Medication intake adherence with real time activity recognition on IoT

Usefulness of health care services is seriously effected by medication adherence. Medication intake is one of the cases where adherence may be difficult for the patients, who are willing to undertake the prescribed therapy. Internet of Things (IoT) infrastructure for activity monitoring is a strong candidate solution to maintain adherence of forgetful patients. In this study, we propose an IoT framework where medication intake is ensured with real time continuous activity recognition. We present our use case with the focus of application and network layers. We utilize an activity classification scheme, which considers inter-activity detection consistency based on non-predefined feature extraction as the application layer. The network layer includes a gateway structure ensuring end-to-end reliability in the connection between a wireless sensor network (WSN) and Internet. Results obtained in simulation environment suggest that the selected application and network layers introduce a feasible solution for the medication intake use case.