With the promotion of global educational equity and quality education under the United Nations Sustainable Development Goals (SDGs), technology-assisted instruction has become a critical pillar in engineering education. This study aims to apply computer vision and deep learning technologies to electronics circuit lab teaching, enhancing learners' efficiency and addressing fairness issues in experimental assessment. Traditional assessment methods for electronics experiments focus on experiment success or failure, which may fail to accurately reflect the efforts of some students. To address this, the study developed an intelligent experimental assessment system based on computer vision, capable of accurately identifying the number and position of components amidst environmental noise. Combined with deep learning for experiment classification, the system provides dynamic scoring and subsequent operational suggestions based on the recognition results, thereby improving learners' understanding and interest in electronics experiments. The system leverages image processing to filter external noise and accurately extract key experimental features. To enhance adaptability and versatility, the system supports automated recognition and scoring across various experimental scenarios. Experimental results demonstrate that this system not only reduces the workload of educators but also significantly increases learner engagement and learning outcomes. The study contributes to integrating technological innovation into engineering education, achieving the SDGs' vision of educational equity and quality education.

The evolution of human-robot interaction (HRI) demands robust evaluative methods that capture not only technical performance but also the intricate social dimensions of robotic systems. This workshop examines the application of the ROSaS questionnaire as a systematic instrument for assessing social aspects and uncanny valley perceptions across diverse user profiles in different stages of the design. Drawing on empirical studies, the session explains how the ROSaS tool elucidates users' perceptions of robot sociability and human-likeness, and how these perceptions enrich the understanding of formative affinity in the design of social robots. By providing participants with practical experience on prototyping social robots, this workshop aims to elucidate the determinants of robot warmth. Through scholarly presentations and in-depth discussions, participants will engage in a collaborative analysis of current methodologies and challenges, with the ultimate objective of refining evaluation practices that guide the design of human-robot interfaces in social robots. This workshop welcomes participants of all expertise levels.

Education in science, technology, engineering, and mathematics (STEM) is essential to achieving continued technological advancement. The most critical years for instilling knowledge are during childhood, and a strategic way to accomplish this is through playful materials. Therefore, there is a need to develop more inclusive solutions to achieve the good inclusion of visually impaired (VI) learners in this learning area. Despite their importance, specific design guidelines are scarce for developing playful, educational solutions for VI learners in STEM. Qualitative research was conducted through interviews and observations of the interactions between VI Learners playing an audio game and tactile 3D printed blocks, which covered an age range of young participants aged 8–18 years and adults aged 30–40 years in Taiwan. Surprisingly, the results showed that the combination of tangible and audio elements for playful purposes opens the way for students to show interest during educational interactions and, at the same time, allows them to understand the concept, especially when presented in different game missions but repeating the same principle/concept. In conclusion, there is a need for more inclusive strategies and approaches for playful STEM tools for VI learners, and one important aspect of achieving this is design guidelines. This study aims to understand the educational context of VI learners and their interactions when playing with educational materials to learn STEM concepts and develop design guidelines for the future development of playful STEM educational games.

The proliferation of Internet of Things (IoT) applications prompts extraordinary demands for the collaboration of large amounts of computational resources provided by IoT devices in edge networks, and these applications are mostly delay-sensitive. Generally, these resources are encapsulated as IoT services. Thereafter, IoT applications can be performed, such that the collaboration of their sub-tasks is achieved through the composition of functionally complementary and geographically contiguous IoT services. The status of computational resources in IoT devices may change continuously along with their occupancy and release by IoT services. Considering the resource-scarceness of IoT devices, when the workload of IoT devices increases due to more services to be processed, certain IoT devices may hardly have enough remaining resources to co-host more instances of certain IoT services prescribed by forthcoming IoT applications with strict constraints. As a result, the delay satisfaction of both on-running and forthcoming IoT applications may be negatively impacted, or even hardly be satisfied any longer. To solve this issue, this paper proposes a rEsource-Efficient service Configuration ($E^{2}$rC) mechanism, which aims to optimize the configuration of computational resources provided by IoT devices with respect to complex requirements prescribed by IoT applications, through service migration techniques. This service migration problem is formulated as markov multi-phases decisions, which is solved through our enhanced Deep Reinforcement Learning (DRL) approach with a two-layer Q-network. Extensive experiments have been conducted upon the dataset of our testbed system. Evaluation results show that our $E^{2}$rC is more efficient than the state-of-art counterparts in satisfying delay constraints of IoT applications, while reducing the energy consumption and improving the resource utilization efficiency of IoT devices.

