
- Client :
- Category :Optimization, Pedestrian
- Project Url :
- Date :February 4, 2025
Military Gym Layout Optimization
The Challenge
This project focused on analyzing people flow inside a gym facility to compare two different layouts and identify bottlenecks that could impact movement efficiency and thus delay the transitions between different areas in the location. The facility accommodates groups of 150 to 200 people who transition through four workout zones to do a workout routine before exiting. After a certain amount of time in each zone, people move to a new zone going around obstacles such as workout stations, machines, and walls.
![]() | ![]() |
Understanding how different layouts influence flow efficiency was critical in determining the best design for minimizing delays between zones. The key challenge was identifying congestion points and evaluating the overall throughput of each layout based on movement patterns.
The Solution
A simulation model was built in AnyLogic to represent the gym facility in 2D. The Pedestrian Library was used to capture individual movement dynamics and interactions with obstacles and other people.
The model was built in a flexible way such that the user can control the input parameters such as the number of people, time at each station, movement speed, as well as the layout to be used from an excel sheet. Moreover, the model had enough flexibility to incorporate additional layouts for testing if needed.
We used the simulation model to generate raw data that was used to find the throughput and the bottlenecks.
Throughput
To obtain the throughput we wanted to know how much time it takes for each individual to move from their initial position in one zone to their final position in the next zone. The information was exported on an individual level to perform population analysis.
Bottlenecks
To find the bottlenecks, we generated a set of uniformly distributed squares in the layout. Each time step, we calculated the density in the squares. Also, for each square we calculated the amount of time an individual spends in that square. A square with high density or where people spend more time than normal is a potential bottleneck square.
Results
The raw data exported from the model was used to generate different dashboards in Power BI. The figures below show some results related to one of the layouts.
The client was able to test layout variations and make data-backed decisions before implementing physical changes, leading to a more efficient and user-friendly environment.