Thursday, June 5, 2014

Data from smart phones

In my last post, we looked at the 5 steps needed to go from data to insight. I think of the first step, Get the Data, as really the fuel that powers everything else. Data's ever-increasing availability and variety of form is spawning the innovations introduced in the remaining four. What good is a database if you have nothing interesting to store in it? Not much. Why bother with a Hadoop processing system if you are working with data that fits nicely into one Excel workbook? You wouldn't.

The cool, headline-grabbing technologies are there because data is exploding. So where specifically is this stuff coming from?

Our mobile devices throw off an incredible amount of information. Traffic in Google Maps works by crowd-sourcing location information telecommunication companies pull from phones. While not every person stuck in traffic along the Capitol Beltway is a proud owner of an Android, enough of these devices are out there so that Google knows when to paint I-495 a heartbreakingly dark shade of red.

Similarly, phone-generated location data in developing countries is helping the medical community manage malaria outbreaks. By understanding patterns of movement across at-risk regions, public health professionals can craft better communications and target areas to spray with more precision.

When our devices aren't passively generating data from our everyday motions, they are generating data based on actions we take. As an avid bike rider, I am a regular user of the Strava Bike App. The app allows me to track my ride and easily share how much slower I shredded Skyline Drive in the Shenandoah Valley than the 300 people ahead of me. Every time I push the record button, I am generating information that is useful not just to me the individual, but also to transportation planners looking to re-think city infrastructure, and companies looking to market to active types such as myself.

For each activity, Strava records my ride with an incredible amount of detail. I get great summary information like average or max speed, vertical feet climbed, and calories burned. If I want more detail, I can get a graph like the one below plotting speed on the y-axis versus mileage on the x. I can even playback my ride to see at which points on the map I achieved what speeds and what elevations.


This is great for me, but also great for Strava and hopefully the rest of the United States' growing cycling community. Oregon's Department of Transportation recently purchased Strava data from Portland for $20,000 and is using it to understand bike traffic at a level of detail that traditional methods such as traffic surveys couldn't approach. Every time I record a ride through Strava, I am contributing to a body of evidence that city planners may one day use to re-think sketchy intersections and head-spinning traffic circles. On that subject, points to whoever can explain the circle inside a circle-type intersection near American University!


Smart phones are just one of a handful of technologies tuned to record the world around them and make data available for later use. From a data privacy perspective, the problem is that data from these devices is at the individual level. So without proper care or intent, it becomes very easy for another person or organization to learn about you the individual - where you spend your time, what you like to eat, what you spend your money on - rather than the group you belong to - your generation, your income bracket, your likely political leanings. In data speak, we have moved from inference to here is what is actually going on with this person.

On a personal level, I am happy to share my rides data with Strava to the extent that they use it to make biking better and safer. But I am still working through how I would feel if they began selling my information so outdoor companies could improve their marketing to me. I find a lot of the personalized marketing I receive now to be annoying because it is ill-timed and imprecise. But suppose a bike company got it right and threw an offer for the perfect mountain bike at a time when I was considering an upgrade. Would I really turn it down?




No comments:

Post a Comment