Space Apps Challenge 2017

Satellite data for rain prediction is very useful when integrated with ground data we get interesting results. It's possible to get a risk of land slide when processing sensor data with NASA's satellite (pre-processed by us to reduce execution time and increase response). All devices uses a Photon board which is a production ready device with IoT integration and a very important feature which is Smart Config which makes it easy to have a device ready for production and usage wich integrated with the Phonton smart-phone application we can configure the device internet access without the need of re-programming the device. These devices are located in different zones of La Paz city which are always in danger of land slides.

With the usage of MQTT protocol -a light protocol for low internet connections, which is the case of La Paz city-, devices sends data to our AWS-based backend.

On the other hand we have a pre-processing application which processes NASA satellite data which is in HDF5 format. We are using the level 3 algorithm which the most useful for our case.

Once device data gets to the backend server, it processes risk of land slides by integrating and interpreting pre-processed NASA data. Once processed it returns via a web-socket to the browser in order to give a real time look-and-feel.

The technologies used for this project are:

• Backend Application using NodeJS express framework
• Pre-Processing for NASA's satellite data in MatLab
• Sensor-backend connectivity layer using Mosquitto Message broker
• Frontend User Interface using AngularJS framework
• Infrastructure running on AWS
• Continuous Integration and Continuous Delivery using Jenkins and Red Hat Ansible