fire team| Spot That Fire V3.0

Team Updates

Inspiration


We have been able to review in the different challenges and papers that there are many incidents of great magnitude and many people lose material and / or personal property.

The network technology to interconnect the sensors that has been chosen is the following:

LoRaWAN ™ is a specification for Low Power Wide Area Network (LPWAN), specifically designed for low power consumption devices, operating in local, regional, national or global networks.


Among the main advantages of LoRa are the following:


  • High tolerance to interference.
  • High sensitivity to receive data (-168dB)
  • Based on “chirp” modulation
  • Low Consumption (up to 10 years with one battery *)
  • Long range 10 to 20km.
  • Low data transfer (up to 255 bytes)
  • Point-to-point connection.


We have put together a set of digital tools which allow us to carry out an awareness campaign in the shortest possible time with the help of social networks and with information from NASA's FIRMS to be able to analyze and sketch analyzes of machine learning models.


Presentations:

Video: https://drive.google.com/file/d/12s_Vo0aNKQWmBov_KudJIKU6Dv-eNovB/view

PPT: https://drive.google.com/file/d/1OgtFiTnlWPSwUwK2XI2VJEJWTU_374BR/view?usp=sharing


Action plan


The project is divided into 2 major implementation milestones: the digital part and the electronic part. Which can be achieved thanks to the use of Machine learning techniques with a KNN (k-Nearest Neighbor) model that allows us to predict possible safe areas for the part of fires in progress and the possibility of reports with images or videos of people surrounding the area. Additionally, for areas with higher risk, being able to deploy fixed devices that allow us to better monitor the areas of higher risk.


Electronic Part:

This analysis which will feed an app developed in Django which can be consulted through the FIRETEAM APP. In which you can see the areas with active fires and safe areas; using social networks to generate aid in the short and medium term, and thus better reduce the economic and social impact.


Goals

  • Being able to generate social and community awareness to fight possible fires.
  • Make known Tips and recommendations for possible high risk events.
  • To be able to contribute to responsible tourism.
  • Know how to react better against fires.


Results

  • We have carried out a small campaign in anticipation of the launch of the APP and an interest in the project could be found.
  • The analysis of the data gave us a vision towards better optimizations to the model.
  • Verify the interest and possible generation of communities that allow facing such incidents.
  • Rescue the option of implementing this same solution for different natural or social incidents by integrating many other means apart from those in charge of fire control. 



Technologies used:

  • Google Colab, 
  • Machine learning
  • Python

Server:

  • Django,
  • Postgresql, 
  • Api,

APP

  • Flutter

Graphic design

  • Illustrator
  • Premier

Social networks

Repository 


E
Enrique Emilio Bonifaz Gutierrez