With the pivotal role that dark mode plays in user interface design, its widespread adoption across various applications and operating systems is evident. This study aims to investigate the potential effects of different background modes (light and dark) using cognitive ability tests and collect demographic variables for analysis. A total of 173 participants from diverse geographic regions worldwide completed an online survey comprising cognitive tests. The experimental results demonstrate that cognitive scores were higher in light mode compared to dark mode. Additionally, younger adults performed significantly better than older adults in light mode, while participants with academic education scored higher than those without in dark mode. In both modes, men outperformed women. A majority of females prefer light mode, while a higher proportion of males feel comfortable with both modes. These findings address the gap in understanding the impact of dark mode, offering practical insights in inclusive design practices.

With the natural decline in health, older adults increasingly require various forms of assistance. As the global senior population continues to grow, the demand for professional caregivers rises correspondingly. Throughout history, assistive technologies such as canes, walkers, and hearing aids have played a crucial role in senior care. Among the extensive range of available assistive technologies, robots and virtual assistants are particularly notable for their wide array of applications for both assistants and caregivers. By integrating a social robot with a virtual assistant, an assistive solution can be developed where their functionalities complement each other. This study aimed to create an assistive solution focused on the care of older adults by leveraging the combined functionalities of social robots and virtual assistants. To validate the proposed assistive solution, an experiment was conducted in which potential users simulated care and schedule management scenarios to examine the interaction between older adults and the employed technology. The positive feedback from the experiment indicated that the assistive solution would be both useful and appreciated, not only for care-related activities but also for daily tasks.

Digital banking adoption in Small Island Developing States (SIDS) faces unique cultural and technological challenges, such as limited digital literacy, infrastructure deficits, and socioeconomic disparities. This paper introduces the Technological Influence on Digital Banking Adoption (TIDiBanD) Framework, a novel approach to contextualizing how backend technologies, data assurance practices, and frontend innovations interact with cultural barriers in digital banking adoption. The framework provides researchers with a structured method to examine the influence of technologies such as social robots, intelligent virtual assistants (IVAs), and generative AI tools in fostering digital banking acceptance. To demonstrate the utility of TIDiBanD, we conducted an empirical study evaluating the persuasiveness of social robots and IVAs in promoting digital banking services in a SIDS population. Results indicate that while IVAs were perceived as more competent and easier to interact with than social robots, social robots evoked stronger positive emotions. The study also finds that consumers in that state may prefer familiar technologies over novel ones for financial interactions, highlighting the need for culturally aligned digital banking strategies. These findings provide actionable insights for banks and policymakers aiming to enhance digital financial inclusion through persuasive technologies.

A service robot is an Internet of Things (IoT) or a cyberphysical system comprising a robotic body integrated with one or more Cloud-based services. This architecture enables sophisticated human-machine interaction and expands the functional capabilities of conventional robotic systems. A service robot can take the form of anthropomorphic, zoomorphic, or even theomorphic design. Our research team is collaborating with the Canadian National Institute for the Blind (CNIB) to develop a prototype service dog robot. Due to this, this paper presents our current experiences in designing, developing, and evaluating a Do-It-Yourself (DIY) articulated robotic tail toolkit for zoomorphic robots at Ontario Tech University. We detail the design process that led to the development of a robotic tail mechanism, driven by microservices, which emphasizes reliability and expressive while utilizing readily accessible components. We range also conducted experiments and evaluations of the robotic tail during a hands-on workshop with students from diverse backgrounds, using a services computing approach. We continue to develop an effective DIY educational tool that promotes student engagement in Human-Robot Interaction (HRI) design for future service computing education.

