St. Francis Institute of Technology (SFIT) in Mumbai, India is an engineering college named after Francis of Assisi, the 12th-century Italian saint. The college is accredited by the National Board of Accreditation, approved by the AICTE and is affiliated to University of Mumbai.The college is run by the Society of Franciscan brothers, with a special consideration to Roman Catholic students. It offers undergraduate, post-graduate and doctoral courses in Engineering. It is the first Catholic technical institute in India to acquire Minority Status..
A tremendous growth of Internet of Things (IoT) devices and the corresponding requirement of low-latency services have resulted in cloud-based standalone architectures to be insufficient to fulfil the demand for realtime commerce applications. This study investigates the computing efficiency of Cloud-only, Edge-only as well as Hybrid Edge-Cloud system using a digital commerce notice delivery system developed on a Raspberry Pi. Two prototype models are developed and evaluated empirically. Model 1 deploys Google Drive as the cloud backend with passive edge caching over Web Sockets. On the other hand, Model 2 employs Supabase as a structured cloud database combined with RFID-based edge processing activity over MQTT. Latency measurements across multiple trials resulted in a Hybrid Edge Cloud approach with lower latency in displaying as well as uploading. The results indicated that active edge intelligence, combined with lightweight messaging protocols and structured cloud storage, substantially enhances computational efficiency in digital commerce deployments.
With the growing demand for smart and sustainable infrastructure, educational institutions require intelligent solutions that ensure both security and energy efficiency. Integrating IoT-based automation with RFID-enabled access control offers a robust framework for managing classrooms and laboratories effectively. Traditional classroom and laboratory setups often result in unnecessary power consumption, while conventional lock-and-key systems expose facilities to risks such as unauthorized entry, key duplication, and administrative inefficiencies. To address these limitations, this project introduces an integrated solution with two core objectives. The first objective proposes Laboratory Automation System (LAS) that employs infrared sensors, NodeMCU ESP8266 modules, and the Blynk platform to automate the operation of lighting and fans based on human occupancy. In the second objective an RFID-enabled door access system that integrates RFID readers and tags with existing locking mechanisms to provide authorized entry. The system supports remote programmability for administrators to grant or revoke access permissions, while maintaining centralized logs for monitoring. The project provides a holistic, low-cost, and scalable framework that unifies IoT-based automation for energy efficiency and RFID-based security for access control, creating a smart, secure, and sustainable educational infrastructure.
The digital world has been reshaped, largely thanks to the swift expansion of e-commerce and Over-The-Top (OTT) streaming services. These platforms use incentive mechanisms that can strongly influence consumer behaviour. Usually these two platforms operate independently, which limits coordination and reduces the potential impact of shared strategies. This paper proposes a way to integrate a framework that links online shopping activity with OTT content access through a profit-aware reward mechanism. A rule-based engine evaluates items in a user’s shopping cart based on its product category, profit margin, and purchase volume, and assigns reward points accordingly. These points can be redeemed for time-limited access to movies and series on a connected streaming platform, creating a continuous engagement cycle between commerce and entertainment. In contrast to conventional loyalty programs that provide fixed rewards, the proposed system dynamically modifies incentives to achieve a balance between customer satisfaction and business profitability. The experimental results based on transaction data show that this approach encourages customers to buy more products from high-margin categories. This model offers a fresh strategy for enhancing customer retention, fostering more personalized interactions, and ultimately, driving revenue growth within digital environments.
The G.723 audio codec uses error concealment as way to regain the lost, corrupt audio data, and maintain clear speech by continuously transmitting audio, despite the network failures of a transmission, this method significantly improves speech quality at low bit rates by utilizing a lossy speech codec, thereby providing a smoother and more reliable listening experience; however, the effectiveness of the error concealment will diminish under high loss environments, introducing audio artifacts, and degradation when the recovery did not meet requirements in high volume. To improve these elements, in this paper we propose advanced error concealment in the G.723 audio codec, by the use of Nested UNet Generative Adversarial Network (Nested UNet-GAN), and Adaptive Neuro-Fuzzy Inference System (AERC-NUNet-GAN-ANFIS). Essentially, the primary goal of the proposed technique is to increase the error resilience, and quality of audio transmission in poor network conditions using the G.723 codec. In the earlier stages of the methodology, speech samples from the DARPA TIMIT Acoustic–Phonetic Continuous Speech Dataset are used as input data. To achieve semi-realistic conditions and mimic real-world transmission errors, Bit Error Rates (BERs) of 2
The secure, interoperable, and regulation-compliant management of electronic health records (EHRs) continues to be a critical challenge in modern healthcare due to fragmented data governance, privacy breaches, and limited scalability of centralized digital infrastructures. Existing blockchain-based healthcare systems, although promising, often suffer from performance bottlenecks, weak access control enforcement, and a lack of integration with healthcare interoperability standards such as FHIR/HL7. These limitations hinder the practical deployment of blockchain systems in real-world clinical environments. To address these challenges, this study proposes FabricMedChain, a decentralized healthcare framework built on Hyperledger Fabric for privacy-preserving, auditable, and scalable medical data management. The proposed model integrates a Delegated Proof of Stake (DPoS) consensus mechanism to enhance throughput and reduce latency, a chaincode-level Role-Based Access Control (RBAC) system to enforce patient consent and policy-driven data governance, and FHIR-compatible transaction mapping for standardized interoperability. Experimental evaluation using Hyperledger Caliper shows that FabricMedChain achieves up to 3.5× lower latency (0.07 s vs. 0.24 s), 50
