Jhulelal Institute of Technology

Synthesis and luminescence properties of Li2CeO3:Eu3+ phosphors are described. By virtue of host sensitization, this phosphor exhibited a broad band excitation which overlapped very well with 365 nm chip output. It was successfully used to prepare blends which lead to fabrication of white LED lamps. Two types of lamps were designed. Blend of Li2CeO3:Eu3+:Cs3ZnCl5:Mn2+: Sr5(PO4)3Cl:Eu2+ was used to produce a white LED lamp with improved CRI of 82. Another blend of BaAl2O4:Eu2+ and Li2CeO3:Eu3+ provided a lamp with even better properties. A human centric lighting could be achieved with this blend and 365 nm chip. Circadian action factor (CAF) could be brought down from 1 to 0.432.

Abstract — This paper presents the design and implementation of a Smart Household Management System (SHMS), an IoT based platform that integrates real-time environmental sensing, safety alerting, energy monitoring, and household productivity management into a unified web interface. The system employs an ESP32 microcontroller interfaced with a DHT11 temperature and humidity sensor, an MQ-2 gas concentration sensor, and a voltage sensing module. Sensor data is published over a local Wi-Fi network via HTTP and consumed by a Python Flask backend that aggregates, enriches, and exposes the data through RESTful APIs. A responsive single-page web dashboard renders live readings, sparkline trend visualizations, safety alerts, task management, and expense tracking. Threshold-based automated safety alerting triggers critical notifications when gas concentrations exceed 9,000 PPM or ambient temperature surpasses 45°C. Experimental results demonstrate stable real-time operation with a 3- second data refresh cycle, effective noise reduction through ADC sample averaging, and a reliable browser based interface for household monitoring. This work demonstrates an accessible, low-cost approach to smart home automation suitable for residential environments, with clear pathways for enhancement in security, persistence, and scalability. Keywords — Internet of Things (IoT); ESP32; smart home; environmental monitoring; gas detection; Flask; real-time systems; DHT11; MQ-2 sensor; home automation.

The requirement of effective information retrieval mechanisms has been brought to the fore in the modern datadriven world where information is excessive and varied. One of the pillars that have been formed to cope with this challenge has been Natural Language Processing (NLP) technologies where computers can understand, interpret, and produce human language. Question Answering (QA) systems are one of the areas of interest in the variety of NLP applications. QA systems are the beginning of a paradigm change in the way we communicate with information. Historically, users have been using key word searches or browsing databases manually to find the desired information. Nonetheless, the QA systems allow the users to ask questions using the natural language and get correct and contextually appropriate answers. The paradigm shift leads to better user experience and ensures much better efficiency and effectiveness of information retrieval procedures. QA systems development and assessment are of enormous importance in numerous areas. The aim of this study on QA systems based on the Distilled Bidirectional Encoder Representations from Transformers (distilbert) is the model distilbert-base-uncased, distilbert-base-cased-distilled-squad, and fine-tuned-distilbert, which aims at improving the extraction of valuable information in text documents. The main task is to create systems that can effectively retrieve and comprehend queries posed by the users to deliver the correct response according to the Stanford Question Answering Dataset (SQuAD). Our model had a validation accuracy of 89.86 after fine-tuning, where precision was 81.61, and recall was 80.99, thereby demonstrating that the research was successful.

Abstract—Health insurance premium pricing remains one of the most complex and consequential challenges in the global healthcare and financial services sectors. Premiums directly determine the affordability and accessibility of health coverage for individuals, families, and enterprises, while simultaneously dictating the financial viability and risk exposure of insurance providers. Despite its critical importance, the conventional process of premium determination relies heavily on rule-based actuarial tables and manual underwriting protocols that are rigid, opaque, and often inadequate in capturing the multidimensional nature of individual health risk. This paper presents a comprehensive machine learning-based Health Insurance Premium Prediction System that integrates demographic attributes, lifestyle indicators, geographic factors, and medical history variables to estimate insurance premiums in an accurate, transparent, and personalized manner. The proposed system trains and rigorously compares four supervised regression algorithms—Linear Regression, Decision Tree Regression, Random Forest Regression, and XGBoost Regression—on a real-world structured healthcare dataset of 1,338 records sourced from the Kaggle Medical Cost Personal Dataset. Comprehensive preprocessing including missing value treatment, feature encoding, normalization, and feature Error (MAE) of 1,978 USD, and Root Mean Square Error (RMSE) of 3,312 USD on the held-out test set. SHAP (SHapley Additive exPlanations) value analysis is employed to interpret model predictions and quantify individual feature contributions, confirming that smoking status, age, BMI, and number of dependents are the dominant risk factors. Beyond prediction, the system incorporates a three-tier risk classification engine (Low, Moderate, High Risk) and is deployed as an interactive web application accessible to policyholders, insurance agents, and healthcare organizations. Future directions include integration of real-time wearable health data, federated learning for privacy-preserving distributed training, and deep learning architectures for longitudinal risk modelling. Keywords—health insurance premium prediction, machine learning, supervised regression, Random Forest, XGBoost, SHAP explainability, risk categorization, actuarial pricing, healthcare analytics, feature engineering

ABSTRACT: ELECTRICAL SAFETY IS VERY IMPORTANT IN OUR HOMES, OFFICES AND FACTORIES. WE USE ELECTRICAL DEVICES, AND THEY CAN BE RISKY IF NOT USED PROPERLY. THIS STUDY IS ABOUT DEVICES THAT PREVENT SHOCKS AND FIRES. WE LOOKED AT EARTH LEAKAGE CIRCUIT BREAKER (ELCB) AND RESIDUAL CURRENT CIRCUIT BREAKER (RCCB). THESE DEVICES HELP KEEP US SAFE BY STOPPING THE POWER SUPPLY WHEN THERE’S A PROBLEM. ELCB IS A DEVICE THAT CHECKS THE VOLTAGE BETWEEN THE EARTH AND ELECTRICAL EQUIPMENT. IF THERES A PROBLEM IT STOPS THE POWER.. IT NEEDS GOOD EARTHING AND IS NOT VERY SENSITIVE. RCCB IS A DEVICE THAT CHECKS THE CURRENT IMBALANCE BETWEEN LIVE AND NEUTRAL CONDUCTORS. IT'S MORE SENSITIVE AND RELIABLE THAN ELCB. WE ALSO LOOKED AT DEVELOPING DEVICES THAT CAN DETECT PROBLEMS FASTER AND MORE ACCURATELY. THESE DEVICES CAN WORK WITH ELECTRICAL SYSTEMS AND PROVIDE BETTER PROTECTION. WE FOUND THAT RCCBS ARE MORE EFFECTIVE THAN ELCBS. WE NEED TO KEEP IMPROVING THESE DEVICES TO MAKE THEM MORE RELIABLE AND EFFICIENT. THE STUDY ALSO EXPLORES HOW TO MAKE CORRECTIVE DEVICES BETTER. THIS INCLUDES MAKING THEM MORE SENSITIVE, TRIPPING FASTER AND WORKING WITH PROTECTIVE DEVICES. ADVANCED RCCBS USE TRANSFORMERS TO DETECT SMALL LEAKAGE CURRENTS AND RESPONDS QUICKLY. THIS KEEPS US SAFER. IN CONCLUSION ELECTRICAL SAFETY DEVICES LIKE ELCB AND RCCB ARE CRUCIAL. RCCBS HAVE REPLACED ELCBS BECAUSE THEY ARE MORE ACCURATE AND RELIABLE. WE NEED TO KEEP DEVELOPING DEVICES TO PREVENT ELECTRICAL ACCIDENTS.