We chose to work on the Confront: Better Together challenge to address human rights inequalities faced by migrants, refugees and human trafficking victims. We want to use satellite images to create situational awareness of high risk migration routes. Situational awareness is visualizing critical situations such as irregular border crossings or emergencies to address whether interventions are required.
Thousands of migrants escape dangerous situations in their home countries by taking long, risky journeys. Oftentimes these journeys bring them to situations even more dangerous than their home countries. When migrants arrive to their destinations they are either fined, arrested or turned away. Many of these migrants do not bring many resources with them, so they are unable to pay fines or to sustain nutrition for the journey back. Since 2014, 36,481 migrant deaths have been recorded and there are countless others that go unnoticed.
We want to make satellite data accessible to NGOs and humanitarian organizations working on relief efforts for migrants. Being able to identify migrant/ human trafficking boats in real time, organizations can go intervene with a mitigation plan before disaster strikes. With our app we aim to monitor weather patterns, dangerous water conditions, unidentified boats and coast patrol for migrant/human trafficking issues.
We believe that all individuals have an equal right to a life that is safe and healthy. Migrating in search of a better quality of life should not be a death sentence. We hope to use data from NASA, CSA and partner organizations to work with humanitarian/relief organizations and potentially save thousands of lives.
Satellite images are generally used for environmental and scientific purposes, however we believe that this data is beneficial for humanitarian causes as well. As a group of young diaspora, we empathize with the refugees and migrants around the world. With constant news reports of migrants facing disastrous situations during their journeys or being turned away from their destination we want to develop prevention and mitigation methods.
Our approach to develop this project has been heavily focused on how to make the data affordable and accessible to humanitarian relief organizations. We used Python Anaconda editor, Python SHA #256 Hash function and Jupyter notebook. We also used data from CSA, NASA, Sentinel HUB, Open Street Maps, AIS, Skywatch and Github.
One of the main challenges of our project is our concern about the privacy and safety of migrants. We want to ensure this data is not easily accessible for everyone and only neutral, non-governing agencies have access. We are using block-chain and are planning to add privacy engineering to our model.
Our aim is to utilize as many data sources as we can, particularly satellite imagery where boats and other vessels are seen from space. We want to process images where we can use machine learning and deep learning to predict and model accurate detection of objects. Ideal images are 10m/px, however, we acknowledge that acquiring high resolution data is expensive. After speaking with a CSA mentor, retrieving RCM data requires high level security clearances, thus, these are our limitations to acquire high quality imagery. Nonetheless, there is an abundance of open source data with mixed datasets with anywhere from low to high resolution data. We know it is possible to acquire images of similar areas of interests using different satellite sources. Along with satellite data, we will be leveraging AIS and weather data for real time monitoring as well as using historical records to predict future movements and weather patterns.

https://missingmigrants.iom.int/
https://www.marinetraffic.com/en/ais/home/centerx:-12.0/centery:24.9/zoom:4
https://www.eodms-sgdot.nrcan-rncan.gc.ca/index_en.jsp
https://landsat.gsfc.nasa.gov/
https://github.com/robmarkcole/satellite-image-deep-learning
https://towardsdatascience.com/object-detection-accuracy-map-cheat-sheet-8f710fd79011