EAI Endorsed Transactions on AI and Robotics

E-Nose design and structures from statistical analysis to application in robotic: a compressive review

2023-02-20T00:45:39+00:00

Since 1982, the olfactory system of creatures has piqued the interest of academics who seek to create a comparable system. Despite its mysterious nature, the first stage has been successfully completed with the development of the E-nose. Its extended applications have opened new doors for researchers, ranging from food quality testing to bomb detection and even, more recently, identifying those infected with the coronavirus. In this talk, we will review the structure and sensor behavior of the E-nose, as well as its applications, such as odour source localization and various applications in agriculture. The challenge of odour identification has prompted researchers to employ robots with sensors to investigate and locate odour sources. The present study aims to synthesize documented research and provide a fresh perspective on odour localization research efforts and tests conducted. The study highlights previous attempts to equip robots with sensors to explore the real indoor or outdoor environment. Initially, a review was conducted to investigate various aspects of the sector and the obstacles involved.

Review of Image Classification Algorithms Based on Graph Convolutional Networks

2023-06-17T03:22:22+00:00

In recent years, graph convolutional networks (GCNs) have gained widespread attention and applications in image classification tasks. While traditional convolutional neural networks (CNNs) usually represent images as a two-dimensional grid of pixels when processing image data, the classical model of graph neural networks (GNNs), GCNs, can effectively handle data with graph structure, such as social networks, recommender systems, and molecular structures. In this paper, we will introduce the problems that graph convolutional networks have had, such as over-smoothing, and the methods to solve them, and suggest some possible future directions.

The Role of Biometric in Banking: A Review

2023-08-03T18:01:34+00:00

Biometrics plays a pivotal role in enhancing security, ensuring accurate identification, and offering convenient solutions across diverse industries. Its uniqueness, reliability, and potential for future advancements establish it as a crucial and valuable field in today's digital landscape. Fingerprint authentication in ATMs presents primary advantages such as heightened security through distinctive identification and user convenience by eliminating the reliance on PINs or passwords. This research paper focuses on conducting a comprehensive review and comparative analysis of various approaches for fingerprint identification, aiming to contribute to the understanding of effective and efficient methods in the context of ATM authentication.

Hearing loss classification via AlexNet and Support Vector Machine

2023-03-07T10:31:53+00:00

This paper presents a new method for detecting hearing loss. Our approach is first to use AlexNet to extract the features. Then, we use the Support Vector Machine as a classifier to classify the images. 10-fold cross-validation results showed that the sensitivities of the healthy control group, the left-sided hearing loss group, and the right-sided hearing loss group in this method were 94.67%, 94.00%, and 95.17%, respectively, achieving a very good effect compared with other hearing loss detection methods. In conclusion, our method is effective for the identification of hearing loss.

Emergency Evacuation Based on Cellular Automata

2023-03-10T05:40:17+00:00

With the increasing number or scale of large-scale assembly activities, emergency evacuation has become increasingly important. In order to understand the evacuation behavior of people, this paper uses European distance, simulation and other methods to optimize the research and design of evacuation in emergency. The main contents of this paper are as follows: First, based on the principle and update rules of cellular automata, determine the three key factors of gender, age and emotional intensity, and the evacuation model based on cellular automata, and then use MATLAB software to simulate the change of people flow under different anxiety conditions, and output relevant data and visual images. Finally, the differences of the experimental results caused by the key factors are discussed. Secondly, study the evacuation rate change curve, as well as the key factors and their effects. Based on the average speed and average flow model of the system, simulate and output the corresponding visual images. Thirdly, in order to study and determine the best simulated evacuation route and the dynamic process of personnel evacuation in the case of different door widths, based on the evacuation route selection model with the shortest time to reach the two doors, the best pedestrian evacuation route is selected according to the shortest time rule. Fourth, in order to study the impact of reduced visibility in the hall on the whole process. In this paper, the regional discretization method is used to establish the perception range model under visibility and analyze the factor changes in the whole process.

Integrating Virtual Reality and Robotic Operation System (ROS) for AGV Navigation

2023-03-24T14:45:28+00:00

The use of AGVs (Automated Guided Vehicles) is rapidly expanding in various applications and industries, meeting the growing demand for automated material handling systems. However, AGV control and navigation remain a challenge. To address this issue, robotics simulators such as ROS (Robot Operating System) have become widely used, reducing the cost and time of checking robot performance. Furthermore, the integration of virtual reality technology into the robotics field has facilitated the study of various robot behaviors in realistic environments, replicating the robot’s real-life size and dimensions. In this study, the TurtleBot2i and RAZBOT AGV robot platforms were integrated into the 3D Unity environment and controlled using ROS. Using Unity as a simulator for the robot’s working environment offers several benefits, including high-quality graphics and a detailed examination of the robot’s behavior. The results of the study demonstrate the accurate simulation and control of the AGV platforms in both ROS and Unity environments.

The Potential Use of ChatGPT for Debugging and Bug Fixing

2023-04-24T15:25:31+00:00

ChatGPT is a cutting-edge language model that has been making waves in the field of natural language processing. However, its capabilities extend far beyond language-based applications. ChatGPT can also be used as a powerful tool for debugging software code. As software applications become increasingly complex, the need for efficient and accurate debugging tools has become more pressing. ChatGPT's ability to analyze and understand code makes it a promising solution to this challenge. Debugging is a critical part of the software development process. Bugs, or errors in code, can have serious consequences for the functionality and security of software applications. Identifying and fixing bugs can be a time-consuming and labor-intensive process, requiring the expertise of experienced developers. ChatGPT has the potential to streamline this process and make it more accessible to a wider range of developers, regardless of their experience level. In this article, we will explore the capabilities of ChatGPT as a debugging tool, the advantages and limitations of using it, and best practices for integrating it into the software development workflow.

Immersive Horizons: Exploring the Transformative Power of Virtual Reality Across Economic Sectors

2023-05-27T01:59:19+00:00

The scholarly discourse surrounding the manifold advantages, applications, and limitations of implementing Virtual Reality (VR) in the contemporary milieu has burgeoned over time. VR holds immense potential, attracting fervent interest from governmental and private entities alike. Nevertheless, the existing body of literature pertaining to the expanding utilization of VR in diverse economic sectors remains scant. Therefore, the primary objective of this study is to furnish a comprehensive literature review encompassing VR applications across various economic domains while elucidating concerns surrounding its integration within engineering education. A total of 108 publications were extracted from prominent databases such as Scopus, Elsevier, Science Direct, and Google Scholar, with a subsequent review of 51 relevant works. These scrutinized journals were published between 2015 and 2022 and were predominantly authored in English. The reviewed publications encompassed VR applications in education, robotics, healthcare, transportation, sports, agriculture, governance, security, and media. The study’s findings unveiled significant advancements in VR implementation within engineering education, medical training, cognitive augmentation, aircraft assembly, governance, and diverse other spheres. Notwithstanding these achievements, impediments to VR deployment were identified, stemming from financial exigencies, cultural and conventional norms, with scant evidence of VR’s prevalence in underdeveloped nations, given that all the assessed research originated from developed economies. Additionally, the limitations of this review encompassed a small sample size and a narrowly focused demographic in the examined articles. Nevertheless, despite these constraints, the research highlights substantial progress in VR utilization over the preceding decade.

Multi-class Classification of Imbalanced Intelligent Data using Deep Neural Network

2023-06-25T15:01:48+00:00

In recent years, studies in the field of deep learning have made significant progress. These studies have focused
more on datasets with balanced classification, and less research has been done on imbalanced datasets, which
are of great importance in the real world and present significant challenges for classification. This article
studies the problem of classifying imbalanced data, introduces dynamic sampling for deep neural networks,
investigates the imbalanced multiclass problem, and proposes a dynamic sampling method for deep learning.
In our proposed method, all samples are fed to the current deep neural network for each training iteration,
and the accuracy, precision, and mean error of the deep neural network are estimated. The proposed method
dynamically selects informative data for training the deep neural network. Comprehensive experiments were
conducted to evaluate and understand its strengths and weaknesses. The results of 13 imbalanced multiclass
datasets show that the proposed method outperforms other methods, such as initial sampling techniques,
active learning, cost-sensitive learning, and reinforcement learning.

