Anjali Goswami
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Anjali Goswami has a total of 7 co-authors. Here are the top co-authors by number of shared publications:
- Anjali Krushna Kadao
Kalinga University
3 shared publications - Roohee Khan
Kalinga University
3 shared publications - A Haja Alaudeen
Department of Computer Applications, B S Abdur Rahman Crescent Institute of Science and Technology, Tamil Nadu, India
1 shared publication - Abdumalikova Guli-Ra’NO Farruh Qizi
Faculty of Economics, Tashkent State University of Economics, Tashkent, Uzbekistan
1 shared publication - F. Rahman
Kalinga University, Raipur, India
1 shared publication - N. Arvinth
Research and Innovation, National Institute of STEM Research, Tamil Nadu, India
1 shared publication
Latest Publications
- 1 August 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1617-1622
Advanced Regularization Techniques for High-Dimensional Data in Robust Statistical Learning
- 1 August 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1617-1622
Advanced Regularization Techniques for High-Dimensional Data in Robust Statistical Learning
- 30 June 2026
- Published by IGI Global Scientific Publishing
- p. 355-374
Intelligent Transport Systems on Employing Wireless Sensor Networks for Smart Vehicle Communication
The urbanization rate and increasing traffic volume have placed enormous strain on available transport infrastructure; therefore, solutions are needed to optimize safety, efficiency, and sustainability. Intelligent Transport Systems (ITS) are one such promising technology being emphasized, with particular emphasis on the potential of Wireless Sensor Networks (WSNs) to facilitate intelligent vehicle communication. WSNs, a collection of geographically distributed sensors, hold the promise for real-time monitoring and collection of vital information with the potential for real-time vehicle-infrastructure interaction. They all contribute fundamental functions in preventing accidents, managing traffic, and improving the management of the road network. This work concerns integrating WSN and ITS for vehicular communication, focusing on developing a model that minimizes latency, enhances data quality, and ensures safe transmission.
- 30 June 2026
- Published by IGI Global Scientific Publishing
- p. 355-374
Intelligent Transport Systems on Employing Wireless Sensor Networks for Smart Vehicle Communication
The urbanization rate and increasing traffic volume have placed enormous strain on available transport infrastructure; therefore, solutions are needed to optimize safety, efficiency, and sustainability. Intelligent Transport Systems (ITS) are one such promising technology being emphasized, with particular emphasis on the potential of Wireless Sensor Networks (WSNs) to facilitate intelligent vehicle communication. WSNs, a collection of geographically distributed sensors, hold the promise for real-time monitoring and collection of vital information with the potential for real-time vehicle-infrastructure interaction. They all contribute fundamental functions in preventing accidents, managing traffic, and improving the management of the road network. This work concerns integrating WSN and ITS for vehicular communication, focusing on developing a model that minimizes latency, enhances data quality, and ensures safe transmission.
- Open Access
- 29 June 2026
- Published by International Academic Institute for Science and Technology in International Academic Journal of Innovative Research
- Vol. 13 (2) , 82-89
Data Driven Approaches for Industrial Systems Optimization and Process Efficiency Improvement
In highly competitive manufacturing industries, intelligent optimization techniques to improve operational efficiency, reduce production costs, and enhance system reliability are urgently needed. Large-scale industrial data, real-time monitoring, and unpredictable process variations are difficult for conventional optimization techniques to handle efficiently. This research proposes a data-driven methodology for industrial systems optimization and process efficiency improvement, leveraging predictive and analytical techniques. By combining industrial sensors, statistical data, machine-learning-based predictive models, and process optimization algorithms, it is possible to evaluate and monitor industrial processes. Data pre-processing, anomaly detection, predictive maintenance, and efficiency-based optimization models are incorporated to identify inefficient processes and improve overall productivity effectively. Using the proposed methods, the experimental analysis showed that machine downtime was reduced by up to 28.4%, production efficiency increased by up to 21.7%, and operational energy consumption decreased by up to 16.3% compared to the conventional industrial management strategy. The predictive maintenance models achieve an operational prediction accuracy of 93.5%, thereby significantly enhancing system reliability and avoiding unexpected equipment failures. Also, according to statistical analysis, the reduction in process variability and improvement in resource utilization efficiency were up to 19.8% and 24.1%, respectively, after implementing the proposed methods. This study demonstrates that a data-driven optimization approach is a promising and effective way to manage industrial processes by enabling real-time decision-making, prediction, and adaptation, which supports, to some extent, initiatives in sustainable manufacturing and Industry 4.0. The proposed model provides an Engineering and Applied Sciences contribution by delivering a scalable and efficient solution to improve industrial productivity, operational stability, and long-term efficiency.
