Wildfires - BlazeNet

• 

  Network Architecture

• 

  MESH Network of Resources

• 

  Synergetic Coordination of Firefighting Resources

• 

  Decision-making Agent

Inspiration

The reconstruction period after a natural disaster can either equitably redistribute resources and reframe society by uplifting our most vulnerable populations, or further exacerbate existing inequalities. Current resource allocation strategy and alert systems employed by most firefighting departments are insufficient. Manned firefighting aircraft have cost and risk issues

What it does

It optimizes resource allocation by utilizing a deep Actor-critic network with human-in-loop for feedback control. It also incorporates firefighting UAVs in a MESH network as well as accepts information from multiple sources.

How we built it

We started by doing some research on the public policy side and finding out the imperfections of these policy

Challenges we ran into

No first-hand experience in firefighting

Accomplishments that we're proud of

We are able to propose a solution within the required time frame and received some valuable feedback from domain experts

What we learned

Real-life resource allocation is extremely complicated and hard to model

What's next for BlazeNet

We plan to develop a cross-platform frontend application and collaborate with local governors and firefighting departments, as well as getting a preliminary testing on our experimental decision-making network.

Built With

• actor-critic
• http://caltopo.com/
• nation-fire-danger-rating-system
• reinforcement-learning