The idea

Big data, but for health care and treatment. That’s the idea that Jukka-Pekka “JP” Onnela, PhD had that led to the creation of the Beiwe app, the mobile side of the Beiwe research platform that uses modern smartphone technology to collect data for clinical studies. Rocket Farm Studios was honored to work with Onnela on this mobile app that is providing a better way to gather better information.

Onnela is a transplant from Finland brought to Boston as a result of winning a Fulbright scholarship, and is now an Assistant Professor of Biostatistics at Harvard University. There, he runs the Onnela Lab which focuses on statistical network science and digital phenotyping. Explains Onnela, “in the former, we study social and biological networks and develop new mathematical models for them. The latter refers to moment-by-moment quantification of human social and behavioral data from personal digital devices.”

In essence, Onnela and his team are looking to use data to explain how and why people act like they do. One tangible area Onnela is researching is in patient care. Traditionally, gathering patient outcome data before or after treatment has been difficult. “We have worked with CDR (call detail records) data to learn about social networks, but people are generally bad at recalling events and the networks are very large,” Onnela said. “So we thought smartphones would be a good tool to collect real-time data to create social and behavioral phenotypes for people who suffer from certain diseases.”

“Medical data collection or intervention usually takes place in hospitals, but this only happens every so often. But the paradigm is starting to change. For example, one collaborator is a neurosurgeon who operates on people with brain tumors. How do we know if the patients are doing well? My collaborators meet with his patients twice, once before and once after surgery. Then the patients are sent surveys by mail. Traditionally, there’s about four points of data collection. Our idea is to use digital smartphone “breadcrumbs,” collect much more data, and turn them into biomedically useful information.”

The design

When Onnela connected with Rocket Farm Studios to build the iOS app, there were many design factors to consider. “We wanted the app to be simple,” said Onnela. “Since this app would be in the hands of patients, perhaps with cognitive or physical issues, it had to be easy to use.”

The app was designed to collect both passive (sensor-based) and active data (surveys) to best monitor cognitive and physical functioning. “What’s the user’s cognitive functioning? Social functioning? Has he or she left the home? How long did that take? For how long? Can the person respond to text messages?” Onnela continued, “using GPS data, we can calculate mobility metrics. For example, how much time does someone spend at home? And screen activity, like the number of times a screen is on during a segment of time. We can quantify sleep patterns accurately with this data.”

The Beiwe app was also designed to be used pre and post treatment to show behavioral changes. “We can collect data 2 to 3 months before an operation, and then 6 months after. With this app, we can collect data multiple times a week instead of once a month.”

Of course, because of the sensitive patient data, security was of the utmost importance. Said Onnela, “we had to think about privacy from the start. We scramble potential identifiers and all data is encrypted at all times. We have about 10 studies approved now and that’s a testament to how we handle privacy.”

the customer

"Rocket Farm was good to collaborate with,” said Onnela. I liked that they had a designer with us from the very beginning who was great. The team did a great job there."
"We quickly learned from Rocket Farm that a user’s expectation is different for an iOS app compared with an Android app. You can’t just copy it over, you have to modify it to the iOS user’s expectations. We had to build the app using Apple’s ResearchKit, and that forced us to show only one survey question at a time [in contrast to the Android app]. We liked this design so much, we’re going to implement this on our Android app too."
"I was very happy with how it worked out. In our case, we wanted a customizable platform. The app had to interact with an established back-end, and I was impressed with how Rocket Farm worked with the back-end developers to get it done."

Check out the app on