NOU CAMP – Sport facility management and maintenance
This testbed focuses on the management of facilities involving the flow of large crowds, both during normal operation and during maintenance and construction projects. The Nou Camp digital twin study applies Machine Learning and Agent-Based Modeling for pedestrian simulation.
Current crowd management systems are not capable of seizing large parallel computational power, and their usability for rapid question answering is limited. This pilot will be performed during the renovation of Nou Camp Nou, the home stadium of Football Club Barcelona – the largest sport facility in Europe with a capacity of almost 100 000 seats. FCB will reconvert all the area and facilities into the best sporting and entertainment complex in the world. The strategic plan not only includes the renovation of the football stadium and expanding the capacity to up 105.000 spectators, but also opening all the private area around the stadium (28.000m2) to all public while integrating it harmoniously with the neighbourhood. To achieve this ambitious goal, we integrate several technologies: an IoT infrastructure, high performance computing, machine learning and big data analytics and modelling.
This testbed aims to analyze how crowds move both historically and in real-time using a robust IoT and big data infrastructure to collect, transmit and process data in real-time. The models developed on the basis of data will be optimized towards preserving usability and safety of the building for visitors, and for aiding the decision process concerning the location of the flows of equipment, machines and workers at each construction phase.
Demonstrate the feasibility of integration of IoT, Machine Learning, Big Data, and simulations into a production-level environment involving large flows of people
Optimize pedestrian flow by using Machine Learning and Agent-Based Modelling techniques into production-ready pedestrian flow predictors for weekly safety assessment and construction design
Improve code performance (e.g., reduction of memory usage and time) of data pipelines and Agent-Based Modelling for elastic cloud resources
Use the analytics and insights gained to minimize disturbance of construction activities to ormal operation of the facility: as well as to optimize normal operations
Develop a cloud twin that enables the generation of real-time evacuation plans in large facilities, with a more lightweight edge twin for local adaptation and refinement of evacuation instructions
To preserve usability and safety of the facilities for visitors, and for aiding the decision process concerning the location of the flows of equipment, machines and workers at each construction phase
To enhance customer experience and satisfaction by improving the location of services, commercial and leisure stands/shops