The project concerned a high school in Eindhoven, the Netherlands. Walking patterns of 1200 students were simulated on an average schoolday.
The goal of this project was to provide the school with appropriate advice on how to reopen at the end of the intelligent lockdown while keeping the risk of spreading Covid-19 to a minimum. Which measures should the school implement to protect both student and teacher while retaining the ability to give quality education? We used a digital twin simulation model to answer this question.
A digital twin is a simulation of an existing environment with moving agents, in case a high school with moving students and teachers. These agents move during the day through the hallways from classroom to classroom. All agents react to one another, meaning that a busy hallway or door could lead to congestion.
In order to reach an appropriate advice, we started with establishing a baseline that would mirror the situation of the school before corona and the introduction of physical distancing. The student and teacher behavior was implemented in the simulation by uploading class schedules provided by the school. In an evaluation meeting with the school board we got confirmation that the flow of agents matched the flow of students and teachers in real life and the baseline was confirmed
By simulation the current state, only adding the individual desire to keep 1.5 meter distance, it was quickly concluded that this would not provide a feasible solution: Hallways become extremely overloaded and there appeared congestion that did not resolve, causing agents to not reach their classrooms on time. Extra measures were needed.
In cooperation with the school we created different scenarios that combined a selection of measures. Some effective measures were: Opening emergency exits, partly digital lectures and the introduction of one-way traffic in hallways. Every scenario was modeled in the simulation and analyzed separately.
The performed experiments were evaluated through the following analyses:
- A heatmap showing in which locations the desired 1.5 m distance was most difficult to maintain and for how long.
- A walking time analysis to see how many agents would be late for their classes in every scenario.
- A graphic display in video format.
In combination these analyses showed which scenario would be the best option.
At the end of the project the advice was given to open certain emergency exits to increase flow and make one-way traffic possible in every hallway of the school. Several measures were also advised to reduce traffic during the lesson changes. Finally, keeping the number of physically present students low and having teachers change classroom instead of students were essential measures.
“The project made me feel like playing a real-time strategy (RTS) game”
Berend Steenhuisen – Logistic Optimization Engineer
“Interesting to be able to apply our knowlegde of logistics and simulation in the context of safety.”
Dreas de Kerf – Industrial Engineer