Workload Assessor

This repository contains the code module for real-time Mental Workload (MWL) detection from pilot control behavior using behavioral entropy. The code implements the proposed HTM-WL metric along with other established measures such as Fessonia, DNDEB, and a Naïve Forecaster. It also includes a real-time MWL spike-detector and evaluation criteria for detection lag and precision.

Repository Structure

• source/pipeline/run_pipeline.py: Main script to execute the data processing pipeline.
• configs/run_pipeline.yaml: Configuration file for the pipeline.
• src/: Additional modules and helper functions.
• requirements.txt: Python dependencies.
• README.md: This file.

Requirements

• Python 3.7 or higher.
• Required packages (see requirements.txt):
  • numpy
  • pandas
  • scikit-learn
  • tensorflow (or the appropriate ML library for HTM)
  • matplotlib
  • (and others as needed)

To install the dependencies, run:

pip install -r requirements.txt

Running the Pipeline

1. Clone the Repository:

   git clone https://github.com/gotham29/workload_assessor.git
   cd workload_assessor

2. Configure the Pipeline:

   Open the configuration file at configs/run_pipeline.yaml and update the following fields:

   • input_dir: Set this to the directory containing your input data (e.g., the NASA dataset provided as supplementary material).
   • output_dir: Set this to the directory where you want the output results to be saved.

3. Execute the Pipeline:

   Run the pipeline by calling the main script with the configuration file:

   python source/pipeline/run_pipeline.py --config_path configs/run_pipeline.yaml

Data Availability

The NASA dataset used in this study is provided as supplementary material with the submission. Please ensure that the dataset is placed in the directory specified by the input_dir parameter in configs/run_pipeline.yaml.

Output

After running the pipeline, the output directory will contain:

• Processed data files.
• Plots of detection lag and precision distributions.
• Performance metrics for the evaluated MWL measures.

Troubleshooting

• Missing Dependencies:
  If you encounter errors regarding missing packages, please run:

  pip install -r requirements.txt

  to install all necessary dependencies.

• Path Issues:
  Verify that the paths specified in configs/run_pipeline.yaml are correct and accessible.

• Error Messages:
  Check the terminal output for any error messages; they may provide guidance on resolving issues.

License

This project is licensed under the MIT License.

Contact

For questions or contributions, please contact Sam at [email protected]