- 29 June 2026
- Published by International Academic Institute for Science and Technology in International Academic Journal of Innovative Research
- Vol. 13 (2) , 82-89
Open AccessData Driven Approaches for Industrial Systems Optimization and Process Efficiency Improvement
In highly competitive manufacturing industries, intelligent optimization techniques to improve operational efficiency, reduce production costs, and enhance system reliability are urgently needed. Large-scale industrial data, real-time monitoring, and unpredictable process variations are difficult for conventional optimization techniques to handle efficiently. This research proposes a data-driven methodology for industrial systems optimization and process efficiency improvement, leveraging predictive and analytical techniques. By combining industrial sensors, statistical data, machine-learning-based predictive models, and process optimization algorithms, it is possible to evaluate and monitor industrial processes. Data pre-processing, anomaly detection, predictive maintenance, and efficiency-based optimization models are incorporated to identify inefficient processes and improve overall productivity effectively. Using the proposed methods, the experimental analysis showed that machine downtime was reduced by up to 28.4%, production efficiency increased by up to 21.7%, and operational energy consumption decreased by up to 16.3% compared to the conventional industrial management strategy. The predictive maintenance models achieve an operational prediction accuracy of 93.5%, thereby significantly enhancing system reliability and avoiding unexpected equipment failures. Also, according to statistical analysis, the reduction in process variability and improvement in resource utilization efficiency were up to 19.8% and 24.1%, respectively, after implementing the proposed methods. This study demonstrates that a data-driven optimization approach is a promising and effective way to manage industrial processes by enabling real-time decision-making, prediction, and adaptation, which supports, to some extent, initiatives in sustainable manufacturing and Industry 4.0. The proposed model provides an Engineering and Applied Sciences contribution by delivering a scalable and efficient solution to improve industrial productivity, operational stability, and long-term efficiency.
- Open Access
- 16 June 2026
- Published by Atlantis Press SARL
- p. 793-802
Explainable Federated Reinforcement Learning Framework for Intelligent and Ethical Smart City Systems
- 16 June 2026
- Published by Atlantis Press SARL
- p. 793-802
Open AccessExplainable Federated Reinforcement Learning Framework for Intelligent and Ethical Smart City Systems
- Open Access
- 5 June 2026
- Published by EDP Sciences in ITM Web of Conferences
- Vol. 86
Advancing AI for NLP to Improve Contextual Understanding and Human-Computer Interaction
The rapid evolution of Natural Language Processing (NLP) has provided a symptomatic insight into the gaps in contextual knowledge, namely, long-range dependencies, continuity in multi-turn language exchange, and non-textual signal fusion. The current study suggests a new synthesis of architecture aimed at improving the semantic modeling and adaptive learning of human-computer interaction. It is described by a four-stage pipeline, including a Contextual Encoder Layer to process linguistic information, a Knowledge Retrieval Layer that grounds the information in facts, a User-Adaptive Reinforcement Layer to make the interaction more personal, and a Multimodal Understanding Layer to integrate various signals of input information. In order to confirm this framework, an intensive experimental design was used with a data split of 70-15-15% and k=5-fold cross-validation. There was a vast amount of preprocessing of data, such as min-max scaling and dimensional reduction, to guarantee that features are significant. Findings show that there is a significant improvement in performance compared to the standard transformer, as the absolute error percentage improvement is 9% with an accuracy of 0.82 to 0.91. Additionally, the model obtained high-performance standards whose Precision and Recall were 0.89 and 0.88, respectively, and a consolidated F1-Score was 0.88. Such statistical observations affirm that a systematic combination of retrieval and reinforcement processes is very effective in minimizing ambiguity and enhancing the surface of response coherence. Although the research study under consideration was carried out in a controlled environment, the results offer a solid ground to be applied to practice in healthcare and education. The study concludes that the change in extract algorithms enhancement to closed architecture synthesis is central to the subsequent generation of persuasive AI systems.
