Effective human-robot collaboration in disaster response is often hindered by the high cognitive demands placed on operators, who must process large volumes of multimodal data under time-critical conditions. We present CADRI, an adaptive multimedia interface that dynamically adjusts the modality and priority of information delivery based on real-time estimates of operator cognitive load. The system integrates visual dashboards, mixed reality overlays, haptic feedback, and auditory cues, enabling context-driven adaptation that reduces information overload and enhances situational awareness. A semantic debriefing module further filters and prioritizes mission-critical messages using caption generation and sentiment-based relevance scoring. We evaluate CADRI in simulated disaster scenarios involving victim search, hazard detection, and equipment inspection. Results show that, compared to a non-adaptive baseline, CADRI improves task completion speed, reduces perceived workload, and enhances interface usability, highlighting its potential for improving decision-making efficiency in high-risk environments.

Fire accidents have severe consequences in terms of life loss, injuries and damages to the property. In hazardous areas, firefighters risk their life to safeguard others. In search of minimizing these risks has contributed to the development of small firefighting robots that can respond quickly. This work aims to develop a state-of-the-art firefighting robot where the robot is designed to help firefighters and save lives. In the paper, the firefighting robot is designed with the combination of hardware devices and deep learning algorithms. The system has different hardware devices like Raspberry Pi, Arduino Uno, Pi camera, 3 flame sensors, servo motor, motor driver, DC motors, water pump etc. The three flame sensor cover almost 180 ^∘ area and each flame sensor can detect fire within 30 cm. The system moves to three different directions, like left, right and front, based on which flame sensor detected the fire. After detecting fire by the flame sensor, the Pi camera captures images and send to the YOLOv8 model for identifying the exact location of the fire in the image. The system calculates the distance and angle of fire from the water pipe using Projectile Motion formulas. Finally, the servo motor moves according to the calculated angle to point the water pipe in the exact position of the fire and water pump activates to give water for putting down the fire. The constructed YOLOv8 model for the system is trained with 6520 online images collected from Roboflow Universe and tested with 371 real time images captured by the Pi camera of the proposed system. The advanced firefighting robot achieves accuracy of 76.4

Automatic scene generation is an essential area of research with applications in robotics, recreation, visual representation, training and simulation, education, and more. This survey provides a comprehensive review of the current state-of-the-arts in automatic scene generation, focusing on techniques that leverage machine learning, deep learning, embedded systems, and natural language processing (NLP). We categorize the models into four main types: Variational Autoencoders (VAEs), Generative Adversarial Networks (GANs), Transformers, and Diffusion Models. Each category is explored in detail, discussing various sub-models and their contributions to the field. We also review the most commonly used datasets, such as COCO-Stuff, Visual Genome, and MS-COCO, which are critical for training and evaluating these models. Methodologies for scene generation are examined, including image-to-3D conversion, text-to-3D generation, UI/layout design, graph-based methods, and interactive scene generation. Evaluation metrics such as Frechet Inception Distance (FID), Kullback-Leibler (KL) Divergence, Inception Score (IS), Intersection over Union (IoU), and Mean Average Precision (mAP) are discussed in the context of their use in assessing model performance. The survey identifies key challenges and limitations in the field, such as maintaining realism, handling complex scenes with multiple objects, and ensuring consistency in object relationships and spatial arrangements. By summarizing recent advances and pinpointing areas for improvement, this survey aims to provide a valuable resource for researchers and practitioners working on automatic scene generation.

