We Predict Human Behavior in Cities

Case Study 3

Bike Shares and Mobility


Understanding real-time demand. Even in cities we haven't been to.



The Challenge

Every time a customer opens a bike share application, unlocks a bike, rides a bike, and drops a bike off, that customer is sending location data back to the company. Using that data to make actionable decisions isn't as easy. How do you predict demand to minimize the distance between where your customers are opening your app and picking up a bike? How do you know where to redistribute bikes next week on Tuesday when it's overcast and 61 degrees? How do you know where to put your first three bike racks in a brand new city?

The Solution

Zoba brings together thousands of variables that show how a city works as a system. These data are both spatial and temporal in nature and allow us to look at historical events in the context that they occurred. 

The combination of our data and our algorithms with a bike share's spatial data, we can predict when and where customers will have demand for bikes on any timescale. And we can do that in new cities as well. 

Screen Shot 2018-05-09 at 11.39.09 AM.png

Screen Shot 2018-05-09 at 11.41.50 AM.png


By being predictive, bike shares can be more efficient with logistical redistribution, and guarantee that their expansion strategies will produce results in new cities.