- 5 June 2026
- Published by EDP Sciences in ITM Web of Conferences
- Vol. 86
Open AccessAdvancing AI for NLP to Improve Contextual Understanding and Human-Computer Interaction
The rapid evolution of Natural Language Processing (NLP) has provided a symptomatic insight into the gaps in contextual knowledge, namely, long-range dependencies, continuity in multi-turn language exchange, and non-textual signal fusion. The current study suggests a new synthesis of architecture aimed at improving the semantic modeling and adaptive learning of human-computer interaction. It is described by a four-stage pipeline, including a Contextual Encoder Layer to process linguistic information, a Knowledge Retrieval Layer that grounds the information in facts, a User-Adaptive Reinforcement Layer to make the interaction more personal, and a Multimodal Understanding Layer to integrate various signals of input information. In order to confirm this framework, an intensive experimental design was used with a data split of 70-15-15% and k=5-fold cross-validation. There was a vast amount of preprocessing of data, such as min-max scaling and dimensional reduction, to guarantee that features are significant. Findings show that there is a significant improvement in performance compared to the standard transformer, as the absolute error percentage improvement is 9% with an accuracy of 0.82 to 0.91. Additionally, the model obtained high-performance standards whose Precision and Recall were 0.89 and 0.88, respectively, and a consolidated F1-Score was 0.88. Such statistical observations affirm that a systematic combination of retrieval and reinforcement processes is very effective in minimizing ambiguity and enhancing the surface of response coherence. Although the research study under consideration was carried out in a controlled environment, the results offer a solid ground to be applied to practice in healthcare and education. The study concludes that the change in extract algorithms enhancement to closed architecture synthesis is central to the subsequent generation of persuasive AI systems.
- 17 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1074-1082
Edge AI Integration for Real-Time Control in Embedded Computing Systems
Embedded systems that integrate Artificial Intelligence (AI) have dramatically improved the efficiency of a wide range of applications with real-time control requirements, enabling fine-grained, low-latency decisions. This document reviews the Design, Implementation, and Performance of Edge AI models in embedded computational systems, specifically for real-time control tasks in resource-constrained scenarios. Edge AI differs from traditional Cloud AI because it is executed directly on edge devices (microcontrollers, FPGAs, or SoC platforms), which offer the advantages of reducing communication overhead, minimizing latency, and lowering power consumption. An overview of the primary architectural considerations for integrating edge AI, including hardware accelerators, model compression techniques (pruning, quantization), and power-efficient neural networks, is presented first. The significance of edge inference in uncertain and dynamic systems is illustrated with examples of real-time Control applications: autonomous robots, smart grids, and factory automation. The work also discusses the integration of AI techniques into traditional control systems and the improvement of system flexibility and robustness offered by such hybrid approaches. The system, a lightweight convolutional neural network running on an ARM Cortex-M processor to control a motor, is presented as an example of an autonomous drive system. The system provides inference latency of under 20 milliseconds and maintains reliability across different ambient disruptions. Experimental results suggest that edge AI controller systems outperform traditional PID controllers in terms of Response and adaptability. Also covers future challenges and opportunities for scaling Edge AI for advanced control tasks, including real-time scheduling, model generalization, security, and interfacing with diverse computing architectures. The results indicate that Edge AI must indeed be, and increasingly will be, embedded in control systems that need to respond immediately and are resource-constrained. This paper summarizes the current developments, accomplishments, and prospects of Edge AI for embedded Real-time Control and provides the basis for researchers and practitioners to advance intelligent embedded systems.
