• Add to Collection
  • About


    A computerized model showing the effects of GPS devices on traffic flow.
GPS and Traffic Aggravation
A graphical model to elucidate the flaws of a near-perfect product
In January of 2011, I took a class on complex systems at MIT. I was the only undergraduate in the whole class. My group, which was primarily made up of Math PhDs from England, decided to model the effects of global positioning systems on traffic.

This project was very heavy on math and computer science, but it has some very direct applications to product design and development. The first GPS was an ingenius solution to the problem of finding your way around a place you don't know. Using satellite technology and complex search algorithms, GPS devices have been able to direct drivers on the most efficient paths to their destinations. But as these devices become more widespread, a new problem has arisen:

wouldn't GPS devices' routing people to the same 'efficient' path make that path inherently inefficient?

We modeled this problem by running various scenarios with different concentrations of GPS agents, Naive agents (symbolized the average non-GPS user), and Taxi agents (symbolized drivers that know the streets well enough to navigate efficiently around traffic). We measured average congestion on each trial - the results are above the title - and found that the most congestion occurs when everyone is using GPS devices.

These results suggest that empirical product design research is very useful in assessing the effectiveness of a product. If I was to present this research to a company that specializes in GPS devices like Garmin, I would suggest including some mechanism in each device that tracks where other devices are around it as well as a routing algorithm that is unique to the device.