How AI influenced ISRO
What is Artificial intelligence ?
Artificial intelligence (AI) refers to intelligence demonstrated by machines rather than human intellect. Thinking, understanding, reasoning, problem-solving, logical reasoning, vision, motion, and handling, as well as, to a lesser extent, social intelligence and creativity, are all behaviours associated with human intelligence that AI systems will typically display.
Also Read https://www.mindstick.com/news/957/how-nasa-bring-technical-advancements-with-the-aid-of-ai-and-ml
Artificial intelligence (AI) is the mimic of human intelligence in robots that have been trained to think and act like humans.
What is Machine Learning ?
Machine learning is a branch of artificial intelligence (AI) that allows computers to learn and improve on their own without having to be explicitly programmed. Machine learning is concerned with the creation of computer programmes that can access data and learn on their own.Huge amounts of previous data are fed into the machine learning model, which then makes predictions.
How AI and ML used in ISRO ?
> Chandrayaan 2 : AI powered ' pragyan ' rover .
• ISRO launched the Chandrayaan 2 spacecraft into Earth orbit on July 22, 2019, as part of the second lunar mission.
• The AI-powered Pragyan rover aboard Chandrayaan 2 can only connect with the Lander and incorporates a piece of motion technology developed by IIT-Kanpur researchers to let the rover manoeuvre on the moon's surface and aid in landing. The algorithm will assist the rover in tracing water and other minerals on the lunar surface, as well as sending images for research and analysis.
• The rover is a six-wheeled robotic vehicle that can undertake payload tests in real time. It is driven by AI tools and frameworks, runs on solar power, and can only connect with the Lander. The Alpha Particle X-ray Spectrometer (APXS) and Laser Induced Breakdown Spectroscope (LIBS) are two of the Pragyan Rover's payloads (LIBS).
Chandrayaan-2 Pragyan shows how AI is helping space exploration.
> Multi Object Tracking Radar (SDSC-SHAR) :
Read more about : how radar detector aided to resist collision of Nasa's LRO & ISRO chandrayaan-2 satellite
• The task of developing a space object tracking solution to ensure satellites' safe survival in the challenging environment of outer space, which contains millions of unknown objects that could interfere with any ISRO-sponsored mission.
• The goal is to create a Multi Object Tracking Radar. Target identification was first created by ISRO utilising machine learning techniques and MOTR radar data.
• Range, Azimuth, Elevation, and Signal to Noise Ratio are all components of radar data (SNR). Target size can be characterised based on range and SNR correlation. Target nature can be determined just by SNR variation over a single track duration. A model should be trained on radar monitored data using Machine Learning methods (Range, Azimuth, Elevation and SNR). A goal nature (controlled or uncontrolled) and size should be identified by the trained model. Models that use common libraries in Python Machine Learning Algorithms are now feasible.
> Image Processing and Pattern Recognition (IIRS)
Using IIRS NASA gives an overview of delhi smoke and pollution .
Read more : https://www.mindstick.com/news/918/smoke-captures-the-delhi-nasa-analysis-on-delhi-s-pollution
The objective for ISRO in the 1980s and 1990s was to develop an effective and cost-neutral image processing and pattern recognition solution for planned missions in the following decade. As a result, unmanned image processing and pattern recognition are becoming increasingly popular (IIRS).
• ISRO used Artificial Neural Networks (ANN), a generic term for a vast class of machine learning algorithms, the majority of which are taught using the back propagation process. ISRO's team explored a variety of deep learning algorithms in diverse earth observation data applications, including self-learning based categorization, prediction, multi-sensor spatial analysis in crop/forest species identification, and remote sensing time series data processing.
> AI-enabled monitoring system for forest conservation :
• The National Remote Sensing Centre (NRSC), which ISRO conceived and constructed, is a monitoring system that uses optical remote sensing, geographic information systems, artificial intelligence, and automation technologies to watch forest cover change and battle deforestation.
• ISRO develops a machine learning technique that examines imagery to detect small-scale deforestation and increase reporting frequency.
• It also allows scientists to interpret satellite information more quickly, cutting the time it takes to produce new reports from a year to a month. The goal of the NSRC is to prevent detrimental changes in the green cover and to safeguard animals.
> Autonomously Navigating Robot for Space Mission (IISU) :
ISRO's task was to design and launch unmanned robots to assist in the retrieval of important space data on many trips throughout the year.
A half Vyomnoid with 3D vision and dexterous manipulative ability to carry out prescribed crew responsibilities in an unmanned mission or aid crew in manned missions.
Full autonomy with 3D vision, dynamically controlled mobility in zero , Artificial Intelligence / Machine Learning allowed real-time decision making with vision optimization and path planning algorithms were among the capabilities of the FULL Vyomnoid.
To tackle this, ISRO used Artificial Intelligence-enabled Path Navigation algorithms.
SRO has recommended scientists and researchers to concentrate on developing generalised parameter extraction software based on artificial neural network (ANN) learning methods that use multidimensional ANN approximation to map microwave filter features. A communication satellite has a huge number of microwave filters, each of which must be fine-tuned after manufacture.
With AI-based sensors embedded in propellants, ISRO plans to develop high-end propulsion technology to assure cost-effective, re-usable, recoverable, restartable, and dependable space missions.
Read more at :
1) https://www.mindstick.com/news/955/how-does-a-satellite-works
