An Open Interactive Tool for Applying Interdependence Analysis in Human–Autonomy Teaming
The Interdependence Analysis (IA) Dashboard is an open-source web application built with Plotly Dash that supports the practice of Interdependence Analysis (IA), a core method in Coactive Design for engineering effective Human–Autonomy Teams (HAT).
The dashboard provides an interactive environment for loading, visualizing, and editing IA tables.
The tool is designed to facilitate the jump from theory to practice: rather than working through IA tables in spreadsheets, users can directly interact with a purpose-built interface.
- Parametric team configuration — auto-detects team structure and agent roles from any CSV.
- Interactive IA table editor — edit coordination levels cell-by-cell directly in the browser
- OPD requirements panel — structured input fields for specifying observability, predictability, and directability needs per task
- Capacity charts — aggregated bar charts summarizing capacities to execute tasks by agent and by color level
- Interdependence pie charts — visual overview of the interdependence level across the joint activity
- Automation proportion metric — computed according to Liu & Kaber (2025)
- CSV import / export — load your own IA table; export the edited version at any time
- Bundled example — ships with a sample IA table (
V7/IA_V7.csv) based on an smart-electronic-checklist in an aviation scenario
Prerequisites: Python 3.9+
# Clone the repository
git clone https://github.com/Bertonlome/Interdependence-Analysis.git
cd Interdependence-Analysis
# Create and activate a virtual environment (recommended)
python -m venv .venv
source .venv/bin/activate # Windows: .venv\Scripts\activate
# Install dependencies
pip install dash pandas plotly
# Run the app
python python_dash_generic.pyOpen your browser at http://localhost:8050
A production-ready Docker configuration is provided under hf_deployment/.
cd hf_deployment
docker build -t ia-dashboard .
docker run -p 7860:7860 ia-dashboardRefer to hf_deployment/README.md for full instructions on deploying to a Hugging Face Space.
The dashboard auto-detects team structure from any CSV that follows this convention:
| Column | Description |
|---|---|
Row |
Row index (optional) |
Procedure |
Procedure or phase label |
Task Object |
Task or sub-task or required capacities description |
AgentName* |
Performer column — column name ending with * |
AgentName |
Supporter column — same base name, no * |
OPD Requirements |
Free-text field for coordination requirements |
Color values in agent columns must be one of: green, yellow, orange, red.
The bundled example (V7/IA_V7.csv) provides a reference for structuring your own IA table.
This tool is part of the Coactive Design process developed by Johnson, and aims to guide designer of human-autonomy team beyond the substition-based automation philosophy, by focusing on interactions that promotes mandatory and optional dependencies between team-members
The IA also lead the designer to ask, for every task and every team configuration alternative:
What needs to be observed from whom, what needs to be predicted from whom, and what needs to be directed to whom?
The color coding scheme:
| Color | Meaning for Performer | Meaning for Supporter |
|---|---|---|
| 🟢 Green | I can do it all | My assistance could improve efficiency |
| 🟡 Yellow | I can do it all but my reliability is < 100% | My assistance could improve reliability |
| 🟠 Orange | I can contribute but need assistance | My assistance is required |
| 🔴 Red | I cannot do it | I cannot provide assistance |
The platform is under active development.
Planned improvements include:
- Adding sequence constraints and what-ifs permutation
- Integration with task network modeling tools
- Export to structured report format (PDF / LaTeX)
Feedback from researchers, and practitioners is welcome.
If you use this software please cite:
Berton, B., & Doyon-Poulin, P. (2026). Interdependence Analysis Dashboard (Version 1.0) [Software]. Zenodo. https://doi.org/10.5281/zenodo.XXXXXXX
This project is released under the GNU-GPL v3 License.
See the LICENSE file for details.
Contributions are welcome.
Benjamin Berton - benjaminberton64@gmail.com
PhD Candidate – Cognitive Engineering
Polytechnique Montréal