Behavio: Opening Up the Power of Mobile Sensing

    by Nadav Aharony
    September 11, 2012

    What three things do you always take with you when you leave your home? Most people would say their keys, their wallet, and their mobile phone. But what if your phone wasn’t just a phone?

    It already isn’t.

    At Behavio, we want to help small developers, researchers, data gatherers, and end-users lower the barrier of tapping into the signals and sensors accessible via mobile devices.


    Today’s smartphones have evolved into incredibly rich sensing and computing devices, that oh-by-the-way can also make phone calls. They are jam-packed with on-board sensors for things like location (GPS), movement (accelerometer), temperature, atmospheric pressure, and more. In addition, these mobile devices can tap into many other signals like what apps are running and when, whether the phone is plugged into power, or if the screen is on or off. Altogether we are talking about dozens of signals that can be combined and used to infer complex and interesting things about us, our environment, and our communities.

    These signals can be used to do everything from monitoring your physical activity levels, to showing you how much time you spend distractedly checking email during lectures, to knowing how much time you really spend with your close friends. Taking things up a notch, your phone can even track the progression of Alzheimer’s and predict disease by noticing small changes in behavior — subtle changes that doctors or family members might miss.

    News media can be augmented with the richness of available sensor data, and give insight into how the photographer got there, what he or she was doing, who else was there, and other details about their environment. This context can be collected automatically and continuously, before they even know an event might be significant.


    Visualization of what your average smartphone can see.

    Data from many phones can be combined for investigating how a society operates, and help us better understand what it means to be human. Data can be contributed anonymously by many users and aggregated to create a collective image of a community or even a whole city. A news story doesn’t have to be the work of one journalist, but can be the emergent property from the data of thousands of citizens. Alternatively, when disaster hits — like a flood, for example — we could use contributed sensing data from people who are in the area to reconstruct, from the ground up, maps of roads that are still operational or find pockets of people who are stranded.

    The Challenges

    All of this is possible, but unfortunately it is presently extremely difficult to tap into these abilities. There exist many challenges to taking these products and applications from a concept, demo, or lab prototype into the real world. To name a few: the computational algorithms for doing the collection, integration, and inference of this “big data.” Running “smart” applications on our phones can quickly drain the battery if not done carefully. Ensuring the data collection is done ethically (informing the users, receiving their consent, protecting user privacy, and dealing with their information responsibly).

    Presently, only the big players have the resources to access and utilize these capabilities, and when they do, the code/functionality is usually proprietary and not accessible to developers. In addition, the data itself is often not available to the end users, if they are even aware it is being collected. Leading smaller organizations end up duplicating basic functionality, over and over again.

    Introducing Behavio

    Behavio grew out of the MIT Media Lab, where we worked for the past several years on using mobile phones as social and behavioral sensors. We were part of the Human Dynamics group, which is directed by professor Alex (Sandy) Pentland, and has been one of the pioneering labs for wearable computing, mobile sensing, and projects like Google Glass.

    We initially developed our specialized mobile sensing software for our own research experiments (mainly the “Friends and Family Study“). In this study we collected data from a community of 150 users for over a year, in one of the largest mobile experiments done in academia to date. We collected over 1 million hours of research data, and with it we are able to see how friendships form and evolve, and how people influence each other, in our effort to understand what makes us human. We then decided to re-architect our software as a modular and extensible software platform that others can use and build on. We named it Funf, short for “Fun Framework”, with a hint to the original “Friends ‘N Family” study. (Click the image below to enlarge the Funf flow diagram.)


    We released a very preliminary “alpha” version of our software in October, and have been blown away by the hundreds of developers already digging into it and building apps with it. For example, the One Laptop Per Child (OLPC) project is currently using Funf in Africa as part of a study to understand how children can learn to read using only technology, in an area where there are no schools.


    With the Knight News Challenge funding, we will be continuing to develop our open-source project along with instructional and educational material and a hub for developers and the user community. Expect different kinds of outreach, like hackathons, workshops, and talks, to connect developers and researchers from a wide spectrum — from computer scientists to social scientists and practitioners. We also hope to rally other companies and organizations around the open ecosystem vision, and show them the value of contributing to it.

    The old proverb says: “Give a man a fish and you feed him for a day. Teach a man to fish and you feed him for a lifetime.” Our goal is not to build all of the possible apps and services, but help others build amazing new apps and services that would be very hard for developers, researchers, and non-profits to do on their own.

    Nadav Aharony is co-founder and CEO at Behavio. He completed his PhD at the MIT Media Lab’s Human Dynamics group, where he investigated the use of mobile phones as social and behavioral sensors, conducted one of the largest mobile data experiments done in academia, and initiated the open source mobile sensing platform that became Funf.org. Aharony was a fellow for three years at the MIT Center for Civic Media, working on mobile and social activism topics. He holds Ph.D. and M.S. degrees from the MIT Media Lab, and a bachelor’s degree in electrical engineering cum laude from the Technion-Israel Institute of Technology. Aharony holds patents in social mobile networking, machine learning, network algorithms and sensor technologies.

    Tagged: behavior big data inference mit media lab mobile open source phone project intros sensing signals

    Comments are closed.

  • Who We Are

    MediaShift is the premier destination for insight and analysis at the intersection of media and technology. The MediaShift network includes MediaShift, EducationShift, MetricShift and Idea Lab, as well as workshops and weekend hackathons, email newsletters, a weekly podcast and a series of DigitalEd online trainings.

    About MediaShift »
    Contact us »
    Sponsor MediaShift »
    MediaShift Newsletters »

    Follow us on Social Media