One of the results of the BD4QoL market analysis has been the appreciation of differences existing among the two major mobile OS platforms – iOS and Android – when it comes to behaviour measurement, which is needed to implement several mHealth applications that target patient empowerment and self-care.

This can be seen, for instance, by examining toolkits that have been specifically implemented to allow such measurement. Three of them are:

• Funf Open Sensing Framework, a ready-made platform for conducting data collection experiments ranging from health and wellness, social and psychological studies. It allows the collection of sensors data such as location, movement, phone use, social proximity and performs data transfer to selected data repositories for data analysis, visualization and notifications. The platform is only available for Android OS
• Radar-Base, allows passive collection of user’s behavior data from a number of mobile smart devices (smartphones, smartwatches, wristbands etc.) for both iOS and Android. The platform collects data from movement sensors, light sensors, phone interaction (on-off; use of applications), social activities (SMS log, number of contacts in agenda, number of Bluetooth devices connected), weather. However, for iOS most of the data cannot be collected in background, as needed by BD4QoL, and two data categories, necessary for BD4QoL (App usage and Phone logs, needed for social activity estimation) cannot be collected at all (see figure)
• Beiwe, is specifically designed to collect high-throughput data from smartphones, both iOS and Android. This open-source platform supports collection of a range of passive data including accelerometer, GPS, power state, phone usage, screen on/off, identifier, wi-fi connections, Bluetooth connections, proximity, gyro, magnetometer, device motion, reachability. However, limitations similar to those mentioned for Radar Base do apply, as they depend on intrinsic limitations of the different mobile OSs.

For the above reasons, BD4QoL decided to support only Android phones and to develop a custom mobile app, that overcomes additional limitations that affect this OS too, especially when a large number of models, including the less modern ones, need to be supported.