This Proof of Concept application that attempts finding a "optimal" fiber network planning for a map area.

The application uses OpenStreetMaps data and parameter provided by the operator to optimize the Fiber-to-the-Home underground cables routes for cost savings. The model is designed consider factors such as:

• Which side of the road would cost less to dig-up
• The extra cost of digging across the road
• Obstacles such as bridges, rivers, monuments
• Cost of fiber per KM
• Cost of trench per KM
• Cost of trench to cross the road
• Cost of equipment
• Limits of PON/GPON networks
• Cost of hardware (Street cabinets and new powered decentralize locations)

The model should output a detailed planning along with itemized material and labour estimates.

Relationship Diagram

conda config --prepend channels conda-forge
conda create -n ftth_planner --strict-channel-priority osmnx

streamlit run ftth_app.py

If you have a idea on how to fix any bug, please send me a pull request.

• OpenStreetmap data often doesn't contain building address data, so it il not work in areas that have less than 96 addresses registered

• If building don't have address information we would of course look for the closest trench, but this would be computationally expensive to do for all buildings and all trenches

• Buildings with address information sometimes do not find the closest trench, not sure why