The increasing range of applications for home robots has heightened the need to address their security challenges. Evaluating the security of these robots requires considering not only potential vulnerabilities but also the diversity of their applications and architectures. This study highlights the limitations of prior research on home robot security and underscores the necessity for more comprehensive investigations. By incorporating insights from analogous systems such as Cyber-Physical Systems (CPS) and smart home technologies, we draw relevant comparisons to enhance understanding. Home robots, as integral components of smart home ecosystems, demand a focus on safety, privacy, and security across CPS architecture layers. Effective deployment requires tailored security frameworks guided by reference models and architectures. Constraints in computing power and energy necessitate cloud and fog computing for robust solutions. Additionally, mapping threats, vulnerabilities, and mechanisms while addressing AI-induced challenges, such as expanded attack surfaces and ethical concerns, is critical for optimal risk management.

This paper presents canine likeness features in robotic quadrupeds that influence their social perception. We adopted Contrastive Language-Image Pre-Training (CLIP), a neural network that has demonstrated signatures of the Uncanny Valley effect, to explore how the perception of quadrupeds evolves as their level of canine likeness intensifies. Seven models were tested, ranging from a fully robotic quadruped to a living dog with 252 images. Our findings indicate that the Uncanny Valley effect also develops in quadruped robots. This finding is a reference to selecting an appropriate level of realism for canine likeness fourlegged robots in Human-Robot Interaction (HRI).

Developing effective platforms for economic energy management is considered a pivotal issue in the field of Electric Vehicles (EVs). To implement a cost-effective Energy Management Platform (EMP), developers must overcome two major challenges. The first challenge lies in the environmental dynamic nature such as EV location, energy price fluctuations, storage levels, and parking availability at charging stations. This causes most traditional one-shot optimizations to fail. The second challenge pertains to the lack of regulation in EV energy exchanges. To address these challenges, we propose a cost-aware two-stage EMP based on blockchain and deep reinforcement learning (DRL), namely TEMP. Specifically, TEMP first develops a sharding-based blockchain energy management framework, which guarantees trust, security, privacy, traceability, and accountability without the need for intermediaries. Then, considering the complex and high-dimensional environment, TEMP devises a two-stage cooperative scheduling scheme by combining ant colony optimization (ACO) with proximal policy optimization (PPO) to enhance learning effectiveness. Evaluations show that TEMP outperforms the two state-of-the-art baselines by 12.3% and 4.4% in terms of long-term profits while reducing costs by 6.7% and 2.8%, respectively. Moreover, energy transaction efficiency can be ensured when the EV number of blockchain networks is gradually increased.

Given the technological progress, smart toys have become relevant in the toy market. Toy companies adopt different requirements and web services to create smart toy features in different shapes and purposes. Each company usually has its requirements and implementation process, including semantic information and risk management guidelines. In other words, there is no common knowledge base related to the smart toy domain, in which the organizations could share information and reuse standardized knowledge, mitigating interoperability issues. Our work aims to build a smart toy's privacy context ontology, bringing general concepts and privacy-related, machine-readable, offering organizations and software agents a common knowledge base related to privacy on smart toy's context to reuse for smart toys design and features implementation.

Autoscaling is critical for ensuring optimal performance and resource utilization in cloud applications with dynamic workloads. However, traditional autoscaling technologies are typically no longer applicable in microservice-based applications due to the diverse workload patterns and complex interactions between microservices. Specifically, the propagation of performance anomalies through interactions leads to a high number of abnormal microservices, making it difficult to identify the root performance bottlenecks (PBs) and formulate appropriate scaling strategies. In addition, to balance resource consumption and performance, the existing mainstream approaches based on online optimization algorithms require multiple iterations, leading to oscillation and elevating the likelihood of performance degradation. To tackle these issues, we propose PBScaler, a bottleneck-aware autoscaling framework designed to prevent performance degradation in a microservice-based application. The key insight of PBScaler is to locate the PBs. Thus, we propose TopoRank, a novel random walk algorithm based on the topological potential to reduce unnecessary scaling. By integrating TopoRank with an offline performance-aware optimization algorithm, PBScaler optimizes replica management without disrupting the online application. Comprehensive experiments demonstrate that PBScaler outperforms existing state-of-the-art approaches in mitigating performance issues while conserving resources efficiently.