Efficiently Guiding K-Robots Along Pathways with Minimal Turns

2023-06-27T09:26:21+00:00

INTRODUCTION: This paper addresses the navigation of a team of k protector robots within pathways, focusing on minimizing the total number of turns. These robots utilize orthogonal routes known as watchman routes, which prioritize finding the shortest path while maintaining visibility of all points in the environment from at least one robot on its designated route. The main objective of this research is to optimize robot navigation by reducing the overall number of turns.
OBJECTIVES: The primary objective of this study is to develop a linear-time algorithm that efficiently processes and determines routes for k robots within a specified area. By minimizing the number of turns, this algorithm aims to enhance the navigation capabilities of watchman robots, enabling them to effectively traverse complex environments.
METHODS: This research employs techniques derived from computational geometry to investigate the navigation of protector robots. The focus is on developing an algorithm that can efficiently process and determine the optimal routes for the robots, considering factors such as visibility and shortest path length. The algorithm is designed to minimize the number of turns while ensuring efficient coverage of the environment.
RESULTS: The main results of this paper include the development of a linear-time algorithm for determining routes for a team of k protector robots. The algorithm efficiently processes the input data and produces separate routes for each robot. By minimizing the number of turns, the algorithm improves the overall navigation efficiency of the robots. The results demonstrate the effectiveness of the algorithm in optimizing robot paths and reducing the complexity of navigation in real-world scenarios.
CONCLUSION: In conclusion, this research contributes to the field of robotic systems by addressing the navigation challenges faced by a team of protector robots. The introduced linear-time algorithm optimizes the routes for k robots, aiming to minimize the total number of turns. The outcomes of this study have significant implications for the advancement of watchman robots, enhancing their coverage and surveillance capabilities in real-world applications. The algorithm’s efficiency and effectiveness in minimizing turns open new opportunities for developing efficient navigation strategies in complex environments.

Breast cancer detection via wavelet energy and feed-forward neural network trained by genetic algorithm

2023-06-29T17:50:49+00:00

Enhancing the precision of breast cancer detection is the primary objective of this investigation, given its status as the most prevalent cancer among women worldwide. Timely identification of breast cancer can significantly improve the likelihood of successful diagnosis. To achieve this, we propose a innovative way that combines wavelet energy and a feedforward neural network. Our method employs the genetic algorithm and undergoes 20 iterations of 10-fold cross-validation for robustness. Via utilizing wavelet energy as a feature extractor and a feedforward neural network as the classifier, our method outperforms three alternative algorithms.

Imbalanced Multiclass Medical Data Classification based on Learning Automata and Neural Network

2023-07-04T04:52:56+00:00

Data classification in the real world is often faced with the challenge of data imbalance, where there is a
significant difference in the number of instances among different classes. Dealing with imbalanced data is
recognized as a challenging problem in data mining, as it involves identifying minority-class data with a
high number of errors. Therefore, the selection of unique and appropriate features for classifying data with
smaller classes poses a fundamental challenge in this research. Nowadays, due to the widespread presence
of imbalanced medical data in many real-world problems, the processing of such data has gained attention
from researchers. The objective of this research is to propose a method for classifying imbalanced medical
data. In this paper, the hypothyroidism dataset from the UCI repository is used. In the feature selection stage,
a support vector machine algorithm is used as a cost function, and the wrapper algorithm is employed as
a search strategy to achieve an optimal subset of features. The proposed method achieves high accuracy,
reaching 99.6% accuracy for data classification through the optimization of a neural network using learning
automata.

Concealing Robots in Environments: Enhancing Navigation and Privacy through Stealth Integration

2023-07-08T14:19:25+00:00

With the continuous advancement of robotics technology, the integration of robots into diverse human environments has become increasingly prevalent. However, the presence of robots in public spaces can often elicit discomfort or unease among individuals. To address this concern, the concept of concealing robots in various settings has emerged as an innovative approach to improve robot navigation and interaction while minimizing intrusion on human privacy. This paper explores the motivations, challenges, and potential benefits of hiding robots in different environments, particularly within the context of swarm robotics where multiple interconnected robots form a cohesive swarm. Equipped with onboard processing, communication, and sensing capabilities, these robots can autonomously interact with each other and adapt to the environment.
The paper investigates the problem of maximizing the number of hidden orthogonal swarm robots, considering scenarios in which robots need to navigate and operate within polygonal environments. Specifically, it presents a 4-approximation algorithm for computing a maximum hidden robot set in such environments. The algorithm offers a practical solution for determining an efficient arrangement of robots that minimizes their visibility while ensuring effective swarm operation.
By concealing robots in diverse environments, several benefits can be achieved. First, it helps to alleviate discomfort or unease among individuals, allowing for smoother integration of robots into public spaces. Additionally, concealing robots enhances their navigation capabilities by leveraging stealth techniques, allowing them to move seamlessly and unobtrusively within the environment. This approach also promotes improved human-robot interaction, as the reduced visibility of the robots can alleviate concerns and foster a more natural and comfortable interaction between humans and robots. The paper sheds light on the current state of the field, discussing the motivations behind concealing robots in different settings and highlighting the challenges that need to be addressed. Furthermore, it presents insights into future directions, including the development of more advanced stealth technologies, ethical frameworks for integrating hidden robots, and the potential impact on urban planning and infrastructure. In conclusion, hiding robots in diverse environments offers a promising approach to enhancing robot navigation, interaction, and privacy. The research presented in this paper contributes to the understanding of this emerging field and provides a foundation for further exploration and development of hiding strategies for swarm robotics in various settings.

Advancing Robot Perception in Non-Spiral Environments through Camera-based Image Processing

2023-07-19T20:01:21+00:00

Robot perception heavily relies on camera-based visual input for navigating and interacting in its environment. As robots become integral parts of various applications, the need to efficiently compute their visibility regions in complex environments has grown. The key challenge addressed in this paper is to devise an innovative solution that not only accurately computes the visibility region V of a robot operating in a polynomial environment but also optimizes memory utilization to ensure real-time performance and scalability.
The main objective of this research is to propose an algorithm that achieves optimal-time complexity and significantly reduces memory requirements for visibility region computation. By focusing on sub-linear memory utilization, we aim to enhance the robot's ability to perceive its surroundings effectively and efficiently.
Previous approaches have provided solutions for visibility region computation in non-spiral environments, but most were not tailored to memory limitations. In contrast, the proposed algorithm is designed to achieve optimal time complexity that is O(n) while reducing memory usage to O(c/log n) variables, where c < n represents the number of critical corners in the environment. Leveraging the constant-memory model and memory-constrained algorithm, we aim to strike a balance between computational efficiency and memory usage.
The algorithm's performance is rigorously evaluated through extensive simulations and practical experiments. The results demonstrate its linear-time complexity and substantial reduction in memory usage without compromising the accuracy of the visibility region computation. By efficiently handling memory constraints, the robot gains a cost-effective and reliable perception mechanism, making it well-suited for a wide range of real-world applications.
The constant-memory model and memory-constrained algorithm presented in this paper offer a significant advancement in robot perception capabilities. By optimizing the visibility region computation in polynomial environments, our approach contributes to the efficient operation of robots, enhancing their performance and applicability in complex real-world scenarios. The results of this research hold promising potential for future developments in robotics, computer vision, and related fields.

Security and Privacy in Fog/Cloud-based IoT Systems for AI and Robotics

2023-07-25T06:54:22+00:00

Integration of Internet of Things (IoT) systems based on the fog or the cloud with Artificial Intelligence (AI) and Robotics has prepared the way for breakthrough advancements in a variety of different fields of business. However, these cross-disciplinary technologies present significant difficulties in terms of maintaining confidentiality and safeguarding data. This article digs into the issues of establishing robust security and protecting user privacy in IoT systems that are based in the fog or the cloud and are utilized for AI and robotics applications. This study gives insights into the possible hazards encountered by such interconnected systems by conducting an in-depth review of existing security threats, vulnerabilities, and privacy concerns. In addition, the study investigates cutting-edge security mechanisms, encryption approaches, access control strategies, and privacy-preserving solutions that can be utilized to safeguard data, communications, and user identities. The results of this study highlight the demand for comprehensive security and privacy solutions to support the mainstream deployment of Fog/Cloud-based Internet of Things systems in the field of artificial intelligence and robotics.