Chapter 9 - In-Home Deployments

Abstract

In-home deployments can be used to study current behavior patterns or to evaluate prototype systems. Compared to other options such as lab studies, interviews, or surveys, in-home deployments offer unparalleled realism due to the in-situ nature of the deployment as part of the participants’ normal lives. However, in-home deployments require careful planning and considerable effort. This chapter will help in the planning and successful execution of home deployment studies by sharing insights drawn from our combined experience running many different studies in the home, and by using the deployment of the PreHeat prototype as a concrete example. Among the lessons we highlight are the importance of carefully considering the minimum viable prototype to build for deployment, the value of remote monitoring to catch problems, and the importance of flexibility and robustness in deployed systems to cope with unexpected issues in the home environment. By discussing the challenges we faced and the lessons we learned, we hope that others will be able to more easily conduct in-home deployments and gather the rich and informative data they provide. Our experiences have led us to develop and open source two platforms that strive to reduce the engineering effort required for deployments: .NET Gadgeteer (http://www.netmf.com/gadgeteer/), a prototyping platform for custom devices, and Lab of Things (http://www.lab-of-things.com/), an SDK (software development kit) that provides features such as remote monitoring and updates for home deployments.