Problem:
The Eiffel Tower receives millions of visitors each year, causing long queues and crowding, especially near the elevators and top floors.
Solution:
Engineers built an AnyLogic simulation model that reproduced visitor arrivals, movement, and elevator operations.
Results:
- Simulated visitor behavior during peak days.
- Estimated queues and congestion points.
- Showed that redesigning floors would not fix the issue.
- Identified better visitor flow within the existing layout.
- Helped improve crowd management without major construction.
To efficiently manage visitors at the most popular tourist site in Paris, the Eiffel Tower, the operating company (SETE) used AnyLogic simulation software from The AnyLogic Company.
The model designed by the software simulates tourist arrivals and their behavior in open spaces. The program estimated queues in front of elevators, taking into regard elevator operation rules and capacity. The main issues of waiting lines and crowd management were on the Tower’s top floor.
AnyLogic has an Agent Based simulation modeling tool, which means it actually simulates the individual behavior of each person (in this case, visitors at the Eiffel Tower). The project members said the Eiffel Tower had 6.7 million tourists in 2010. 27,000 program agents are in attendance on the days of highest visitor traffic. The Tower model with its interior and visitor flows was created by the French EMSYSS consulting company. AnyLogic Europe also took part in the project.
The AnyLogic Company Business Development Director, Timophey Popkov, reported that the project took 30 business days to be completed. “Within that time we gathered data to create a Tower model, took photo and video footage of the visitors’ behavior, and looked into the project to reconstruct the 3rd and 4th floors of the Tower,” he said. The second stage included the model designs. As a result of various experiments, the engineers came to realize that new designs would not solve the problem of crowd management, and thereby would not be reasonable in terms of investments. “The model determined that the new designs would not change the passability issue,” said Popkov. “Finally, we experimented with the model to try and optimize the existing visitor itineraries.”