We were inspired from the next Generation of putting out fires with using military planes.

What it does

The pilot won't know which is the best area to put off first so, we made an AI which grades all areas, assuming how much will be burnt

How we built it

we created labeled data and augmented it a bit or so called added noise. We wanted a sage maker ensembled by linear regression and CNN, but we accomplished only a linear regression.

Challenges we ran into

Learning SageMaker is a curve of studying, not everything is build in, documentation isn't correct sometimes. It's not as simple tf+keras. We were billed too, so adding the CNN would have cost us much more.

Accomplishments that we're proud of

Surviving SageMaker with zero experince with it. It has nice debugging.

What we learned

fun working in a new team, and machine learning experience

What's next for put out wildfire

to build the prototype hardware which puts out the fire

Share this project: