1

A Study of Quantum-Resistent Cryptography for Long-Term Medical Data Protection
A Sivabalan Settu, Deepak Gupta , Divya Mithun

Abstract - Quantum computing looks set to shake up the basics of our current cryptographic setups. It poses a real risk to systems that safeguard electronic medical records and long-term healthcare archives. Things like genomic sequencing profiles, paediatric developmental histories, neonatal imaging, and biometric identifiers all need strong confidentiality that lasts well over a century. The classical public-key methods we rely on, mainly RSA and ECC, just wont hold up once big quantum computers start running Shors algorithm effectively. This work puts forward a full quantum-resistant security setup designed specifically for national healthcare systems. It pulls together lattice-based encryption, hash-based signatures, blockchain for solid integrity checks, and a special engine to model quantum threats. We ran tests with simulated quantum attackers and experiments mimicking hospital data flows on a large scale. The design shows strong staying power, cutting exposure risk by 92 percent. It keeps system performance in line with what clinical workflows can handle without issues. These results really highlight how pressing it is to start shifting to post-quantum methods. That way we can lock down medical data confidentiality for decades ahead.

2

Respiratory Disease Outbreak Prediction Using Swarm Intelligence in Hybrid Time Series Forecasting Models Prediction
N. Narmadha

Abstract-Outbreaks of respiratory diseases present considerable difficulties to global public health. This paper presents a new method that integrates swarm intelligence into hybrid time series forecasting models to improve the accuracy of forecasting outbreaks. The proposed approach uses the well-established ARIMA model alongside swarm intelligence techniques like Particle Swarm Optimization (PSO) and Firefly Optimization (FFO) algorithms to identify the most optimal hyperparameters. The ARIMA model was tuned with an optimal hyperparameter and then trained. This is reinforced with a residual prediction model, resulting in a significant boost in prediction accuracy. The proposed hybrid ARIMA model is compared with traditional methods to demonstrate its effectiveness in developing more efficient strategies to combat respiratory diseases.

3

Reputation of Artificial Intelligence Enclosure into Climatic Action: Prognostic, Responsive and Moral Solutions
V. Diviya Prabha, S. Vanitha

Abstract- Climate change has become one of the most pressing problems of the modern era that needs rapid and mass and intelligent solutions. Artificial Intelligence (AI) has proved to be an efficient way of combating climate change, as we can make a data-driven decision, make certain predictions, and the programs can be autonomous. This paper focuses the subject of doing research on the use of AI as a form of climate action by proposing different approaches that involve deep learning-based climate forecast, reinforcement learning in resource management through adaptation, and explainable AI in transparent environmental policies. The framework combines satellite images, sensor data, and past climatic models to enhance precision of prediction of extreme weather conditions, observation of carbon emission, and simplifying disaster risks. Along with that, it is implied that the energy efficiency of smart grids and transportation networks can be improved through AI-based optimization algorithms. By means of these approaches, the provided solution will be able to promote the Sustainable Development Goals (SDGs) and in the case, SDG 13: Climate Action. The ethical application, transparency of data, and cross-sector cooperation are also mentioned in the paper to make the AI applications effective and accountable towards achieving climate resilience and sustainability.

4

Impact Response and Residual Performance of Hybrid Graded Sandwich Panels for UAV Structural Applications
Sivabalan Settu, Divya Mithun, M. E. Annadorai

Abstract- Composite sandwich structures are widely employed in unmanned aerial vehicle (UAV) components due to their high stiffness- to-weight ratio; however, their susceptibility to impact induced damage remains a major limitation. Conventional uniform foam and honeycomb cores often exhibit either excessive permanent indentation or localized brittle collapse, leading to significant degradation in post-impact stiffness. In this study, a combined experimental and numerical investigation is conducted on full-scale hybrid graded sandwich panels designed to enhance impact tolerance and residual structural performance for UAV applications. Sandwich panels incorporating carbon-fiber-reinforced polymer (CFRP) face sheets and three core configurations—uniform foam, uniform honeycomb, and hybrid graded foam– honeycomb—were fabricated and tested under 25 J low-velocity impact (n=3 per configuration) followed by residual bending. A validated finite element framework was developed using Abaqus/Explicit, integrating Hashin intralaminar damage for the composite face sheets, cohesive-zone modeling for skin–core delamination, and crushable-foam plasticity for the core. Numerical predictions of impact force history, absorbed energy, and damage morphology showed good agreement with experimental results, with deviations within 10%. The hybrid graded core configuration exhibited a stable force–displacement response, improved energy absorption efficiency, and effective damage confinement. Most notably, hybrid panels retained more than 90-92% of their original bending stiffness after impact, significantly outperforming panels with uniform core architectures. The results demonstrate that graded core designs offer a practical and lightweight solution for improving the impact tolerance and damage resilience of UAV sandwich structures.

5

Streamlit-Gemini AI Powered Customer Support Chatbot: A Multimodal Approach to E-Commerce Assistance
S. Indra

Abstract- Customer support has become an essential part of running a successful online business, especially as customers now expect fast, accurate help around the clock. This paper explores the creation of an AI- powered chatbot designed to make customer service more efficient and user-friendly. Built using Streamlit for a clean and simple web interface, the chatbot leverages Google Gemini AI to generate clear, helpful responses to user queries. It’s specifically tailored to handle questions related to Crocs and Apple products, including topics like order tracking, returns and delivery details. What sets this system apart from traditional support tools is its ability to accept both text and voice input, allowing users to interact in a more natural way. It also supports multiple Indian languages through a built-in translation feature and includes a text-to-speech module to improve accessibility. Customers can even upload product images when reporting issues like damage or delivery mistakes. By combining natural language processing with voice technology, the chatbot delivers quicker responses, greater accuracy and smoother user experience. The findings suggest that tools like this can significantly boost customer satisfaction while cutting down on manual work and operational costs.