Periodic Reporting for period 3 - CrowdDNA (TECHNOLOGIES FOR COMPUTER-ASSISTED CROWD MANAGEMENT)
Reporting period: 2023-03-01 to 2024-10-31
Crowd management is a difficult task. Large crowds gathering or heavy pedestrian traffic are events of serious concern for officials tasked with managing public spaces. In practice, crowd management is essentially performed without support of dedicated technologies. Some management scenarios can be studied from simulation but require simulation parameters that can be difficult to estimate. The EU-funded CrowdDNA project proposes a new technology based on innovative crowd simulation models. It facilitates predictions on the dynamics, behaviour and risk factors of high-density crowds, addressing the need for safe and comfortable mass events. The project suggests that the analysis of some specific macroscopic characteristics of a crowd such as its apparent motion can offer important information about its internal structure and allow the exact assessment of its state.
CrowdDNA is a radically new concept to assist public space operators in crowd management, i.e. mass event organisation, heavy pedestrian traffic management, crowd movement analysis and decision support. The CrowdDNA technology is based on a new generation of crowd simulation models capable of predicting the dynamics, behaviour and risk factors of extremely dense crowds. The basic idea behind CrowdDNA is that the analysis of some specific macroscopic features of a crowd, such as its apparent motion (which can be easily measured in real crowd events), can reveal valuable information about its internal structure and provide an accurate estimate of a crowd's state. This challenges existing paradigms in the field, which rely on simulation techniques and require the measurement of simulation variables to initialise them, such as density, number or individual features, each of which is difficult to estimate. This vision raises a major scientific challenge, which can be summarised as the need for a deep understanding of the relationships between the smallest scales of crowd behaviour (e.g. contact and shoving at the limb scale) and the largest up to the whole crowd. CrowdDNA is the first attempt to combine biomechanical and behavioural simulation in complex scenarios of crowd interaction. The project proposes technological solutions the practice of crowd management to meet the demands of modern society for safety and comfort at mass events or in crowded transport facilities. CrowdDNA lays the foundations for new research on crowds and opens up new opportunities for studies of physical interaction in cognitive science and biomechanics, as well as robotics and autonomous vehicles for safe navigation among people.
CrowdDNA is a radically new concept to assist public space operators in crowd management, i.e. mass event organisation, heavy pedestrian traffic management, crowd movement analysis and decision support. The CrowdDNA technology is based on a new generation of crowd simulation models capable of predicting the dynamics, behaviour and risk factors of extremely dense crowds. The basic idea behind CrowdDNA is that the analysis of some specific macroscopic features of a crowd, such as its apparent motion (which can be easily measured in real crowd events), can reveal valuable information about its internal structure and provide an accurate estimate of a crowd's state. This challenges existing paradigms in the field, which rely on simulation techniques and require the measurement of simulation variables to initialise them, such as density, number or individual features, each of which is difficult to estimate. This vision raises a major scientific challenge, which can be summarised as the need for a deep understanding of the relationships between the smallest scales of crowd behaviour (e.g. contact and shoving at the limb scale) and the largest up to the whole crowd. CrowdDNA is the first attempt to combine biomechanical and behavioural simulation in complex scenarios of crowd interaction. The project proposes technological solutions the practice of crowd management to meet the demands of modern society for safety and comfort at mass events or in crowded transport facilities. CrowdDNA lays the foundations for new research on crowds and opens up new opportunities for studies of physical interaction in cognitive science and biomechanics, as well as robotics and autonomous vehicles for safe navigation among people.
The CrowdDNA project is based around 4 main activities that represent the different pillars of a crowd management assistance technology.
The first pillar of the project concerns data. The project has acquired a very large number of innovative data sets relating to dense crowds, which for the first time make it possible to understand the physical interactions that occur between individuals. The project has collected for the first time detailed whole-body motion data for a set of individuals forming a compact group in a variety of scenarios.
The second pillar of the project focuses on the modelling and simulation of dense crowds. The project
explores innovative models which, like the data collected, can simulate the details of physical interactions between individuals in dense crowds. These models make it possible to reproduce the dangerous phenomena observed in real crowds and to study them in silico, but also to produce valuable synthetic data sets for training crowd movement analysis methods.
The third pillar explores methods for analysing crowd movement images and mobility data to predict crowd behaviour and density variations, or to detect the warning signs of potential dangers associated with dense crowd movements.
The fourth and final pillar of the project establishes invaluable links between the academic world and the world of crowd management professionals, with a number of objectives: to demonstrate and apply the project's technologies in the field, to disseminate the knowledge produced, to provide case studies and data to validate our approaches and, finally, to help bring these two worlds, which are still too far apart, closer together.
The first pillar of the project concerns data. The project has acquired a very large number of innovative data sets relating to dense crowds, which for the first time make it possible to understand the physical interactions that occur between individuals. The project has collected for the first time detailed whole-body motion data for a set of individuals forming a compact group in a variety of scenarios.
The second pillar of the project focuses on the modelling and simulation of dense crowds. The project
explores innovative models which, like the data collected, can simulate the details of physical interactions between individuals in dense crowds. These models make it possible to reproduce the dangerous phenomena observed in real crowds and to study them in silico, but also to produce valuable synthetic data sets for training crowd movement analysis methods.
The third pillar explores methods for analysing crowd movement images and mobility data to predict crowd behaviour and density variations, or to detect the warning signs of potential dangers associated with dense crowd movements.
The fourth and final pillar of the project establishes invaluable links between the academic world and the world of crowd management professionals, with a number of objectives: to demonstrate and apply the project's technologies in the field, to disseminate the knowledge produced, to provide case studies and data to validate our approaches and, finally, to help bring these two worlds, which are still too far apart, closer together.
Project's results and impacts concern:
- Scientific contributions to a new generation of crowd simulators: new avenues for crowd research and transposable principles from the physical modelling of local interactions to other research fields (robotics, transport and sociology).
- Scientific contributions to the study of physical interactions: new datasets and models which have no equivalent yet, development of new methods for the acquisition of data and a better understanding of human behaviour to enlarge the horizon of Cognitive Sciences.
- Technological contributions: a prototype for computer assisted on-line management of mass events, a prototype for computer simulation of mass events and the concept of “Living crowd observatories” to demonstrate prototypes.
- Social Impact: increase the level of safety and comfort of mass events and high, accurately predict crowd movement and help to identify increased risk and potentially test management strategies.
- Economical Impact: better operational management, increased reliability and reduced costs, optimization of the quality of services offered by crowded public places and a new generation of more realistic crowd simulators.
- New market opportunity: put EU in a leadership position with regards to crowd management.
- Scientific contributions to a new generation of crowd simulators: new avenues for crowd research and transposable principles from the physical modelling of local interactions to other research fields (robotics, transport and sociology).
- Scientific contributions to the study of physical interactions: new datasets and models which have no equivalent yet, development of new methods for the acquisition of data and a better understanding of human behaviour to enlarge the horizon of Cognitive Sciences.
- Technological contributions: a prototype for computer assisted on-line management of mass events, a prototype for computer simulation of mass events and the concept of “Living crowd observatories” to demonstrate prototypes.
- Social Impact: increase the level of safety and comfort of mass events and high, accurately predict crowd movement and help to identify increased risk and potentially test management strategies.
- Economical Impact: better operational management, increased reliability and reduced costs, optimization of the quality of services offered by crowded public places and a new generation of more realistic crowd simulators.
- New market opportunity: put EU in a leadership position with regards to crowd management.