Wearable haptic systems integrated with mixed reality (MR) headsets have the potential to enrich remote and virtual interactions by providing tactile cues that align with visual feedback. This paper presents the design and hardware-in-the-loop (HIL) evaluation of a dual-glove MR-haptic framework integrated with Microsoft HoloLens 2 to enhance touch perception in immersive environments. The system consists of a Master Glove that captures tactile and motion data using force-sensitive resistors (FSRs), flex sensors, and an inertial measurement unit (IMU), and a Follower Glove that reproduces corresponding vibrotactile feedback through fingertip-mounted eccentric rotating mass (ERM) actuators. Sensor data from the Master Glove is transmitted via Bluetooth and Wi-Fi to a Mixed Reality Control Unit (MRCU)—implemented in Python and Unity—with a WebSocket bridge enabling low-latency, bidirectional communication. Within the MR environment, a spatially aligned virtual hand interacts naturally with virtual objects, and contact events trigger proportional vibrotactile feedback on the Follower Glove. A hardware-in-the-loop testing framework comprising programmable force application and automated signal logging quantifies latency, mapping fidelity, and actuator repeatability without requiring human participants. Experimental results demonstrate approximately 50 ms end-to-end latency, high correlation between sensed and rendered forces, and consistent haptic response across multiple trials. The results validate the proposed system as a robust and scalable platform for MR-based haptic interaction, suitable for future user-validated applications in training, teleoperation, and collaborative tasks.

Collaborative robotics research shows multi-robot systems’ utility in various applications. This study proposes a novel approach to knowledge sharing and learning across heterogeneous robots. The suggested architecture allows robotic agents to easily share vision, control, and decision-making algorithms. We solved object detection, global position estimation, and reconstruction using a stage-wise modular method, demonstrating good performance even in changing situations. The system’s initial findings show that RealSense technology’s rigorous calibration allows for minimum global position estimate errors. The PCA-based orientation module was highly accurate. The paper examines the bounding box estimation module’s complexity and Z-axis orientation flexibility. The reconstruction module needs improvement due to intrinsic conversion problems between pixel data and 3D points, however the system estimates height and width accurately. This research improves adaptability and interoperability in complex multi-agent circumstances by fostering collaborative intelligence in different robot teams.

Disease detection using medical images enables early and precise diagnosis. Despite the growing success of deep learning models, accurate classification remains a significant challenge. Medical images often exhibit characteristics such as limited spatial resolution, subtle visual differences between disease categories (low inter-class variance), and substantial variation within the same class (high intra-class variability). These factors collectively hinder the ability of standard vision models to generalize effectively, frequently resulting in misclassification. These challenges highlight the need for efficient architectures that can focus on critical spatial and contextual features for reliable performance. To mitigate these challenges we propose MedNet, a lightweight CNN architecture that combines depthwise separable convolutions with the CBAM attention mechanism to efficiently extract and refine spatially and contextually relevant features. The core ResidualDSCBAMBlock captures local patterns while CBAM enhances important spatial and channel-wise information, followed by adaptive pooling, dropout, and fully connected layers for robust classification. The model is trained and validated on the DermaMNIST, BloodMNIST, OCTMNIST which from MedMNIST, and Fitzpatrick17k datasets. MedNet matches or exceeds CNN baselines across these medical image datasets achieving higher accuracy with significantly fewer parameters and lower computational cost, demonstrating its effectiveness and efficiency in medical image classification tasks. Code is available at https://github.com/Md-Ferdous/MedNet .

Teeth-based human identification is particularly valuable in forensic investigations due to the durability of dental structures under challenging conditions. This study investigates the use of Oriented Bounding Boxes (OBBs) for the geometric representation of tooth structures. Unlike Horizontal Bounding Boxes (HBBs), OBBs accurately localize rotated or irregularly shaped objects. OBBs align with the natural orientation of the teeth, providing a more accurate geometric representation of the individual dental structures, since each tooth has a unique orientation. Unlike existing works, we propose a matching approach that uses the Structural Similarity Index Measure (SSIM) for scale correction and Rotated Intersection over Union (RIoU) for feature matching, ensuring scale-aware identification. Furthermore, we account for the variability in the interspace between the upper and lower jaw at different time periods, we apply a statistical mean to determine the shift vector, thereby making our identification method shift invariant. The experimental results confirm the effectiveness of our strategy, with the highest precision and recall observed at IoU thresholds ranging from 0.5 to 0.9. These findings highlight the potential of OBB-based detection in forensic and biometric applications. The proposed method can be applied to disaster victim identification, criminal investigations, and scenarios where traditional biometric data (e.g., fingerprints or facial recognition) are unavailable.

In recent times, aerial robots have gained prominence in critical applications such as natural disaster response and social security. However, existing control interfaces like joysticks and touch screens often pose significant challenges for untrained users, particularly in complex tasks and remote operations. This study introduces an innovative, novice-friendly drone control system, leveraging the familiar framework of a conventional car steering system combined with a VR headset for visual feedback. This design aims to provide an immersive, carlike experience to users unfamiliar with UAV operation. The core objective is to bridge the knowledge gap for inexperienced users, enabling them to effectively control drones in critical scenarios. To evaluate the system’s efficiency and user experience enhancement, we devised an indoor search-and-rescue experiment simulating an earthquake disaster scenario. The study includes a comparative analysis between traditional joystick control and our car steering-based system. Additionally, we will conduct a comprehensive user study to assess the improvement in user experience offered by our interface. This research contributes to making drone technology more accessible and effective for general users, particularly in high-stakes environments.

Mental health challenges are on the rise in our modern society, and the imperative to address mental disorders, especially regarding anxiety, depression, and suicidal thoughts, underscores the need for effective interventions. This paper delves into the application of recent advancements in pretrained contextualized language models to introduce MindGuide, an innovative chatbot serving as a mental health assistant for individuals seeking guidance and support in these critical areas. MindGuide leverages the capabilities of LangChain and its ChatModels, specifically ChatOpenAI, as the bedrock of its reasoning engine. The system incorporates key features such as LangChain's ChatPrompt Template, HumanMessage Prompt Template, ConversationBufferMemory, and LLMChain, creating an advanced solution for early detection and comprehensive support within the field of mental health. Additionally, the paper discusses the implementation of Streamlit to enhance the user experience and interaction with the chatbot. This novel approach holds great promise for proactive mental health intervention and assistance.

As our world increasingly faces various disastrous events, there is an urgent need for improved methods of disaster response, particularly for the detection of victims trapped in the debris. Deep learning presents a promising avenue for this task. However, the applicability of advanced detection models trained on popular datasets like COCO is limited, given their focus on objects in unobstructed, everyday conditions. In stark contrast, victims in disaster scenarios are often buried or partially hidden by rubble, presenting a unique challenge for detection models. Additionally, the collection of real-world disaster victim images for training is a daunting and ethically complex task. This study seeks to address these challenges by proposing an all-in-one solution for generating realistic, synthetic images of disaster victims using a framework based on the Poisson equation. This harmonious composite image generation provides a versatile and accessible means of training deep learning models for victim detection, circumventing the difficulties associated with gathering natural disaster images. We leverage the YOLO architecture for training and testing our model, applying it to the synthesized harmonious images. Our approach seeks to balance the need for effective, reliable victim detection in robot-assisted search and rescue missions with the practical and ethical constraints of model training. By generating synthetic yet realistic representations of disaster scenarios, we aim to create a solution that generalizes well and thus can potentially enhance the efficacy of disaster response efforts. We have achieved the best Average Precision (AP) with 50% Intersection over Union (IoU) called ( AP 50 ) using YOLOv5x. The ( AP 50 ) for YOLOv5x after just training was 29.3% but it reached 92.4% after harmonization.

This paper introduces a novel AI and Robotics education model featuring cascade mentoring and project-based learning. It empowers junior members to become active mentors, fostering reciprocal mentoring experiences and enhancing their understanding of computer science, AI, and Robotics. We establish a collaborative pedagogy, connecting K-12 students with a university research lab using standards-based curricula. Our research demonstrates the positive impact of cascade mentoring on both mentor students and mentees in computer science and physical science education, improving their attitudes toward these subjects. Moreover, we explore the potential of educational robotics to address societal concerns such as green energy and sustainability.

