Bapatla Engineering College

Wildfires have become a global threat in recent times. Not only do they cause the loss of forest resources and ecological imbalance, however, they also result in the loss of human lives. The economic impact is also substantial, with significant costs associated with property damage, firefighting efforts, healthcare, and loss of productivity. Furthermore, wildfires contribute to climate change by releasing large amounts of carbon dioxide and other greenhouse gases into the atmosphere. Early fire detection can save human lives and mitigate these extensive economic and environmental damages. This proposed work explores the possibility of detecting forest fires from images using a zero-shot approach. In recent times, several visual foundation models have been trained on massive datasets, enabling us to apply these models to a variety of tasks, including zero-shot classification, segmentation, and object detection. The zero-shot approach allows us to classify or detect objects in images without any prior training. Recently, Contrastive Language Image Pre-training (CLIP) models and Large Multimodal Models (LMM) have gained attention due to their excellent zero-shot classification and detection capabilities. With text prompts alone, images can be accurately classified. In this work, we utilize several CLIP models and LLaVA (Large Language Vision Assistant) to classify images as either containing fire or not. The effectiveness of these models is demonstrated by applying them to six wildfire datasets. Classification accuracy is computed, and a comparison of the results from both CLIP and LLaVA models is presented. Simulation results show that these models achieve accuracies beyond 95

Concrete is one of the most versatile materials used in the construction industry in which cement is used as main ingredient playing a dominant role in gaining of strength. Despite of it, the production of cement leads to depreciation of natural materials. Regarding this aspect, in this study an attempt has been made to use GGBS (Ground Granulated Blast Furnace Slag) as a substitute material to cement with a percentage replacement of 50% .Specimens with various geometric shapes like cubes, cylinders and prisms were casted and tested with(High Volume Slag Concrete) and without replacements of GGBS, varying water-cement ratios of 0.55, 0.45, 0.36 & 0.27 for mechanical properties, Young’s Modulus of elasticity, impact strength with and without the addition of steel fibers and also abrasion resistance after 28 days and 90 days of curing.

The present study addresses research gaps on future hourly and quarter-hourly precipitation trends. It studies in global urban context under various SSPs, 1,2,3, and 5 in climatic change and their further implications. In this study, cities considered are Athens, Bogota, Delhi, Rome, and Tokyo. Future daily projected precipitation of 2060 and 2100 is obtained from Copernicus portal. Hourly and quarter-hourly precipitation is computed using Indian Meteorological Department, IMD formula and Bartlett-Lewis (BL) Model from daily precipitation. Bias Corrected Prewhitening(bcpw) and Bootstrapped Mann–Kendall Trend Test with Optional Bias Corrected Prewhitening(pbmk) are applied to find sub-daily projected precipitation trends of data disaggregated with BLM and IMD formulae, respectively, of chosen cities to obtain accurate results. Most of the trend results are either insignificantly increasing or decreasing. However, there is a slope of trend for all the results obtained of cities considered. With IMD disaggregation formula, following trend results are obtained. Athens will have 4.74 × 10–6, 8.21 × 10–6, − 4.78 × 10–6, and − 1.99 × 10–6 hourly precipitation trend slope under SSP126, SSP245, SSP370, and SSP585 in 2060, respectively. Athens will get 6.92 × 10–6, − 1.09 × 10–5, − 3.92 × 10–5, and 4.04 × 10–6 hourly slope under SSP126, SSP245, SSP370, and SSP585 in 2100, accordingly. Bogota will have − 1.47 × 10–4, 8.04 × 10–4, 1.25 × 10–3, and 4.36 × 10–4 hourly precipitation trend slope under SSP126, SSP245, SSP370, and SSP585 in 2060, respectively. Bogota will get 1.30 × 10–3, 1.87 × 10–3, 8.96 × 10–4, and 2.45 × 10–3 hourly slope under SSP126, SSP245, SSP370, and SSP585 in 2100, accordingly. Delhi will have no slope under SSPs 1,2, and 3, while 9.77 × 10–8 hourly precipitation trend slope under SSP585 in 2060. Delhi will face − 6.25 × 10–16 under SSP126, no slope under SSPs 2 and 3, while, − 1.32 × 10–14 under SSP585 in 2100. Rome will be subjected to − 1.64 × 10–5, 1.96 × 10–6, − 4.51 × 10–6, and − 2.10 × 10–5 hourly trend slope in 2060 under SSP126, SSP245, SSP370, and SSP585, respectively. Rome will face 2.51 × 10–6, − 1.30 × 10–6, − 4.26 × 10–5, and − 1.72 × 10–7 hourly slope in 2100 under SSPs 1, 2, 3, and 5, accordingly. Tokyo will get − 2.19 × 10–5, − 4.98 × 10–5, 2.95 × 10–5, and 3.29 × 10–5 hourly trend slope in 2060 under SSPs 1, 2, 3, and 5, respectively. Tokyo will receive 9.82 × 10–6, 2.02 × 10–5, − 2.13 × 10–5, and − 1.23 × 10–5 hourly slope in 2100 under SSPs 1, 2, 3, and 5, accordingly. All trend slopes are in mm/year. However, there is no robust trend of sub-daily precipitation for all the cities considered in 2060 and 2100 under SSPs even with BL model disaggregation method. Further, implications of sub-daily future precipitations trends need to be studied in detail to assess their impacts on extreme events.

