Tracking Activity at the 2022 World Cup Stadium Sites in Qatar

Qatar’s World Cup stadiums have received a lot of publicity ranging from the innovative cooling technology to the deaths of migrant workers due to labor conditions at the stadium construction sites. Additional controversies include the FIFA debate over adding an additional host country to mitigate the negative publicity associated with the 2020 World Cup or to remove the hosting duties from Qatar entirely.

Orbital Insight’s GO platform leverages remote sensing and computer vision at scale to support infrastructure change detection and foot traffic analysis of areas of interest. The platform generates historical as well as near real-time data using entire catalogs commercial satellite imagery and other geospatial data sources. Users can examine change in various areas of interest over large spans of time in just a few clicks of the mouse.

We created a GO project to examine construction progress at the four stadiums through automated infrastructure change detection. The computer vision algorithms detected stadium progress on a near-weekly basis since construction began at these sites in 2014. We were able to see both historical and near real-time construction/infrastructure progress.

construction detected by the GO computer vision algorithms in 2016
The screenshot above displays construction detected by the GO computer vision algorithms in 2016, and is layered over a Mapbox base map from December 30, 2018 for illustrative purposes.

Qatar’s Al-Wakrah stadium is the first of the country’s four stadiums to be completed in advance of the 2022 World Cup. The stadium in Al-Wakrah was inaugurated when it hosted the Amir Cup Final on May 16th, 2019. To add more color to our dataset, we combined the vehicle detection and foot traffic algorithms to capture the spike in activity at the stadium for the Amir Cup game.

Al-Wakrah foot traffic
A screenshot of Orbital Insight’s GO platform capturing increased foot traffic at Al-Wakrah stadium during the grand opening.

Monitoring change detection over time is useful, but just remembering to look for changes can be its own cumbersome task. We used the GO platform’s alert feature to set up automated notifications that email us when changes or anomalies occur at our selected areas of interest. Users are able to set custom detection thresholds so they can be immediately alerted when observations drop below or rise above certain values.

Alerts can also be cued on a per-algorithm basis; e.g. ‘alert me when there are 50% more cars than usual’ and ‘alert me when there are more than 100 geolocation pings.’ This alerting feature frees up users to do other analysis, rather than actively monitor the play by play happenings of each particular project. For this project, GO sent us an alert when vehicle and foot traffic spikes occurred at the Al-Wakrah stadium.

The GO Platform’s catalog of computer vision algorithms can conveniently be creatively combined to generate a single enriched data set. In this case, the platform’s infrastructure change detection algorithm output provided information about construction progress, while foot traffic data generated insight as to how facilities are being used.

To explore your own creative algorithm combinations using the GO Platform, contact Orbital Insight: