Solar Planner

Inspiration:

Our huge global ecological footprint, which shows our resources getting exhausted almost twice as fast as they can be replenished. The importance of clean energy and sustainability towards our future is undeniable.

What it does:

Currently, our program is capable of analyzing an already processed image to determine usable surface area on buildings in a region. The intention is to provide an application that will model the possibilities of solar energy implementation for urban planners and other entities with similar roles (such as college campuses), and project the effects of this sustainable energy into the future.

How it should work:

With satellite imaging, rooftops can be identified through the use of computer vision/machine learning. These areas, once highlighted, can be analyzed for potential of solar panel installation, and economic/ecological benefits mathematically calculated to quickly and efficiently judge viability.

Challenges we ran into:

Unfortunately, we were not able to implement an object detection strategy for rooftops successfully within the time-frame, so we ended up having to skip over that portion of our project entirely.

Accomplishments that we're proud of:

We think we have a pretty good idea that has the potential to create positive change in favor of sustainability and clean energy, and we're proud of the hard work we did through the night to get where we are now.

What we learned:

We learned some aspects of computer vision that we were previously unfamiliar with. We learned to use Pillow, as well as some more details about solar panel design and operation.

What's next for Solar Planner:

Many improvements, hopefully. We want to implement machine learning and automate the process to create a truly convenient application that will be useful in the ways specified. We would also like to obtain more data for higher precision/accuracy; we assumed several things such as panel size and efficiency, we didn't take into account roof setback (solar panels in NYS need to be at least 3 feet away from roof edges, according to some sources), panel orientation, variable weather conditions, and many other factors. We also hope to add more features to improve usefulness, e.g. costs, return calculations

Built With

• pillow
• python