Disseminating Synthetic Smart Home Data for Advanced Applications

The research in IoT and Smart Homes fields is rapidly growing, leading to the emergence of new services to improve the health and lifestyle of people based on the analysis of data that they produce performing their daily activities. However, researchers report a lack of high-quality publicly-available datasets: conducting experiments gathering such data is long and expensive, especially if the annotation of meaningful information (environment, person’s activity, health status) is required. Moreover, there are even more specific settings (e.g., dementia detection) where data must be related to a change in inhabitants’ behavior. We present a collection of new publicly-available datasets generated with the SHARON simulator. Thanks to this software, researchers can obtain synthetic data suiting their specific requirements. Two classes of datasets are described: one extends existing datasets preserving the original statistical properties, the other is composed of simulations of virtual inhabitant-environment systems. Moreover, we induced behavioral drifts compatible with dementia symptoms, generating further datasets. We believe that these resources may help the progress of research, as long as new real-life high-quality datasets are not available.