Mobility Modeling and Analytics for Traffic Management Applications

Introduction

Vast amounts of data from multiple type of sensors allowing observation of actual trips and point flows enables a fundamentally new understanding of travel demand/mobility in real-time. At the same time, the need for better real-time traffic management support systems is growing in all big cities around the world.

The aim of this PhD is to combine methods in machine learning with traffic modeling to make short-term predictions of travel demand to support traffic management applications.

Project

The purpose of the project is to utilize an existing platform for real-time traffic modelling and estimation to develop a real-time decision supprt tool for traffic management centers. The platform has been developed jointly by Linköping University, Royal Institute of Technology, UC Berkeley, the Swedish Transport Administration and the traffic management center in Stockholm, Trafik Stockholm. Your focus in the project will be on short-term prediction of travel demand. The methodology will combine data-driven methods from machine learning with real-time traffic modelling and include data assimilation and fusion.

You can expect to work with large amounts of real-world city-level sensor data from different sensors, such as radar sensors, license plate recognition data, mobile network data, GPS probe data, bluetooth and wifi detectors.

Research partners

The project is funded by Trafikverket. Project partners are KTH and Trafik Stockholm. The research group has close collaborations with UC Berkeley and UPC Barcelona in the area of traffic management.

PhD position

If you have an interest in mobility, traffic modeling and machine learning we welcome applicants with several engineering degrees, such as electrical engineering, computer science, civil engineering and industrial engineering.

We think you have a solid background in mathematics and are comfortable in implementing algorithms in at least one programming language.

Experience in traffic modeling or machine leraning is an advantage, but not a requirement. Apply by filling in the application form linked here»

More information

Please contact either David Gundlegård or Clas Rydergren to get more information about the project and the position.