The exponential growth of the Internet of Things (IoT) necessitates robust, scalable, and secure identity management solutions to handle the vast number of interconnected devices. Traditional centralized identity systems are increasingly inadequate due to their vulnerabilities, such as single points of failure, scalability issues, and limited user control over data. This study explores a decentralized identity management model leveraging the IOTA Tangle, a Directed Acyclic Graph (DAG)-based distributed ledger technology, to address these challenges. By integrating Decentralized Identifiers (DIDs), Verifiable Credentials (VCs), and IOTA-specific technologies like IOTA Identity, IOTA Streams, and IOTA Stronghold, we propose a proof-of-concept framework that enhances security, scalability, and privacy in IoT ecosystems. Our implementation on resource-constrained IoT devices demonstrates the feasibility of this approach, highlighting significant improvements in transaction efficiency, real-time data exchange, and cryptographic key management. Furthermore, this research aligns with Web 3.0 principles, emphasizing decentralization, user autonomy, and data sovereignty. The findings suggest that IOTA-based solutions can effectively advance secure and user-centric identity management in IoT, paving the way for broader applications in various domains, including smart cities and healthcare.

We perform a series of optimisations on the convolution operator within the ARM CMSIS-NN library to improve the performance of deep learning tasks on Arduino development boards equipped with ARM Cortex-M4 and M7 microcontrollers. To this end, we develop custom microkernels that efficiently handle the internal computations required by the convolution operator via the lowering approach and the direct method, and we design two techniques to avoid register spilling. We also take advantage of all the RAM on the Arduino boards by reusing it as a scratchpad for the convolution filters. The integration of these techniques into CMSIS-NN, when invoked by TensorFlow Lite for microcontrollers for quantised versions of VGG, SqueezeNet, ResNet, and MobileNet-like convolutional neural networks enhances the overall inference speed by a factor ranging from 1.13× to 1.50×.

This paper investigates a serverless edge-cloud architecture to support knowledge management processes within smart cities, which align with the goals of Society 5.0 to create human-centered, data-driven urban environments. The proposed architecture leverages cloud computing for scalability and on-demand resource provisioning, and edge computing for cost-efficiency and data processing closer to data sources, while also supporting serverless computing for simplified application development. Together, these technologies enhance the responsiveness and efficiency of smart city applications, such as traffic management, public safety, and infrastructure governance, by minimizing latency and improving data handling at scale. Experimental analysis demonstrates the benefits of deploying KM processes on this hybrid architecture, particularly in reducing data transmission times and alleviating network congestion, while at the same time providing options for cost-efficient computations. In addition to that, the study also identifies the characteristics, opportunities and limitations of the edge and cloud environment in terms of computation and network communication times. This architecture represents a flexible framework for advancing knowledge-driven services in smart cities, supporting further development of smart city applications in KM processes.

In this work, we assess the performance and energy efficiency of high-performance codes for the convolution operator, based on the direct, explicit/implicit lowering and Winograd algorithms used for deep learning (DL) inference on a series of ARM-based processor architectures. Specifically, we evaluate the NVIDIA Denver2 and Carmel processors, as well as the ARM Cortex-A57 and Cortex-A78AE CPUs as part of a recent set of NVIDIA Jetson platforms. The performance–energy evaluation is carried out using the ResNet-50 v1.5 convolutional neural network (CNN) on varying configurations of convolution algorithms, number of threads/cores, and operating frequencies on the tested processor cores. The results demonstrate that the best throughput is obtained on all platforms with the Winograd convolution operator running on all the cores at their highest frequency. However, if the goal is to reduce the energy footprint, there is no rule of thumb for the optimal configuration.

Purpose: This study investigates how federated learning (FL) and human–robot collaboration (HRC) can be used to manage diverse industrial environments effectively. We aim to demonstrate how these technologies not only improve cooperation between humans and robots but also significantly enhance productivity and innovation within industrial settings. Our research proposes a new framework that integrates these advancements, paving the way for smarter and more efficient factories.Design/methodology/approach: This paper looks into the difficulties of handling diverse industrial setups and explores how combining FL and HRC in the mark of Industry 5.0 paradigm could help. A literature review is conducted to explore the theoretical insights, methods and applications of these technologies that justify our proposal. Based on this, a conceptual framework is proposed that integrates these technologies to manage heterogeneous industrial environments.Findings: The findings drawn from the literature review performed, demonstrate that personalized FL can empower robots to evolve into intelligent collaborators capable of seamlessly aligning their actions and responses with the intricacies of factory environments and the preferences of human workers. This enhanced adaptability results in more efficient, harmonious and context-sensitive collaborations, ultimately enhancing productivity and adaptability in industrial operations.Originality/value: This research underscores the innovative potential of personalized FL in reshaping the HRC landscape for manage heterogeneous industrial environments, marking a transformative shift from traditional automation to intelligent collaboration. It lays the foundation for a future where human–robot interactions are not only more efficient but also more harmonious and contextually aware, offering significant value to the industrial sector.