Solar Panel AI: How can we encourage citizens and businesses to move to cleaner energy?

Jurisdiction: New Zealand

The goal is to identify and segment all the building roofs of a city and identify which way they are facing (a roof could be segmented by identifying the ridgeline). This information could be later on stored into a GIS and be queried to filter all the roofs facing a certain orientation to the sun. It is important that we optimize sun exposure for north facing roofs. This information will help energy companies, councils and others to identify where solar panels could be installed and evaluate the potential in the city.

How can we encourage citizens and businesses to move to cleaner energy?

Councils want to work with residents and business owners to look at areas to activate where solar can be of benefit. It is especially important to cross-reference this against social-economic data, so the best roofs fit the people most in need, for subsidising, etc. The council wants the city to have energy independence and also consume clean energy, eg. help people move off wood burners.

Council and the public thus have this open data, together working to a solution.

How can we encourage citizens and businesses to move to solar?

Additional Information:

Below are examples of how you might wish to approach this challenge. Please use this as an example only and not as guidance.

For example:

ONE - Provide a city manager’s city-wide dashboard for the area potential of all north facing roofs:

• Cross-reference the dashboard with lower socioeconomic areas of the city
• Make the direction an adjustable parameter +/- North (the sun is not perfectly aligned with North)

TWO - Provide a citizen interface to take their home address and analyze their roof to indicate their north facing area and if they are available for subsidy (in lower socioeconomic area)

• Provide a mock contact to ECan submission form with their details to look at different options for a subsidised solar scheme

THREE - Exclude any roofs already that have solar installed

FOUR - Identify any roofs with a chimney installed


The technical stuff.

You have the choice of using the Trimble AI / Machine Learning tools to help solve this challenge! Everything you need to analyse a roof image, what pixel is what for the neutral network and training engine. Here is a link to the Trimble AI - Getting Started Guide. We have Trimble technical mentors online over the weekend to help answer your questions.

Eligibility: At least one team member must be able to attend and present your challenge at the Te Pae Technology Expo in Oct 2021 if selected.

Entry: Challenge entry is available to all teams in New Zealand.

Dataset Highlight

NZDep Indicator Data Sets - 2018 (Most recent) - New Zealand Indexes of Deprivation / Socio-economic factors

Go to Dataset

Canterbury - Christchurch and Ashley River (NZ) Lidar Index tiles

Go to Dataset

New Zealand Building Outlines

Go to Dataset

Christchurch City (NZ) Aerial Photos

Go to Dataset

Challenge Entries

Back to Challenges