The escalating environmental challenges faced by our world today make it imperative to instill eco-friendly habits and environmental awareness in young children, who will be at the forefront of addressing these issues in the future. In this paper, we present the “RoboRecycle Buddy”, a voice and chat GPT-integrated social robot designed to enhance early childhood green education and foster positive recycling habits through playful interaction. By incorporating engaging, voice feedback-based educational content, the RoboRecycle Buddy aims to make learning about recycling enjoyable, accessible, and relevant to children, while simultaneously sparking their interest in the rapidly growing field of robotics. We discuss the design, development, and implementation of the RoboRecycle Buddy, highlighting its innovative features such as object recognition, face recognition, and natural voice interaction which encourages children to practice responsible waste disposal. The robot leverages advanced voice recognition technology and the GPT language model, enabling it to engage in contextually relevant conversations with children, both through voice and text, further enhancing their learning experience. The robot is equipped with an image recognition module based on a Convolutional Neural Network, enabling it to detect and classify waste materials in a manner similar to a child, i.e., by simply looking at them. We emphasize the importance of providing a supportive learning environment that encourages children to explore, question, and develop a deeper understanding of the recycling process. The paper concludes with a discussion on future research and development directions, including potential improvements and adaptations to the RoboRecycle Buddy, as well as broader implications for the field of Child-Robot Interaction.

Smart connected devices, called Smart IoTs are becoming a heavily adopted product around the world. They have been integrated into their service platform and are spreading rapidly in households as common products. For example, people use them to control lights, get weather updates, play music, set a timer and do a plethora of tasks through various human interfaces in their platforms. Especially, voiced controlled interfaces and virtual assistant technology, such as Google Assistant and Amazon Alexa are quickly embraced in major companies due to user convenience. In addition, they allow third party developers to build Smart IoTs and applications for adopting these voice controlled interfaces in their platform so that they can be controlled by voice. While the technology has created value by helping people perform certain household chores through voice, it is also imperative to acknowledge the security vulnerabilities that lie within these voice controlled devices. In this paper, we demonstrate the exploitation of security and privacy issues in these smart IoTs which are integrated with virtual assistant, especially Amazon Alexa in a custom endpoint and provide possible solutions to the platform provider as redesigning human-IoT interaction processes.

The advancements in immersive technologies allow us to create more sophisticated environments designed to help engage users by merging the physical world with a digital or simulated reality. These can range from completely immersive virtual environments to mixed reality immersive environments, where the virtual world and the real world collide. Virtual reality is a completely immersive environment where the users’ reality is replaced with a simulated environment, and the hardware works to convince the user that they are in a different world. In contrast, augmented reality is a mixed type of reality, combining both the virtual and the natural world by augmenting the real world with digital assets and components. While both types of experiences contribute to creating rich collaborative environments, a limitation, and sometimes inconvenience, is present with the requirement of wearing a head-mounted device (HMD), creating restriction that prevents users from having physical interactions with others. Rather than interacting in the virtual space, we propose a concept that provides the structure for a physical space where users can interact with the shared mixed reality environment, an environment projected to help create the collaborative aspect in this project without any wearable devices. This paper will present the developed system and implemented four-dimensional interactions and demonstrate the feasibility of the structured experience we have created.

Recent advancements in robotic technologies have boosted new robot applications to perform many tasks that used to be limited to humans. Given this trend towards the ubiquity of robotics in day-to-day life, demand for professionals with expertise in robot development and maintenance is highly expected to increase. It is essential that such talents be fostered in future generations at an early age in order to meet this demand. However, several challenges arise in trying to accomplish this mission. First of all, it might be difficult to hold children’s attention when teaching them a new or potentially challenging subject, especially when the subject is not attractive for them. In addition, lack of mathematical understanding, which can make it challenging to comprehend key principles, is another problem with robotics teaching at young ages. In this paper, we propose a highly-accessible drone design and an accompanying education strategy, which, together, can alleviate the aforementioned problems. The major goals of the developed drone are to remain affordable as well as to enhance students’ attention and motivation while maintaining a high level of functionality and safety. In the proposed educational strategy, students will be divided into four different levels; each level will have its own procedure to enable educators to integrate robotics into existing curricula. At each level, moreover, students will learn increasingly more complex robotics subjects by interacting with the drone through a user-friendly visual coding interface.