In manufacturing and process industries, industrial pumps are essential equipment and can lead to a loss of production, equipment damage and maintenance expenses if they malfunction. It is therefore crucial to detect the abnormal operating condition at an early stage to gain the reliability of the system and to perform condition based maintenance. In this paper we introduce PumpSentinel AI, a hybrid quantum-classical autoencoder framework for unsupervised detection of anomalies in multivariate pump sensor data from an industrial facility. The framework proposed in this work uses a classical encoder to compress the sensor signals into a low dimensional latent representation, followed by a four qubit variational quantum circuit that models the interactions of complex features. A quantum generated latent representation is then reconstructed using a classical decoder, and the reconstruction error is used to quantify the deviation from learned normal operating behaviour as the anomaly score. The model is trained only using normal operational samples and wants to avoid to rely on balanced labelled fault dataset, which is suitable for an industrial environment that may have limited amount of failures. Before training the model the preprocessing pipeline includes feature selection, normalizing and interpolating for missing values. The threshold is calculated from the distribution of the reconstruction errors for normal samples; it is used to separate normal and abnormal operating conditions. The proposed framework is also incorporated into a RESTful monitoring infrastructure based on Flask to enable accessing for authenticated users, real-time inference on single records, batch CSV analysis, historical anomaly visualization and logging of predictions for practical deployment. The proposed system essentially features latent representation learning with quantum enhancement and reconstruction-based anomaly detection without the need of manually labeled fault data, which offers an efficient and scalable framework for intelligent health monitoring of industrial pumps.

The multicomponent glass system 3Bi(2)O(3)-10CaO-20P(2)O(5)-(67-x)B2O3: xV(2)O(5) (0 < x < 0.9 mol%) was fabricated by heat treatment techniques. These glasses were investigated by XRD, FTIR, ESR, optical absorption, and dielectric properties. XRD patterns authenticated the glassy phase of specimens. The density of glasses is increased from 2.720 to 2.735 g/cc with increasing concentration of V2O5. The vanadium ion concentration (N-i), interionic distance (R-i), and polaron radius (R-p) of glasses were estimated. The optical absorption spectra displayed a prominent band at 667 nm owing to B-2(2g) -> B-2(1g) transitions of vanadium ions. The direct and indirect optical bandgap energies (E-o) of specimens exhibited a decreasing pattern, but Urbach energy (Delta E) showed an opposite trend. The least value of E-o (indirect) is 2.249 eV where as the highest value of Delta E is 0.551 eV. ESR results confirmed presence of vanadium (V4+) ions, and the estimated spin-Hamilton parameters indicated the octahedral coordination. FTIR spectra specified the structural modifications, including gradual conversion of BO4 units into BO3 units. Dielectric studies were performed over 30-300 degrees C and 1 kHz-1 MHz. Increases in dielectric parameters of glasses dielectric constant (epsilon), dielectric loss (tan delta), and a.c. conductivity (sigma(ac)) and were detected with increasing V2O5 content. The maximum value of sigma(ac) is 4.971 & times; 10(-7) (ohm-cm)(-1) for the glass doped 0.9 mol% of V2O5. This glass is widely used in fabricating photo detectors, optical switches, light sensors, smart windows, display panels, and solid state batteries.