Data Science for the World of Moving Things

Damon Wischik – Computer Laboratory, University of Cambridge
Wednesday 08 February 2017, 16:15-17:15
Lecture Theatre 1, Computer Laboratory.

Trains, buses, taxis, cars, deliveries. These exist in three networks
simultaneously: the physical network of roads and train lines, the data
network of sensing and analytics and control, and the social network of
people who make use of them.

I will tell the story of a data science startup, where my colleagues and I
spent five years making sense of these networks. We started with incentive
schemes, paying and persuading commuters to shift their habits. We moved on
to dashboards and full-resolution reconstruction of the transit network, to
see where there was congestion and whom we needed to target. We built a new
type of database, for rich and high-speed processing of spacetime data; and
we designed a visual querying language to make the system accessible to
urban planners and transit agencies. And, in the end, we joined the Google
Maps team.

We learnt some hard lessons about how to interact with data visually and
tangibly. I believe that over the next decade it will become easy to
understand and argue with big datasets, as easy as Excel is today; and I
will discuss the way ahead.