Oct 9th 2024 Current Affairs

Index:

1. ISRO's PSLV-C37 Mission Upper Stage Re-entry

CONTEXT:

Launch Date: February 15, 2017.
Launch Vehicle: Polar Satellite Launch Vehicle (PSLV-C37).
Payload:
- Main Satellite: Cartosat-2D.
- Co-passengers: 103 smaller satellites including INS-1A, INS-1B, Al-Farabi 1, BGU SAT, and more.

Mission Significance:

Historical Achievement: PSLV-C37 created a record by launching 104 satellites in a single mission.
Orbit Details: Post payload injection, the PS4 upper stage was placed in an orbit of 470×494 km.

Re-entry of PSLV-C37 Upper Stage:

Tracking:
- S. Space Command tracked the PSLV-C37’s upper stage (PS4) with NORAD ID 42052.
- The altitude of the stage began to decay due to atmospheric drag effects.
Prediction:
- ISRO’s Space Operations Management (IS4OM) monitored orbital decay.
- Predicted re-entry: First week of October 2024.
- Orbital decay: Stage decayed to a size of 134×148 km by October 6, 2024.
Re-entry Time:
- October 6, 15:49 UTC, with a margin of uncertainty of about 1 minute.

Compliance with Space Debris Guidelines:

The re-entry complied with international space debris management guidelines, particularly those laid out by the Inter-Agency Space Debris Coordination Committee (IADC).

Objective: Limiting the post-mission orbital life of defunct objects to below 25 years in low-Earth orbit.

2. Nobel Prize 2024 in Physics – Foundations of AI and Neural Networks

Awardees:

John Hopfield (91 years) – Princeton University, USA.
- Contributions: Laid the foundation for understanding how neural networks mimic the human brain’s functioning.
Geoffrey Hinton (76 years) – University of Toronto, Canada.
- Contributions: Backpropagation technique in artificial neural networks, enabling them to learn from mistakes and improve.

Field: Artificial Intelligence (AI) and Neural Networks.

Significance: These discoveries have enabled major breakthroughs in pattern recognition, such as face recognition, image enhancement, and other applications in AI.

How Artificial Neural Networks (ANNs) Work:

Inspired by the Human Brain:
- ANNs consist of layers mimicking the neuronal connections in the brain.
- Input layer: Receives data.
- Hidden layers: Process data.
- Output layer: Provides the final result.
- Learning Mechanism: ANNs learn and improve by adjusting their processing based on training data.

Key Innovations:

Backpropagation (Hinton): Allows ANNs to learn from mistakes and improve accuracy, particularly in tasks like image recognition and speech processing.
Deep Learning: Hinton’s work on deep networks, where multiple layers allow ANNs to learn complex patterns from large datasets.
- Spectacularly demonstrated in the 2012 ImageNet Visual Recognition Challenge, leading to breakthroughs in AI applications.

Applications of ANNs:

Face Recognition.
Speech Processing.
Voice Assistants.
Self-Driving Cars.
Medical Imaging and Diagnostics.

3. MeitY Relaxes AI Compute Procurement Norms to Accommodate Start-ups

CONTEXT: Ministry of Electronics and IT (MeitY) has relaxed norms to enhance participation from start-ups in AI computing capacity procurement. It’s part of India’s larger ₹10,370 crore IndiaAI Mission.

Key Objectives:

IndiaAI Mission aims to establish a computing capacity with more than 10,000 GPUs.
Development of Foundational Models: Models with 100 billion parameters, trained on vast datasets covering Indian languages and priority sectors like healthcare, agriculture, and governance.

Changes in Norms:

Turnover Requirements Reduced:
- For primary bidders: Lowered from ₹100 crore to ₹50 crore.
- For non-primary or consortium members: Lowered from ₹25 crore to ₹10 crore.
- Relaxation to ensure that more start-ups can bid for projects.

Compute Power Requirements:
- TFLOPS (Trillion Floating Point Operations per Second):
  - Reduced from 150 TFLOPS to 300 TFLOPS for the required systems.
  - AI Compute Memory: Reduced from 40GB to 24GB.

Minimum Turnover Requirement: Bidders must have ₹10 lakh minimum turnover over the last three financial years (2020-21, 2021-22, and 2022-23).

Experience Criteria: Experience of companies with local content between 20%-50%.

AI Compute Mission

Focuses on enabling cloud-based infrastructure for AI development.

Supports Indian firms in building large AI systems.

Ensures compliance with Make in India guidelines for components procurement.

Bidders’ Responsibility:

Successful bidders must ensure availability of AI compute capacity for consumption on demand.

A minimum of 10 AI compute houses will be required to meet demand, scaling up to 500 AI compute houses within a month and 1000 houses in a year.