THE ARK | A Flood of Ideas

A Flood of Ideas

Your challenge is to develop a new methodology or algorithm that leverages Earth observation and critical infrastructure datasets to estimate damages to infrastructure caused by flooding. Make a measurable impact on the resilience of nations by helping the Earth observations community contribute to the United Nations’ primary effort to reduce disaster risk!

MET URPS(METROLOGICAL REMOTELY PILOTED SYSTEM)

Summary

The two major problems estimated is:1.That the data collected by NASA can be only obtained from space.2.The data is not collected within a stipulated amount of time and no sufficient time to warn people in case of emergency .Hence we designed a module called as the MET URPS(METROLOGICAL REMOTELY PILOTED SYSTEM) which would actually float or fly to collect data such as the temperature , pressure ,accumulated seismic energy and many more factors responsible for a flood.This data collected is compatibly quicker and much more efficient cause we have implemented our project on raspberry pi system with all the necessary sensors .The data transmitted in fed into deep learning module quickly

How We Addressed This Challenge
In June through mid-July 2018, successive heavy downpour in southwestern JAPAN resulted in widespread, devastating floods. As of 20 July, 225 people were confirmed dead across 15 prefectures with a further 13 people reported missing. More than 8 million people were advised or urged to evacuate across 23 prefectures.  




  • This was a major outbreak in japan causing 100 of lives .The reason was though the data was transmitted before the floods ,the time was not sufficient for the people to evacuate .hence the data has to be transferred much more earlier so that sufficient time is given for the predictions and the evacuations in time of disasters .And hence we have proposed a system that would solve the above mentioned problems and give enough time for the people to evacuate .



What we have developed?



  • We have developed a system called as the MET URPS(METROLOGICAL REMOTELY PILOTED SYSTEM) .These are tiny floating objects (either floating in the water or in the air ) that mostly consist of sensors .
  • First in order to predict the distortion of natural weather or climate ,we need to know the data or the information about the temperature, pressure and much more .Hence the required sensors are fitted inside a resistive material and are let into the troposphere (where more often the weather takes place )or deep down the ocean .
  • These tiny floating objects called CLIME (climate sensors) are remotely controlled in a particular area so that the data's of pressure, temperature ,humidity at that particular area are recorder and stored.


How it works ?

 



  • 100s of CLIMES are let to float in the air (troposphere) and deep oceans and are remotely controlled .
  • The sensors inside it collects data and sends it to us in much faster rate and the predictions are done earlier and there is sufficient time to work on later .


Why is it important ?



  • As addressed before the information collected by NASA can be obtained only from space through images and are mostly 2d and no direct information or data's are obtained also those data has mirror timings
  • Hence sufficient time is not obtained .
  • But in our module, direct data or information has been obtained and accurate data has been obtained by taking the average of information's obtained on the climes (since it is direct ).
  • Data obtained is comparatively faster rate so that predictions and warning could be given before hand and save millions of lives.


What do we hope to achieve ?



  • A true life incident happened in jataka (capital of Indonesia ) on Jan 29 of 2007 ...it was quite a peaceful morning and nobody expected or was ready for the GREATEST FLOODS IN INDONESIA EVER RECORDER.
  • Warning was issued just before minutes the flood strict and people weren't even given time to evacuate .
  • This resulted in 80 deaths and over 40,000 people lost their property .
  • If the warning could have been a bit earlier and people would have evacuated and reduced the loss of lives
  • And that's all because that the data reached so late and this is just an incident ,many such incidents have happened in the history
  • To stop all this conditions and to save millions of lives ..this module has been created to be more accurate and give more time to evacuate.






How We Developed This Project

Why did i choose this challenge?



  • The negative effects of flooding is the damage caused to buildings or structures from the pressure created by the weight of water flowing, and can demolish bridges or buildings in its path. Even minor effects cause not only inconvenience but loss of lower level electrical goods, kitchens and furniture, soaks into dwelling walls that takes months to dry out from costing insurance companies huge amounts in payouts.
  • Also so many loss of lives and property which affected us and we build this up to reduce loss of lives during floods.

Tools and hardware used in this project :

What data do we actually need ?



