons learned for capture and analysis of data from such conversational devices.