Specific Parameters:

- Customer locations as points
- Customer demand in kWh per day at each location
- Passable obstacles as polygons
- Unpassable obstacles as polygons
- Roads as lines

General Parameters:

Number of leaves per demand kWh/day - a leaf is a pole with solar panels on top Type of root per range of demand kWh/day - there will be more than 5 types of roots (battery houses) and a specific root type will be able to serve a range of customers Max service drop length - a service drop is the cable that connects a customer to the distribution cable Max number of service drops per pole Max distance between poles Max distance between Wifi access point and Wifi mesh node Max distance between Wifi mesh nodes Max distance between Wifi mesh node and PV (solar inverter) Number of customers per street light - across an entire site, there will be a specified ratio defining the number of street lights to service the total number of customers Cost of a pole Cost per meter length of distribution cable

Algorithm outputs:

Pole Schedule: Pole ID Site Name Pole Type (one of two types) PV (yes/no) Customers (quantity) Street Light (yes/no) Wifi (yes/no) Corner (yes/no) Longitude Latitude

Root Schedule: Type Longitude Latitude

Customer Schedule: Customer ID Demand (kWh/day) Longitude Latitude Distribution Schedule (Roy can advise if there’s a better way to add up distribution length & connect nodes): Pole ID start Pole ID end Length

Service Drop Schedule: Pole ID start Customer ID end Length

- Maximize the percentage of customers who are within a maximum distance to a service drop pole.
- Minimize the number of service drop poles.
- Minimize the distance from a customer to the nearest service drop pole.
- Limit the maximum distance from a customer to the nearest service drop pole.
- Limit the maximum number of customers per service drop pole.

- Maximize the percentage of service drop poles connected by distribution cable.
- Maximize the percentage of distribution cable that follows a road.
- Minimize the distribution cable length.
- Limit the maximum distance between poles.
- Do not put the distribution cable across unpassable obstacles.

- Minimize the distance of an electric root to each service drop pole.
- Minimize the number of electric roots.
- Limit the maximum distance of an electric root to a service drop pole.
- Do not put an electric root inside a passable obstacle or unpassable obstacle.

- Estimate the daily demand (kWh / day) for each electric root.
- Estimate the number of solar panels needed to satisfy the daily demand for each electric root.
Place a solar panel on poles closest to an electric root.

Place a power meter on each service drop pole.

- Place a power meter on each street light pole.
QUESTION: If a pole has both a service drop and a street light, how many power meters does it need?

Place a wifi root at each electric root.

- Minimize the number of wifi poles.
- Minimize the distance from a solar panel to a wifi pole or wifi root.

- Minimize the summed distance of a customer to a street light over all customers.
- Place a street light on selected poles to satisfy the above constraint.

- Use an 8 meter pole when the angle of the incoming and outcoming distribution line is smaller than 135 degrees
- Use an 8 meter pole for solar panel poles
- Use an 11 meter pole otherwise

EXAMPLE Algorithm Inputs (rarely changed): Demand per customer type: Type 1 customer: 0.5 kWh/day Number of leaves per demand kWh/day - 3 kWh/day satisfied by each leaf Type of root per range of demand kWh/day - Type 1: < 20 kWh/day Type 2: 20 to 50 kWh/day Type 3: 50 to 100 kWh/day Type 4: 100 to 150 kWh/day Type 5: >150 kWh/day Max service drop length - 20m Max number of service drops per pole - 4 Max distance between poles - 50m Max distance between Wifi access point (root) and Wifi mesh node - 1km Max distance between Wifi mesh nodes - 500m Max distance between Wifi mesh node and PV (solar inverter) - 50m Number of customers per street light - 4 Cost of a pole - $1000/pole Cost per meter length of distribution cable - $5/m Notes Consider Z axis