Londoners spend on average over 6 full days waiting in traffic each year, we attempted to tackle this for our project.

What it does

Our solution enables data-driven decisions to tackle traffic disruptions and, ultimately, keep London moving. It does this by integrating with a variety of data sources, from live bus arrivals from TfL to tweets related to traffic disruptions from Twitter. Crucially, these have been setup such that they can be updated in near-to-realtime, as time is of the essence when managing traffic. This drives an operational application that brings together all these data feeds into one place for analysts to consume. This is augmented by modelling that helps users understand the impact of actions they are going to take, principally diverting bus routes.

How we built it

The solution required us to integrate data from a variety of different sources that each return data in a different format. We were able to leverage Palantir Foundry to integrate the data and build a data model that could support the application we wanted to build. This involved setting up the data connections, building robust and speedy data pipelines and crafting a suitable ontology.

We then used the phantom traffic formula to enable users to simulate the effect of actions they are going to take. We had to build data pipelines to ensure we had the right data for the simulations to run.

Built With

Share this project: