Active Projects

Big Data Analytics for Cities

Kharita: Map construction and update using crowd-sourced GPS traces

Kharita is map construction algorithm to build and update road networks using crowd-sourced GPS traces. Kharita is a result of a long-term collaboration between QCRI and QMIC.

People: Sofiane Abbar, Rade Stanojevic, Saravanan Thirumurugunathan, Sanjay Chawla (QCRI), Fethi Filali, and Ahid Aleimat (QMIC)

Publications

Softwares on Github

Input Data: Set of trajectories

Step 2: Inferring edges to connect clusters

Step 1: Clustering of GPS points

Step 3: Graph spanner to remove unnecessary edges

Unthule & RoadRunner: Map construction from aerial images and large scale GPS traces

Unthule (un-thule) is a system designed to build road networks from aerial images using deep-learning techniques. RoadRunner is another system for map inference that relies solely on GPS traces. These systems are the results of a collaboration between QCRI and CSAIL MIT.

People: Sofiane Abbar, Sanjay Chawla (QCRI), Hari Balakrishnan*, Favyen Bastani*, Songtao He*, Sam Madden*, Mohamed Alizahed*, David Dewit* (MIT CSAIL)

Publications

Software on Github

  • Coming soon

Road Tracer: Maps from Aerial Images

Road Runner (Doha)

RoadRunner + RoadTracer

Urban Mobility & City Resilience

This project aims at understanding mobility and resilience in cities around the world.

People: Sofiane Abbar, Abdelkader Baggag, Noora Al Emadi, Tahar Zanouda (QCRI), Javier Borge (CoSN3), Yanyan Xu, Marta Gonzalez (MIT)

Relevant stakeholders: QMIC, SCDL, Qatar Rail

Publications

Softwares on Github

  • Coming soon