Loading...
Pushpendra Kumar Sharma is an Assistant Professor in the Department of Electrical Engineering at Jagannath University, India. His contribution focuses on cloud computing architectures, scalable AI applications, and intelligent cloud infrastructure.
The swift expansion of Artificial Intelligence (AI) has generated a heightened need for computational resources that can accommodate extensive datasets, intricate algorithms, deep neural networks, and real-time intelligent applications. Cloud computing offers a flexible and expandable framework for the development, training, deployment, and management of AI systems. Cloud architectures facilitate the development of AI systems for organisations by integrating virtualised computing resources, distributed storage, high-performance accelerators, containerisation, microservices, serverless computing, and managed machine-learning platforms, thereby eliminating the need for substantial physical infrastructure. This chapter outlines the essential principles of cloud computing for artificial intelligence, scalable cloud frameworks, management of AI workloads, distributed model training, inference services, cloud storage solutions, container orchestration, MLOps, security measures, resource optimisation, and real-world applications. It also examines nascent themes such serverless AI, edge-cloud integration, federated learning, foundational models, and eco-friendly AI computing.
This section offers an extensive structure for self-governing entities in crisis intervention and emergency administration. It analyses disaster-response frameworks, situational awareness, integration of multimodal data, damage evaluation, support for search-and-rescue operations, resource distribution, planning of emergency routes, assistance with evacuations, management of communications, coordination of healthcare, logistics, monitoring of infrastructure, unmanned aerial vehicle and robotic systems, edge-cloud frameworks, and coordination among multiple agents. The chapter also explores human-in-the-loop control, interpretability, uncertainty management, cybersecurity, privacy, ethical implications, resilience, and assessment approaches. A conceptual multi-agent framework is established wherein specialised agents cooperate in the domains of sensing, evaluation, planning, logistics, communication, and emergency response. Special attention is given to constrained autonomy due to the substantial impact that emergency decisions can have on human safety. Future investigative pathways concerning disaster digital twins, multimodal foundational models, autonomous rescue robotics, federated learning, robust communication, and cooperative multi-agent emergency systems are also examined. The section illustrates how self-governing agents can enhance emergency management effectiveness by facilitating information synchronisation, alleviating regular tasks, and aiding prompt, data-driven operational choices.