Climate change disaster aid map
With the increase of natural disasters due to climate change, this map aims to predict what aid will be required before the disaster occurs so that the aid is already in place.
Help communities be resilient to natural disasters/ reduce subsequent suffering
How it works:
Using AI, Machine Learning and satellite forecasts
Need to know type of infrastructure and population information
Working with the UN and aid charities like red cross
Long-term predictions - Educate at risk communities on survival techniques
Short-term - placing aid near to where the disaster is going to happen to shorten the time taken to deliver aid
How it addresses the UN sustainability goals:
3) Good health and wellbeing → medical resources, first aid
9) Industry, Innovation and Infrastructure → learning and protecting
10) Reduced inequalities → aid can be close and accessible to all
11) Sustainable cities and communities → surviving climate change
13) Climate Action
17) Partnerships between UN and other natural disaster charities → coordinate aid so not all in 1 place
Interactive map - alerts will appear to users when a prediction of a natural disaster in their region happens. Another function is for the UN/Charities to pre-plan where aid is coming from and for that decision to feed into a machine learning model. The machine learning model will learn overtime when and where to send aid to and from.
Platform to identify closest and most suitable organisations allowing them to collaborate and coordinate response.
Categories of aid required: medical, sanitation, infrastructure, repair, fire
Climate change is increasing up to 10 times the current scale according to the Institute of Economics and Peace.
According to the charity World Vision it can take up to 72 hours to deliver basic aid to affected areas.
Our team is a multinational group and so have been sharing personal experiences of how different countries are set up to deal with various natural disasters they are prone to. This gave us the idea to use local knowledge and geography of the area to understand the best safe spaces for different disasters.
It is important to get a mix between human and machine input because human error can happen in stressful situations but machine learning only works off data it already has. Feedback from users will allow it to evolve and be optimised. An example of a similar machine learning model (logistics based) is Amazon, they predict what people will order in different locations using machine learning but this doesn’t always take into account external factors like storms etc, overtime the algorithm will account for this but until then some human input is involved.
We have seeked feedback from data scientists and academics on our idea to help develop our concept.
Considerations must be made for the initial set up and ongoing operational costs including costs for app design, machine learning as well as UN and charity contributions.
Hiring talent from the local communities using the app for future operational costs and app development.
The app will continually evolve through expert input and both community and feedback to be shaped to the needs of the end user.
#Sustainability #Infrastructure #Coordination #NaturalDisaster #Aid #HumanitarianResponse #Resilience #ClimateChange #ClimateAction #UNSDGs #PlanningAhead #AI #SDG3 #SDG9 #SDG10 #SDG11 #SDG13 #SDG17 #InternationalCollaboration #Designathon2022 #DisasterPreparedness #D22025
@Helen Aries @Isabel Dodd @Jahanara Hasan Sohana @Boaz Thembo
We would also like to thank our mentors for all their valuable feedback and suggestions.
@Tavish Kotian @Joe Haniff