D4.1: Report on Crowd Observatories

Executive Summary

This document is a redacted version of deliverable D4.1 to enable it to be unrestricted. A restricted, confidential version of D4.1 has also been produced which contains confidential information linked to the crowd observatories and is available on request to the project coordinator by project officer and project reviewers.

This report focuses on the first set of observatories for CrowdDNA, looking specifically at the movement of people in an environment where the movement is uninhibited by controlled experiments. The intention is to create a data set for predicting crowd movement which once applied will enhance crowd safety. Through the identification of a range of event types through categorisation, a target list if observatories were contacted. Due to the annual nature of festival season, the focus was placed on pursuing festival observatories to maximise the number of festival seasons for observation throughout the duration of the CrowdDNA project. A systematic approach was taken to secure the observatories.

Scenarios and types of data to be captured were specified for each observatory type prior to attending, with specific ones being relevant to the festivals. This helped to inform what equipment would be required for data capture and the methodology as to how to capture it. The chosen approach was a combination of photo, video, body motion, Xsens suits, IoT tracking, mobile phone tracking and camera counters. The data capture methodology posed some production considerations and presented the requirement for a common framework to ensure GDPR compliance, as well as legal and ethical considerations.

The installed observatories took place between the 17th June and 20th August 2022, with one festival in France (Hellfest) and three in the UK (Tramlines, Bloodstock and ArcTanGent), all with varying capacities and curation. For each observatory, a detailed account of the data capture process is documented within this report, including maps, photos, and video file references. Observations made on site at each of the festivals has facilitated a feedback report to the owners and operators based on findings (issued separately to this report) but included in bullet point form within this report.

The data captured at each of the observatories may under-go further analysis than what is included in this report to generate algorithms for predicting crowd movement and in turn, increase crowd safety. Learnings from participating in the data capture exercises at observatories can be applied to all observatories moving forward and if required, more extensive data capture can take place at festivals in the 2023 season.