Robots may face difficulty detecting all the required objects with their current vision in a search and rescue (SAR) operation in a disaster scenario. This can be due to a partial view of some objects from the whole disaster scenario, or the training model of the robots may not be perfect enough to recognize all the available objects. This research established a method for SAR robots to learn without human interaction by combining their own and other robots' knowledge. It assumes that all robots have a simple machine learning model that can detect objects from their whole perspective. Yolov8, trained with a custom dataset, has been used as the basic machine learning model in this study to detect the object. Two robots and one object are used for the proof of concept, with one robot seeing the object fully and the other partially. Partial view images were auto-labeled, and the fundamental machine learning model of the robot with partial view was retrained using auto-labeled data until a predetermined condition was met. Then the other robot in the environment received the final retrained model with partial view information. The suggested system was tested in numerous scenarios and usually works well.

. The advancements in immersive technologies allow us to create more sophisticated environments designed to help engage users by merging the physical world with a digital or simulated reality. These can range from completely immersive virtual environments to mixed reality immersive environments, where the virtual world and real world collide. Virtual reality is a completely immersive environment where the users’ reality is replaced with a simulated environment, and the hardware works to convince the user that they are in a different world. In contrast, augmented reality is a mixed type of reality, combining both virtual and the natural world by augmenting the real world with digital assets and components. While both types of experiences contribute to creating rich collaborative environments, a limitation, and sometimes inconvenience, is present with the requirement of wearing a head-mounted device (HMD), creating restriction that prevents users from having physical interactions with others. Rather than interacting in the virtual space, we propose a concept that provides the structure for a physical space where users can interact with the shared mixed reality environment; an environment projected to help create the collaborative aspect in this project, without any wearable devices. In this paper, we will present the developed system and implemented four dimensional interactions and demonstrate the feasibility of the structured experience we have created.

Simulation of emergency response scenarios and routine inspections are imperative means in ensuring the proper functioning and safety of power plants, oil refineries, iron works, and industrial units. By utilizing autonomous robots, moreover, the reliability and frequency of such inspections can be improved. With the exception of facilities located in hazardous areas, such as off-shore factories, where dispatching response teams might be impossible, accidents caused by human mistakes can be prevented by autonomous inspections and diagnosis of facilities (pumps, tanks, boilers, and so on). One of the primary obstacles in robot-assisted inspection operations is detecting various types of gauges, reading them, and taking appropriate action. This study describes a unique robot vision-based plant inspection system that may be used to enhance the frequency of routine checks and, in turn, minimize equipment faults and accidents (explosions or fires caused by gas leaks) caused by human mistakes or natural degradation. This suggested system can conduct facility inspections by detecting and reading a variety of gauges and issuing reports upon the detection of any anomalies. Furthermore, this system is capable of responding to unforeseen anomalous events that pose potential harm to human response teams, such as the direct manipulation of valves in the presence of a gas leak.

One of the main concerns of modern life would be the potential risk of irreversible ecological damage. However, due to the lack of focus on the subject within education, this type of risk will only get worse when being left unattended. This is where the robotics system, known as the “Smart Trashcan Brothers”, can provide better environmental consciousness with the current, younger generation attending primary school. This paper goes over the concepts that make up the Smart Trashcan Brothers system, as well with a functional evaluation to verify that the described parts of the robotics system function as intended. From there, a discussion of future works will be brought up with regards to further Child Human Interaction works.