Go to content

Droidcon Italy 2017 // Smart-Watches as paradigm-shifting player in IoT - Francesca Stradolini

Most of the wearable applications are developed for patients’ smart-watches to detect their body-motion, body-temperature, physical exercises and amount of sleep [6]. Otherwise, monitoring systems are also used in specialized medical applications, as for the home monitoring in dementia care [7] or for patients with specific illness [8,9]. Continuous monitoring concept is extremely important and beneficial in hospital environment too, especially for critical patients in Intensive Care Units (ICUs) [10] or during surgery under anesthesia [11]. To address the necessity of keeping constantly under control vital parameters, we propose a system where the smart-watch becomes the only center of a monitoring network for hospital environments. Indeed, it is connected via Wi-Fi to a series of independent bedside monitoring systems connected to each patient. The monitoring system consists of bio-sensing platforms able to detect different endogenous and exogenous metabolites in patient’s fluids [13, 14] plus an Android device to display the measured data. In this way, a flexible and personalized monitoring is performed. The medical doctor or the caregiver, wearing the smart-watch, is able to remotely monitor multiple molecular parameters of many different patients at the same time. Incoming alert notifications are received on the smart-watch in case of anomalous values that requires medical support. It is important to underline that another strength of the monitoring system is its flexibility in measuring different endogenous/exogenous metabolites by properly functionalizing the biosensors. By only modifying the client bedside application, whatever sensor can be connected and remotely monitored. The architecture of the proposed network consists of three main building blocks: (a) a Client Android interface, running on a bedside tablet, that continuously receives and displays all the data measured by the biosensing platform on the patient, (b) an Android intermediary Server-side, running on a tablet/smart-phone in a central workstation, that collects data from different clients and enable the bidirectional communication with them, and (c) a Smart-Watch application on the doctor watch device. The Android intermediary Server-side is unavoidable to enable the multi-patients monitoring from the Smart-Watch since the wearable can be paired, by default, to only one device. Hereafter, a detailed description is provided separately for each system component. The system has been validated by creating a network with two tablets as client units, one tablet as server and one smart-watch worn by the final user. Adopting this architecture, we tested all its functionalities: • The clients simultaneously connect to the server via Wi-Fi and to the monitoring biosensors via Bluetooth. • The server connects multiple clients storing their IP addresses. Every time a new client is paired, a Spinner List on the smart-watch is automatically updated. The doctor handles/allows the client connection and disconnection. • The smart-watch queries the server for receiving parameter $ from client #. In few seconds, the user receives the desired parameter from the correct patient. We tested the Internet transmission latencies by connecting to a personal hotspot obtaining less than 2 s for on-demand queries and less than 3 s for alerts receiving. Of course, these delays may vary accordingly to the Internet connection available. • The alarm message is sent to the smart-watch if x values received from the monitoring biosensor exceed the physiological threshold. The threshold varies depending on the parameter and the number of acquired samples (x) to be averaged for the comparison with the threshold. This x value changes depending upon the desired medical application. • The correct receiving of alert notifications on smart-watch also in case of Internet unavailability. The alarm is sent as soon as the connection is re-established and a notification is properly broadcast from the smart-watch.

April 6, 2017