An AI-driven project to simplify the preventative management of water main segments. We focus on two main components: prediction of health scores for pipe segments and the optimized scheduling of repairs.
Use Python version >= 3.8 and run pip install -r requirements.txt
.
├── data/ # Our datasets stored in csv format
├── graphs/ # SA and ML graphs
├── map/ # Demo map generation
├── models/ # ML models
├── prediction_and_survival/ # SA
├── research/ # SA and ML notebooks
├── scheduler/ # CSP and LP schedulers
- Run
map/main.py
- This will generate the Plotly map that contains all prone-to-breakage pipes with assigned properties to them
- Run
prediction_and_survival/ML.py
andprediction_and_survival/SA.py
- This will write the CSV output file in the
data/
folder
/scheduler/workforce.py
defines employees with work hours and tasks to be scheduled- The output of the scheduling tasks are interactive gantt charts that open in a new browser window
/scheduler/CSP/main.py
solves the scheduling problem using backtracking/scheduler/CSP/csp.py
contains the constraints
/scheduler/LP/main.py
solves the scheduling problem using pulp's linear programming capabilities