  1. Average temperature ,
  2. Precipitation ,
  3.  Average wind speed 
  4. ,Average relative humidity, 
  5. Vapor pressure , 
  6. Dew point temperature ,
  7. Average local pressure , 
  8. Average sea-level pressure , 
  9. Duration of sunshine ,
  10. Visibility , 
  11. Ground-surface temperature.


In simpler words we need data (sensors ):



  1. Temperature 
  2. Pressure 
  3. Humidity 
  4. Composition of gases 
  5. Accumulated seismic energy (incase of ocean ).




The latest technology devices are the most appropriate and even preferred for their simplicity, ease of use and low cost. The raspberry pi module consisting of temperature ,pressure and precipitation sensors is fitted inside a buoyancy material so that it could float.

Temperature sensors :


These are four sensors we recommend using because they are inexpensive, easy to connect, and give accurate readings; DSB18B20, DHT22, BME280, and Raspberry Pi Sense HAT.




  1. DHT22— This temperature and humidity sensor has temperature accuracy of +/- 0.5 C and a humidity range from 0 to 100 percent. It is simple to wire up to the Raspberry Pi and doesn’t require any pull up resistors.
  2. DSB18T20 — This temperature sensor has a digital output, which works well with the Raspberry Pi. It has three wires and requires a breadboard and resistor for the connection.
  3. BME280 — This sensor measures temperature, humidity, and barometric pressure. It can be used in both SPI and I2C.
  4. SENSE HAT— This is an add on board for Raspberry Pi that has LEDs, sensors, and a tiny joystick. It connects directly on to the GPIO on the Raspberry Pi but using a ribbon cable gives you more accurate temperature readings.


Pressure sensors :




  • BMP180 is one of sensor of BMP XXX series. They are all designed to measure Barometric Pressure or Atmospheric pressure. BMP180 is a high precision sensor designed for consumer applications. Barometric Pressure is nothing but weight of air applied on everything. The air has weight and wherever there is air its pressure is felt. BMP180 sensor senses that pressure and provides that information in digital output. Also the temperature affects the pressure and so we need temperature compensated pressure reading. To compensate, the BM180 also has good temperature sensor.


ELECTRONIC DESIGN OF THIS PROTOTYPE:



  • In this section we elaborate the construction and working of our design . raspberry pi v2 is connected to the sensors with the help of connecting wires which has been coded with python this raspberry is connected to your module designed for to predict using deep learning …(we developed a deep learning model for heavy rain damage prediction using data collected in the week preceding heavy rain damage. The deep learning techniques used for the model development are a DNN, CNN, and RNN, and the ideal deep learning model for the heavy rain damage prediction was proposed by comparing the accuracy of each deep learning technique. To verify the deep learning prediction model proposed in this study, training and testing of the model were performed 30 times for each model. Through the process, the accuracy and the robustness of the deep learning model were evaluated. The results indicated that the mean accuracy was high in the order of the DNN-based model, CNN-based model, and RNN-based model, and the standard deviation was small in the same order. )



PROBLEMS FACED :



  • Data collected from all the climes is not that accurate .in order to be accurate we had a problem collecting data from the sensors that is when taking an average of all the data obtained .
  • Also the problem raised that how energy(that is when two tectonic plates collide under the ocean an accumulated seismic energy is created )which is so strong and how this clime could withstand its pressure
  • and also how the pressure underwater could be detected . (we have to work on these in real time projects on hand )



How We Used Space Agency Data in This Project

NASA greatly helped in understanding the problem :



https://disasters.nasa.gov/floods?page=4
  • The above mentioned page proved a lot of views and ideas that how data could be transferred at a much faster rate .
https://youtu.be/wqLghXCMxBI
  • This YouTube link provided a lot of information is implementing is our project
https://sharaku.eorc.jaxa.jp/GSMaP/
  • This jaxa helped us in finding out the composition of gases in the troposphere .so that we could proceed on the calculation without any sensors needed.
Project Demo

https://www.youtube.com/watch?v=pKg6fvJuFV0

Data & Resources
https://link.springer.com/chapter/10.1007/978-1-935704-09-6_1
Tags
#SAVELIFE #RASBEERYPI #SENSORS #URV
Judging
This project was submitted for consideration during the Space Apps Judging process.