The problem of efficient slotting in the warehouse under multiple objectives and constraints is still an open problem, especially in high-velocity environments, taking into account the weight of items, their frequency of retrieval, and the duration of storage. The existing rule-based and conventional reinforcement learning techniques and methodologies have shown limited potential in efficiently solving this problem, considering the spatial and semantic dependencies and the decision structure under the influence of the imposed constraints. In this paper, a novel reinforcement learning framework, namely DTQNetwork, is introduced, which incorporates a multi-head transformer encoder with a Double Deep Q-Network. The framework is targeted at developing an intelligent system for pallet placement in the warehouse. The introduced framework incorporates the existing constraints, including weight stability, urgency, and frequency, using a carefully designed reward and penalty system. The framework is implemented and tested using a custom OpenAI Gym environment, simulating a dual-section 14 × 14 warehouse grid with Automated Guided Vehicle operations. The results show significant improvements over existing methodologies, including a 60 % improvement in slot utilization and 85 % reduction in time. The framework is expected to provide a solution for developing intelligent storage systems, considering the existing and emerging challenges and complexities. First, we integrate a multi-head Transformer encoder into a Double DQN pipeline for smart warehouse slotting. Second, we design a custom Gym-based simulation framework that enforces physical and operational constraints during both training and inference. Third, through detailed benchmarking and ablation studies, we demonstrate the applicability of DTQNetwork to real-world warehouse optimization.

Timely detection of pests play a major role in agriculture. There exist many pest identification systems, but almost all of them suffer from the misclassification due to lighting, background clutter, heterogeneous capturing devices as well as the pest being partially visible or in the different orientation. This misclassification may cause tremendous yield loss. To mitigate this situation, we proposed an architecture to provide high classification accuracy under the aforementioned conditions using morphology and skeletonization along with neural networks as classifiers. We have considered the crop rice as a use case as it is the staple food grain of almost the entire population of India. The amount of pesticides used is highest in rice as compared to all other food grains. This paper offers a robust technique to identify the pests in rice crops. The performance of the proposed architecture is tested with an image dataset, and the experimental results reveal that our proposed approach provides better classification accuracy than the existing pest detection approaches in the literature. Furthermore, the experimental results also provide the performance comparison among the popular classifiers.

Agriculture is crucial for food production globally, especially with the growing population. Efficient management is essential to meet the increasing demand for food. In some regions, agriculture relies solely on unpredictable monsoon rainfall for water. To meet crop water needs, irrigation delivers water based on soil type and crop requirements, contributing to water conservation. The smart irrigation system utilizes edge and fog computing to provide a cutting-edge solution for efficient water management in agriculture. Sensors collect data on soil moisture, temperature, and humidity, and process it locally at the edge for immediate responses to changing conditions. The data is then analyzed by fog nodes to provide comprehensive insights and facilitate predictive analytics. This approach not only optimizes water usage, ensuring crops receive precise irrigation based on real-time needs but also enhances scalability and reliability. The system demonstrates significant potential in promoting sustainable agricultural practices, reducing water wastage, and increasing crop yields. This paper uses an Arduino Uno R3 microcontroller board based on the ATmega328P to implement the system. This study also utilizes Deep Learning (DL) to predict the irrigation system. The successful implementation and results of the Long Short Term Memory (LSTM) based DL model highlight a promising direction for future research in smart irrigation and environmental sustainability.

Sharding implementations use conservative approximations for determining the number of cloud instances required and the size of the shards to be stored on each of them. Conservative approximations are often inaccurate and result in overloaded deployments, which need reactive refinement. Reactive refinement results in demand for additional resources from an already overloaded system and is counterproductive. This paper proposes an algorithm that eliminates the need for conservative approximations and reduces the need for reactive refinement. A multiple linear regression based machine learning algorithm is used to predict the latency of requests for a given application deployed on a cloud machine. The predicted latency helps to decide accurately and with certainty if the capacity of the cloud machine will satisfy the service level agreement for effective operation of the application. Application of the proposed methods on a popular database schema on the cloud resulted in highly accurate predictions. The results of the deployment and the tests performed to establish the accuracy have been presented in detail and are shown to establish the authenticity of the claims.

The proposed solution includes a detailed and novel description of a data mining model applied to the educational environment and an intuitive and easy-to-use software tool that implements it. This model starts with the presentation of the educational problem to be analyzed and ends by providing the alternatives of possible solutions, i.e., an end-to-end model. What is interesting and novel is that most of the complex tasks to be performed to achieve the objective have been automated using techniques such as Exploratory Data Analysis, ETL’s Scripts, Automated Machine Learning, and Automatic Interpretability and Explainability. In addition, the model takes into consideration the ethical standards that must be met in terms of privacy and security. This solution allows institutions to find patterns within the educational big data that allow them to understand the complexity of their problems and the main factors that generate them.

Traffic sign classification remains a significant challenge for autonomous vehicles, particularly in recognizing targets of various scales and achieving real-time performance. The varying sizes of traffic signs can negatively affect detection accuracy. Additionally, changes in lighting when capturing input images can pose challenges for existing methods. Therefore, this research presents a new approach to enhance the performance of current traffic sign classification systems by combining the benefits of existing machine learning-based and deep learning-based methods through ensemble learning techniques. Experimental results show that the proposed method improved the performance of existing methods using the German Traffic Sign Recognition Benchmark dataset. Under normal input conditions, the proposed method achieved an accuracy of 95.77

Client-centric consistency models define the view of the data storage expected by a client in relation to the operations done by a client within a session. Monotonic reads is a client-centric consistency model which ensures that if a process has seen a particular value for the object, any subsequent accesses will never return any previous values. Monotonic reads are used in several applications like news feeds and social networks to ensure that the user always has a forward moving view of the data. The idea of Monotonic reads over multiple copies of the data and for lightly loaded systems is intuitive and easy to implement. For example, ensuring that a client session always fetches data from the same server automatically ensures that the user will never view old data. However, such a simplistic setup will not work for large deployments on the cloud, where the data is sharded across multiple high availability setups and there are several million clients accessing data at the same time. In such a setup it becomes necessary to ensure that the data fetched from multiple shards are logically consistent with each other. The use of trivial implementations, like sticky sessions, causes severe performance degradation during peak loads. This paper explores the challenges surrounding consistent monotonic reads over a sharded setup on the cloud and proposes an efficient architecture for the same. Performance of the proposed architecture is measured by implementing it on a cloud setup and measuring the response times for different shard counts. We show that the proposed solution scales with almost no change in performance as the number of shards increases.

The advancement in technology and the increase in usage of Internet access has revolutionized the landscape of agriculture using E-Commerce. Several E-Commerce websites are operative in India to promote uniformity in agricultural marketing across the integrated markets by removing information asymmetry between buyers and sellers. Stakeholders are reluctant to utilize this new technology for trading agricultural produces in spite of close opportunities. Pricing mechanism of the online trading portals neither generates maximum revenue during high demand and less supply nor ensures minimum loss due to the decay or down selling of the products. Static pricing mechanism prevents the sellers from joining this online system, as it does not provide many benefits to an online customer. A continuous adjustable dynamic pricing mechanism that can adapt the market condition and quality degradation is crucial for maintaining the seller revenue and customer interest. This paper explains several existing dynamic pricing mechanisms and analyzes their relevance in the field of agro-marketing. In this paper, several research challenges on dynamic pricing approach of E-Commerce have been summarized. The factors like demand, supply, and freshness of the agri-products must +be considered for the development of a pricing mechanism in the dynamic environment of E-Commerce.

Automated traffic monitoring is an essential system in our daily life offering numerous benefits. It helps drivers stay aware of their speed, reducing the risk of accidents and saving lives, while also aiding law enforcement in effectively regulating traffic. The goal of this research aims to compare the evaluation of the car detection while estimating the speed of both Haar Cascade and Yolov8 methods. Experimental results demonstrated that our proposed one outperformed with the accuracy MAE is about 0.77, along with the precision 93

Many plant diseases manifest visible symptoms, typically diagnosed by experienced plant pathologists who visually inspect infected plant leaves. However, this manual diagnostic process is slow and heavily dependent on the pathologist's expertise, highlighting its suitability for computer-aided diagnostic systems. Unlike traditional machine learning approaches that demand meticulous manual feature extraction, there is a demand for models capable of successful classification without extensive preprocessing. In this research, we introduced an efficient module that leverages the advantages of vision transformer and weighted feature fusion across raw images, enhanced images, and edge information. Experimental results demonstrated that our proposed system outperformed existing methods such as EfficientNet or DenseNet. Specifically, our proposed system achieved a detection rate and false alarm rate of 98.18

Digitization is the process by which information is packaged into discrete bits that can be individually addressed. This makes the information machine-readable. The two most crucial factors in today’s world, when everyone is going toward digitization, are data protection and risk management. The transmitted data must be secured against any malicious attackers. Cryptography and steganography are the techniques for addressing the data-hiding process. Unlike cryptography, which transfers data in an encrypted format, steganography refers to confidential communication where the existence of the information, if any, is concealed. The JPEG image is used as the paper’s canopy. Koch’s snowflake fractal with a fourth iteration value is taken into account to provide a stego key. The text and fractal embedding within the host image of fractals have the distinctive property of neither changing the host image nor raising any red flags. The research is novel since it uses image steganography to make it more imperceptible. The host image maintains its mathematical behavior during embedding. Convolutional Neural Network (CNN) model is applied to enhance stego-image analysis and optimize embedding quality. The experimental results demonstrate that, in terms of capacity, robustness, and other statistical parameters, the proposed method performs superiority perceptually. The research is beneficial for hiding an image and sensitive data.

While Retrieval-Augmented Generation (RAG) has shown impressive results across various applications, including question answering, summarization, and dialogue generation, it also has limitations, especially when it comes to personalized answers. One significant challenge is that RAG models typically rely on generalized retrieval methods and do not always tailor responses to individual user profiles or preferences. This can result in answers that lack personalization, failing to account for the user’s specific context, needs, or past interactions. Therefore, this research introduces a new approach to improve the performance of existing RAG-based systems by incorporating three main features: question rewriter, multi-embedding, and user profile information. Experimental results on three datasets—HotPotQA, SQuAD 1.1, and the Vietnamese AI Vietnam dataset—show that the proposed system improves the performance of existing RAG-based systems in terms of generating personalized answers.

As digital commerce settles among consumers, the need to capture these users grows. The demand for techniques, tools and strategies grows as digital sales channels multiply and as consumers continue to adopt new online shopping habits. This is why ecommerce specialists today position themselves as the most demanded profiles in the commercial area. Due to the high growth rate of e-commerce globally, companies that did not have knowledge about this marketing channel had to help themselves to continue their business competitively. For this purpose, a quality model of workflows, metrics and indicators is proposed based on quality standards and information collected from the software industry and computer services in the region, emphasizing an integration architecture with other systems to concretize the digital transformation of companies. companies.

Arrhythmia is a common cardiac condition characterized by irregular heartbeats that, if left untreated, can have major health effects. In this paper, we present a random forest-based binary classification method for identifying arrhythmia. The random forest approach is an ensemble learning methodology that can handle large datasets with little overfitting, making it a good choice for analysing medical data. Moreover, the interpretability of the model is further improved by the Explainable AI in the form of decision tree analysis, which offers vital insights into the individual ECG aspects impacting the classification choices. The science of arrhythmia diagnosis and patient treatment is finally advanced because of this openness, which also helps physicians make wise decisions based on the model’s predictions. The experiment was performed on the MIT-BIH ECG dataset by applying a 75–25 split. The accuracy of 98.99 was achieved using random forest classifier, and then using decision tree explainer, we found out the most significant features for the classification, and the explanation tree illustration helps decode the reason for the classification of the data for normal or arrhythmia classes. This explainer model will help the healthcare personnel to better understand and take decision for the betterment of the patients.

We use ontology to capture the knowledge about domains. There are considerable number of ontologies which have been developed. To develop a new ontology, we use the existing concepts of the ontologies according to the domain needs. The reusability of the concept maintains shareability of knowledge across the domains and prevents multiple interpretation of the defined concept. An Internationalized Resource Identifier (IRI) of the concept is used to connect the various concepts with each other. This work aims to provide an IRI-Debug tool that enables the ontologist to validate their crafted ontology against the standard ontologies using IRI and gauging to what extent their ontology’s concepts or properties, if reused, compliant to the standard ontologies. This tool allows user to select a desired appropriate ontology from the available ontologies and validate the developed ontology with respect to standard ontologies.

Smart manufacturing drives innovation and competitiveness in production. This paper describes a smart manufacturing system including a smart factory, smart warehouse, and autonomous mobile robots (AMRs) to connect them. The system uses advanced technologies like the Manufacturing Execution System (MES), digital twin, Radio Frequency Identification (RFID), Internet of Things (IoT), Warehouse Management System (WMS), and Artificial Intelligence (AI) to improve product quality, reduce lead times, and increase customization. The system improves business experts’ skills and provides excellent educational chances to learn about modern manufacturing technology. The study underlines the potential for unique and important applied research possibilities and industry-academia collaborations by integrating the Festo CP-Factory system and the smart warehouse system into the academic scene. However, such a system involves careful planning, design, and integration of varied components, as well as data security, privacy, and human–robot interaction issues. Smart manufacturing is a major change from traditional production methods and might give companies that adopt it a competitive edge. Technological advancements are predicted to make the smart manufacturing system more accessible, affordable, and scalable, helping many industrial firms.

Knowledge representation and reasoning is a field of ‘Artificial Intelligence’ that encodes knowledge, beliefs, actions, feelings, goals, desires, preferences, and all other mental states in the machine. An ontology is prominently used to represent knowledge and offers the richest machine-interpretable (rather than just machine-processable) and explicit semantics. Ontology does not only provide sharable and reusable knowledge, but it also provides a common understanding of the knowledge; as a result, the interoperability and interconnectedness of the model make it priceless for addressing the issues of querying data. Ontology work with concepts and relations that are very close to the working of the human brain. Ontological engineering provides the methods and methodologies for the development of ontology. Nowadays, ontologies are used in almost every field, and a lot of much research is being done on this topic. The paper aims to elaborate on the need of ontology (from data to knowledge), how does for ontology (from data to knowledge), how semantics come from logic, the ontological engineering field, history from hypertext to linked data, and further possible research directions of the ontology. This paper benefit reader who wishes to embark on ontology-based research and application development.

Current developments of intelligent software technology have been playing an important role in a broad range of innovative applications. Major industries such as automotive, telecommunications, banking, e-business, automation, and recent smart applications in artificial intelligence (AI), machine learning (ML), deep learning, robotics, internet of things and smart devices, smart environment and technology, and many others are highly dependent on intelligent computer software for their basic operations. Moreover, the integration of artificial intelligence with machine learning, deep learning, and block chain are proving to be quite a powerful combination for improving virtually every industry in which they are implemented. AI and ML applications to academics to automobiles to railways to Aviation to healthcare industries are helping to significantly improve reliability, services, safety, and protection to human life and lifestyle. This keynote presentation will address recent applications of intelligent software systems along with the future potential growth of AI. Fundamental concepts and limitations connected with the reliability of intelligent software systems will be introduced. Several issues of software validation techniques involved in maximizing the reliability of a large and complex intelligent software system will be explored. The discussion and presentation may help identify new and innovative research directions for further study and development in computational intelligence with wide range of emerging applications along with the challenges and opportunities.

Binary generators are devices responsible for delivering random or pseudo-random binary sequences, very useful in different branches such as cryptography, simulation, mathematics. These random chains must have high periods and high linear complexity and must also pass statistical tests to ensure that they are effectively random. The generator design process must take the above into account and must be verified at each stage for the final result to be successful. Sometimes the simple combination of poorly designed cryptographic components can lead to a faulty generator. This work shows the steps to follow for the successful development of a good generator. For the generator proposed in this presentation, eight nonlinear feedback shift registers (NLFSR) were used, which are cryptographic components with nonlinear Boolean filtering functions, which were combined by using five multiplexers. Finally, with the device obtained and operating with different keys, pseudorandom binary sequences were obtained that passed the randomness tests to which they were subjected.

Federated learning (FL) ensures data privacy and security, which has been emerging as a promising approach for developing an accurate and efficient model for COVID-19 detection. This paper highlights the FL-assisted systematic review of the current state-of-the-art COVID-19 detection methods. This work conducts a comprehensive study of the relevant literature and identifies relevant articles. The included articles described various FL-based COVID-19 detection models that utilized different types of data and machine learning algorithms. Furthermore, these models have several advantages over traditional centralized models, including privacy protection, data security, and scalability. The findings of this systematic review suggest that FL-based COVID-19 detection models have great potential for accurate and efficient COVID-19 detection while protecting data privacy and security. FL requires further research to validate the findings on different metrics and enhance the performance of the models.