Locusobrutamap

Aaruush, Chennai | Mapping Space Trash in Real Time

Locusobrutamap-Mapping Space Traces in Real Time

High-Level Project Summary

LocusObrutaMap is a combination of three latin word 'Locus' means space, 'Obruta' means debris and Map is map i.e. track. As the name suggest that mapping the space debris or space traces like satellite's wastes, ISS-wastes, rockets or space shuttle's wastes, asteroids etc. , This team's motto is 'Secure future space exploration, space missions and protect our planet from Space debris.' We have used concepts of Deep learning, in which we use Convolutional Neural Networks (CNN) using KERAS in Tensorflow, implemented object detection using YOLO-4 in tensorflow and make testing and Evaluation of results using Tensoflow models. This project is not complete properly and needs more work.

Link to Project "Demo"

https://docs.google.com/presentation/d/1LqwFijWiAh5Sz9Z8iwJMQqw_teOxUbrJlegeFzMSY_M/edit?usp=sharing

Link to Final Project

https://colab.research.google.com/drive/1uS5NTQ76T3jmMF7xP-ZBBFuM678p9dy_?usp=sharing

Detailed Project Description

Locusobrutamap-Mapping Space Traces in Real Time

We have developed a Deep Learning on Codelab named Space Debris Mapping using Tensorflow , in which we use YOLO-4 for object detection using CPUs and GPUs. This program can detect debris from given data and train the model and evaluate the results. This can map debris 90-97% of accuracy, locate each debris along with its % of accuracy of being the debris.

This model takes datasets and gives results as per the learning of the model presently, but later it can be made possible to achieve real time mapping of Space debris using Geospatial application. And further it can use datasets of Near-Earth Object Surveillance Satellite (NEOSSat). It means this project we have done requires lot of work and also we requires more knowledge about it. But lack of support and being a Bachler's student of Computer Science & Engineering, as per our strength to do in this project , we have done and learn more. I have chosen this challenge for Learning and also interest in my space exploration .

GitHub link of the project is https://github.com/devanand369/spaceDebrisMapping.git .

Systems and Software we used

Operating System - Windows 11 pro Insider Preview Build 22468
Hardware - Dell Vostro 3490 , Intel i7 10th gen. processors , AMD Radeon 610 graphic card
Codelab (colab), Jupyter Lab, Anaconda
Python3
Tensorflow for deep learning model
Mapping/virtual globe libraries: openlayers examples, leaflet tutorials.
General orbital parameter information to map space debris: two line element sets, AIAA 2006-6753, sgp4 propagator, celestrak space debris tle, space-track documentation.
Software libraries to read orbital parameters in two-line element set (TLE) format: satellitejs library, nsat/jspredict library, pypi sgp4.
Examples of open-source satellite trackers: stuffinspace, jsattrak, worldwind spacebirds

Space Agency Data

We are using NASA WorldWind and forked it on Github for Geospatial 2D and 3D globe, Learn from NASA's Space Debris Report and Datasets from Canadian Space Agency of datasets of NEOSSAT (Near-Earth Object Surveillance Satellite), Most of the learning materials from NASA websites and NASA videos and take ideas from ISRO mission NETRA to remove Space Debris from space .

Hackathon Journey

NASA Space Apps Challenge is most inspiring and developing real challenges solutions. This is my first global challenge, help to enhancing my thinking ability and helps in grow knowledge in Artificial Intelligence and expertise in programming skills. This challenge helps us to confidence building, team work ability.

Our Team's approach to Mapping Space Debris in Real time is Deep Learning in Artificial Intelligence, using Convolutional Neural Networks (CNN).

Sorry this project not complete presently as per the requirements of the Challenge and does not meet the challenge as Real time debris tracking from 3D or 2D Geospatial, but further it can be improved and meet the requirements of the challenge. It is happening because My team has only two members but other member of my team is not actively participating in the challenge, so I remain only one who lead the project and due to lack of support from team member and lack of knowledge of me and lack of time due to my college work.

Overall this Space Challenge is full of exciting, joyful and inspiring.