- 17 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1074-1082
Edge AI Integration for Real-Time Control in Embedded Computing Systems
Embedded systems that integrate Artificial Intelligence (AI) have dramatically improved the efficiency of a wide range of applications with real-time control requirements, enabling fine-grained, low-latency decisions. This document reviews the Design, Implementation, and Performance of Edge AI models in embedded computational systems, specifically for real-time control tasks in resource-constrained scenarios. Edge AI differs from traditional Cloud AI because it is executed directly on edge devices (microcontrollers, FPGAs, or SoC platforms), which offer the advantages of reducing communication overhead, minimizing latency, and lowering power consumption. An overview of the primary architectural considerations for integrating edge AI, including hardware accelerators, model compression techniques (pruning, quantization), and power-efficient neural networks, is presented first. The significance of edge inference in uncertain and dynamic systems is illustrated with examples of real-time Control applications: autonomous robots, smart grids, and factory automation. The work also discusses the integration of AI techniques into traditional control systems and the improvement of system flexibility and robustness offered by such hybrid approaches. The system, a lightweight convolutional neural network running on an ARM Cortex-M processor to control a motor, is presented as an example of an autonomous drive system. The system provides inference latency of under 20 milliseconds and maintains reliability across different ambient disruptions. Experimental results suggest that edge AI controller systems outperform traditional PID controllers in terms of Response and adaptability. Also covers future challenges and opportunities for scaling Edge AI for advanced control tasks, including real-time scheduling, model generalization, security, and interfacing with diverse computing architectures. The results indicate that Edge AI must indeed be, and increasingly will be, embedded in control systems that need to respond immediately and are resource-constrained. This paper summarizes the current developments, accomplishments, and prospects of Edge AI for embedded Real-time Control and provides the basis for researchers and practitioners to advance intelligent embedded systems.
- Open Access
- 8 December 2025
- Published by Association for Computing Machinery (ACM)
- p. 862-872
Autonomous Blockchain-Powered Smart Contracts for Trust Management in Decentralized Energy Grids
Blockchain technologies incorporated with decentralized energy grids can offer new paradigms in energy management, distribution, and consumption. Trust building among participants in decentralized energy systems is a critical challenge in the absence of a central authority. Foremost, the self-executing, fully autonomous blockchain-based smart contracts and the decentralized automated systems enable real-time, secure, and transparent business transactions. This paper examines the domain of smart blockchain contracts in decentralized energy grids. Smart contracts facilitate self-enforcement, aiding trust-building. Invoice blockchain contracts enable automated energy exchange, thereby facilitating contract execution without third-party involvement. Trust-building, system decentralization, energy producers and consumers bypassing operators, and operational cost reduction for self-sustainability provide system autonomy. The paper also discusses self-governing blockchain contract systems engineering. In smart energy contracts, the paper outlines system challenges, which include load, interoperable infrastructure, and energy consumption. Empirical and case study research, which are the primary focus of the paper, present blockchain contracts for decentralized energy systems as a trust-management and transaction-integration technology.
- 8 December 2025
- Published by Association for Computing Machinery (ACM)
- p. 862-872
Open AccessAutonomous Blockchain-Powered Smart Contracts for Trust Management in Decentralized Energy Grids
Blockchain technologies incorporated with decentralized energy grids can offer new paradigms in energy management, distribution, and consumption. Trust building among participants in decentralized energy systems is a critical challenge in the absence of a central authority. Foremost, the self-executing, fully autonomous blockchain-based smart contracts and the decentralized automated systems enable real-time, secure, and transparent business transactions. This paper examines the domain of smart blockchain contracts in decentralized energy grids. Smart contracts facilitate self-enforcement, aiding trust-building. Invoice blockchain contracts enable automated energy exchange, thereby facilitating contract execution without third-party involvement. Trust-building, system decentralization, energy producers and consumers bypassing operators, and operational cost reduction for self-sustainability provide system autonomy. The paper also discusses self-governing blockchain contract systems engineering. In smart energy contracts, the paper outlines system challenges, which include load, interoperable infrastructure, and energy consumption. Empirical and case study research, which are the primary focus of the paper, present blockchain contracts for decentralized energy systems as a trust-management and transaction-integration technology.