NEW DELHI: Union Minister of State (Independent Charge) Science & Technology; MoS PMO, Personnel, Public Grievances, Pensions, Atomic Energy and Space, Dr Jitendra Singh in a reply in the Rajya Sabha today, updated the House about a series of “Artificial Intelligence” (AI) related initiatives undertaken in the recent years.
Since the member asking the question wanted to know specifically about the “Artificial Intelligence” related projects taken up by the Department of Atomic Energy in the last five years, the Minister stated that the The Department of Atomic Energy (DAE) has undertaken a number of “Artificial Intelligence” (AI) and Machine Learning (ML) based research in last five years, some of which are listed herein below:- AI based Content verification and anomaly detection,AI based Physical Intrusion Detection, Biometric & Face Recognition, Vehicle Identification & Management and Application Behavior & Anomaly Detection.
In addition, Dr Jitendra Singh stated, there are other AI related research areas such as Human Interface Development, Robotics,Image Processing,Medical and Biomedical, AI/ ML development platform and hardware,High-Performance Computing – job scheduling, resource utilization prediction, Nuclear Knowledge Management and optimization problems in Nuclear environments, Medical thermal images for breast cancer, eye diseases and diabetes mellitus, AI-algorithm based application to identify the near optimal value of the operating parameters of the electron synchrotrons for control system of Accelerators, An operating support system developed using Swarm intelligence and AI-based expert system to maximize the beam injection efficiency without operator intervention for control system of Accelerators, AI and ML algorithms/methodologies are used in Machine Vision based Inspection systems for inspection & quality control of Nuclear Fuel & Nuclear Fuel Assembly Components.
Giving more details , the reply stated , there are diverse AI based projects like Qualitative and quantitative analysis of different grades of steel by applying ML-based algorithms trained over spectral data from Laser Induced Breakdown Spectroscopy,Development of ML-based systems/techniques for Raman Spectroscopy,A multi-staged ML model is developed and evaluated to detect fraudulent connections,,,udy on ML models for IPv6 address lookup in large block lists,Study, Design and Development of ML-based technique for SPAM detection and Statistical modelling and ML were used to extract meaningful insights from CHSS Medical Data at RRCAT which includes prevalence and patterns of diseases, correlation among ailments/ symptoms, seasonal patterns, CHSS beneficiary profiling etc.
In the past 5 years, the Department of Atomic Energy collaborated with academic institutions for various AI/ML related research projects and programmes which are Robust Shape based Face Recognition System & Robust and Scalable Computer Vision Systems for Smart Multi Camera Video Surveillance.
The details of the projects and programmes undertaken in collaboration with academic institutions in these domains are: Early detection of breast cancer using thermal/infrared imaging, AI application for classification of the lesions mapped by infrared images, Diabetic eye diseases detection and classification using Convolutional Neural Network (CNN),Unsupervised radiation field mapping inside cyclotron vault,Unsupervised area surveillance by drone,Indian sign language coder and decoder, Video compression for low bit rate video conferencing and Neutron Gamma separation for DAQ system.
Artificial Intelligence and Machine Learning related research is carried out through projects in the last five years. 3 Projects costing Rs. 180 crores were under execution and around Rs. 53 crores have been utilized.
AI and ML research along with high performance computing has enabled processing of large volumes of data. This has resulted in development of applications in all the domains. It has also led to several developments in Security and Cyber Security. Development of AI based video analytics and Cyber Security tools were initiated and has resulted in mature technologies. Indigenous products such as Secured Network Access System (SNAS) have become a backbone of Cyber security.
DAE network is a very large network and millions of events are generated every day. Event Monitoring Systems, designed to analyze massive amounts of data and provide summarized results for display, have been developed and integrated with SNAS. These systems operate seamlessly round the clock (24/7), ensuring continuous vigilance and timely insights.