Dr. D. Y. Patil Institute of Technology, Pimpri, Pune, was established as Dr. D. Y. Patil Women's College of Engineering in July 1998. The college was converted into a co-education college with the name Dr. D. Y. Patil Institute of Engineering & Technology, Pimpri, Pune in the academic year 2002–2003. The institute is situated in the vicinity of a Pimpri-Chinchwad industrial belt, which is one of the biggest industrial belts in Asia. The college is affiliated with Savitribai Phule Pune University. Ranked 172 in the engineering category by the National Institute of Ranking Framework (NIRF) in 2021.The college is named after Dr. D. Y. Patil, former Governor of Bihar State, Republic of India. It has a hostel for both boys and girls. The girl's hostel is just opposite the main gate of the complex and the boys hostel is a few miles away.
The enormous development of Internet of Things (IoT) devices poses unique challenges for interconnected systems in relation to privacy and security. The increasing number of cyber-attacks has made scalable intrusion detection systems (IDS) mandatory for all organizations. When dealing with cyber-attacks, existing methodologies have struggled to capture spatial and temporal features. When the number of edge systems rises, tradeoffs exist due to a lack of diversity. Thus, Federated Learning (FL) based IDS is proposed to ensure privacy preservation, accurate detection of attacks, and cost reduction, which are difficult in old centralized machine-learning-based IDS. The FL-based model offers privacy benefits but still faces vulnerabilities that are evaded by adversaries through the cooperation of user data. Therefore, ensemble models were evaluated as possible solutions for improving attack detection and classification. As a result, the proposed study introduced an ensemble-based deep federated learning model (E-DFL) driven IDS framework for IoT applications. The concept uses a residual auto-encoder, attention capsule network, and stacked bidirectional long short-term memory (Bi-LSTM) as basis classifiers to generate ensemble data. The meta-classifier then performs the final intrusion detection on the test set using a global model created by combining ensemble model updates from different clients. Compared to classical FL approaches, the E-DFL model provides improved user privacy. Edge-IIoT cybersecurity datasets were used to assess the effectiveness of the proposed E-DFL method, and achieved an accuracy of 99.23%.
The present study investigates the performance enhancement of photovoltaic–thermal (PVT) systems using a comparison of conventional water coolant and Al2O3–water nanofluid coolant. Outdoor experiments were conducted in Pune, Maharashtra, India, by varying Al2O3 nanoparticle concentrations in a spiral copper thermal absorber integrated with a polycrystalline PV module. The objective was to evaluate how coolant type and nanoparticle loading influence overall PVT system efficiency with practical feasibility. The study uniquely combines thermal, electrical, hydraulic considerations to identify the realistic applicability of Al2O3 nanofluid cooling in PVT systems with spiral thermal absorber. This holistic approach provides practical insights for optimizing PVT performance in domestic and industrial applications under Indian climate conditions—representing a significant advancement over existing studies. The higher thermal conductivity of the Al2O3 nanofluid facilitated more effective heat extraction from the PV surface, resulting in reduced operating temperature, improved electrical output, and enhanced thermal energy recovery. Al2O3 water-nanofluids proved particularly effective, enhancing thermal conductivity and facilitating heat recovery of 5–12 °C. Improvements were observed in thermal efficiency (57.51
A driving cycle is a graphical representation of speed of a vehicle versus time. A real-time driving cycle is a pattern that represents the actual driving behaviour of vehicles on the roads, as observed in real-life traffic conditions. To promote for sustainable public transportation, electric buses becoming a popular mode of transportation. It plays a significant role in reducing emissions and improving air quality. Real-time driving cycle of electric bus needs to be developed for Indian cities to estimate actual energy consumption, as driving cycles vary by city and mode of traffic. This article reviews the literatures and identifies areas for further research. By collecting real-time data and analyzing research results, it will helpful for electric bus energy estimation and further helpful for energy optimization.
As cyber threats continue to advance, traditional single-factor authentication methods—such as relying solely on passwords—have become increasingly insufficient for protecting sensitive user information. This is particularly critical on social media platforms, where maintaining privacy and security is essential. This research highlights the rising significance of multi-factor authentication (MFA) as a proactive and robust solution to strengthen digital security. The proposed system, Enigma, is a secure social media web application designed with layered authentication mechanisms to enhance user trust and safeguard data. By integrating multiple verification factors, the system reinforces user authentication and minimizes the likelihood of unauthorized access. In addition to conventional password-based login, Enigma incorporates advanced methods such as time-based one-time passwords (OTPs) and passkey-based authentication—an emerging passwordless and biometric-compatible approach. This combination effectively mitigates vulnerabilities commonly associated with traditional login methods, including phishing attacks and credential leaks.
Due to the rapid increase in the use of IoT devices in the industrial sector, there is a growing demand for advanced predictive maintenance systems that can predict equipment breakdowns even before their occurrence. This research paper presents a new approach known as Explainable Deep Learning Framework for Predictive Maintenance (XDL-PM), which is based on the combination of LSTM networks, CNNs, and explainability mechanisms based on SHapley Additive exPlanations (SHAP) and Local Interpretable Model-Agnostic Explanations (LIME). Our proposed architecture can analyze IoT sensors such as vibration sensors, acoustic emission sensors, temperature sensors, and current sensors. Experimental analysis using benchmarks has shown that our proposed model has an accuracy of 97.4%.
