Evaluating Attention Management Systems for Dynamic Monitoring Tasks

This repository contains the source code used in the paper "Evaluating Attention Management Systems for Dynamic Monitoring Tasks". The project investigates how various attention management strategies affect human performance in dynamic monitoring environments, such as those encountered in safety-critical systems or real-time supervision tasks.

The experiment simulates a simplified dynamic monitoring task environment (Senders Clock Task) and evaluates how different system-driven cues impact user behavior and task effectiveness.

Getting Started

To run the experiment locally:

1. Clone the repository

git clone https://github.com/ciao-group/Attention-Management-Systems-for-Dynamic-Monitoring-Tasks.git

2. access the repository

cd Attention-Management-Systems-for-Dynamic-Monitoring-Tasks

3. Install the required packages

pip install -r requirements.txt

4. Run the Simulation

python3 game.py

Configuration Parameters

Below is a list of key configurable parameters used in the experiment environment:

• Screen Size: Resolution of the game window (e.g., 1920x1080)
• Distance Between Clocks: Pixel distance between clock centers
• Radius of Clocks: Radius in pixels of each circular clock interface
• Timestep Length: Length of a timestep (e.g., 0.02)
• Game Length: Total duration of one trial (e.g., 90s)
• Crossing Time Tolerance (CTT): Time window after a crossing during which a pressing is still counted as correct (e.g., 0.5s)
• AMS Crossing Tolerance: Distance threshold between the clock target and marker for triggering an AMS cue (e.g., 100 px)
• Buttons for Activating/Deactivating Clocks 1–6
• Set Bandwidth for Clocks 1–6: Adjusts clock speed or task difficulty
• Choosing AMS System: Selection of attention management strategy
• Logging Mode: Controls data collection mode