Future wireless connectivity is envisioned to accommodate functionalities far beyond broadband data transmission, taking into account environmental parameters and usage scenario characteristics. In this landscape, the concept of integrated sensing and communication (ISAC) introduces a new beyond communications paradigm shift, where communication nodes can be employed to sense the shape of surrounding obstacles, map the environment, detect the presence of objects, focus power, localize users and track mobility. In this article, the roadmap for interactive, immersive and intelligent connectivity is presented with respect to different usage scenarios, involving mobility and environmental awareness. It is analyzed how the ISAC opportunity can be transformed into a powerful 6G enabler for an intelligent beyond-communications network capable of sensing the world. In this context, ISAC is envisioned to emerge from the synthesis of three visionary pillars: sense-to-communicate, communicate-to-sense, and multifunctional ISAC connectivity.

Open-radio access network (O-RAN) seeks to establish the principles of openness, programmability, automation, intelligence, and hardware-software disaggregation with interoperable and standard-compliant interfaces. It advocates for multi-vendorism and multi-stakeholderism within a cloudified and virtualized wireless infrastructure, aimed at enhancing the deployment, operation, and management of RAN architecture. These enhancements promise increased flexibility, performance optimization, service innovation, energy efficiency, and cost effectiveness across fifth-generation (5G), sixth-generation (6G), and beyond networks. A silent feature of O-RAN architecture is its support for network slicing, which entails interaction with other domains of the cellular network, notably the transport network (TN) and the core network (CN), to realize end-to-end (E2E) network slicing. The study of this feature requires exploring the stances and contributions of diverse standards development organizations (SDOs). In this context, we note that despite the ongoing industrial deployments and standardization efforts, the research and standardization communities have yet to comprehensively address network slicing in O-RAN. To address this gap, this paper provides a comprehensive exploration of network slicing in O-RAN through an in-depth review of specification documents from O-RAN Alliance and research papers from leading industry and academic institutions. The paper commences with an overview of the relevant standardization and open source contributions, subsequently delving into the latest O-RAN architecture with an emphasis on its slicing aspects. Furthermore, the paper explores O-RAN deployment scenarios, examining options for the deployment and orchestration of RAN and TN slice subnets. It also discusses the slicing of the underlying infrastructure and provides an overview of various use cases related to O-RAN slicing. Finally, it summarizes the potential research challenges identified throughout the study.

This paper serves as a foundation for researchers and practitioners to explore the key concepts, primary applications, and potential impacts of xApps in the open radio access network (O-RAN) architecture. We provide background information and examples related to xApps and discuss several fundamental concepts to familiarize readers with the key notions presented throughout the paper. In addition, we propose an architectural framework for the development and deployment of xApps within the near-real-time RAN intelligent controller (Near-RT RIC), offering a concise examination of its features, components, and interactions through internal and external open interfaces. Moreover, we explore three major aspects of the Near-RT RIC: the Near-RT RIC platform, the Near-RT RIC-related application programming interfaces (APIs), and the E2 service model (E2SM). Our primary objective is to provide a broader perspective on the theoretical foundation and practical implementation of xApps. Specifically, (a) we examine the Near-RT RIC platform to understand how its functionalities facilitate the deployment and monitoring of xApps, (b) investigate the APIs to comprehend the interconnections and data exchanges between xApps, the Near-RT RIC platform, and other components within O-RAN, and (c) analyze the E2SM, which governs the interactions and actions of xApps with E2 Nodes. Furthermore, we shed light on the management and orchestration of a collection of xApps within the Near-RT RIC, as well as present the lifecycle management of an individual xApp, detailing the phases from design to termination. Subsequently, we demonstrate the practical implementation of an open-source xApp (the key performance indicator [KPI] Monitoring xApp) in our O-RAN testbed. To accomplish this, we offer a detailed overview of the xApp, discuss the setup and architecture of our testbed (including the hardware and software components used), and provide a step-by-step guide to implementing this xApp in the testbed. Following that, we explain xApp-related conflict mitigation, addressing a range of topics, with a particular focus on the types of xApp-level conflicts and corresponding mitigation strategies. We then present several lessons learned from our research on xApps and their practical deployment. Finally, we identify several research and engineering challenges that need further effort to develop novel xApps and ensure their seamless deployment in O-RAN.

TThe evolution toward disaggregated, cloud-native, and multi-vendor architectures, as demonstrated by the open radio access network (O-RAN), together with the stringent performance requirements of sixth-generation (6G) networks, necessitates advanced transport network (TN) management and orchestration (M&O). Such capabilities are critical to ensure deterministic performance, flexible service provisioning, and scalable automation across fronthaul (FH), midhaul (MH), and backhaul (BH). In this context, the study of end-to-end (E2E) TN M&O requires a comprehensive examination of the contributions and perspectives of diverse standards development organizations (SDOs). Despite substantial standardization efforts, E2E TN M&O remains insufficiently addressed in the research literature. To bridge this gap, this paper presents a systematic and in-depth analysis of TN M&O. Specifically, we review and critically assess the contributions of major SDOs, with emphasis on TN M&O frameworks, control abstractions, and interface specifications. We then examine TN M&O in O-RAN, focusing on the role of the service management and orchestration (SMO) framework. Furthermore, we analyze its capabilities and current limitations in achieving a unified E2E orchestration across the RAN, TN, and core network (CN) domains. Moreover, we conduct a comparative assessment of major SDOs, open-source initiatives, and commercial solutions regarding their contributions to TN M&O. Building on this analysis, the paper proposes a unified architectural framework that explicitly integrates TN M&O within the SMO, leveraging key principles from a set of standard-compliant frameworks. The proposed architecture enables coordinated M&O of compute, storage, and TN resources. It supports multi-layer and multi-domain orchestration, as well as closed-loop service assurance, establishing a unified control plane for E2E network operation. Finally, this paper synthesizes the lessons learned from standards activities, open-source initiatives, commercial solutions, and architectural proposals. It identifies critical research challenges that must be addressed to realize fully integrated, scalable, and adaptive TN M&O in 6G. Together, these contributions establish a structured foundation for advancing TN M&O as a core enabler of 6G O-RAN.

Recent advancements in mobile and wireless networks are unlocking the full potential of robotic autonomy, enabling robots to take advantage of ultra-low latency, high data throughput, and ubiquitous connectivity. However, for robots to navigate and operate seamlessly, efficiently and reliably, they must have an accurate understanding of both their surrounding environment and the quality of radio signals. Achieving this in highly dynamic and ever-changing environments remains a challenging and largely unsolved problem. In this paper, we introduce MapViT, a two-stage Vision Transformer (ViT)-based framework inspired by the success of pre-train and fine-tune paradigm for Large Language Models (LLMs). MapViT is designed to predict both environmental changes and expected radio signal quality. We evaluate the framework using a set of representative Machine Learning (ML) models, analyzing their respective strengths and limitations across different scenarios. Experimental results demonstrate that the proposed two-stage pipeline enables real-time prediction, with the ViT-based implementation achieving a strong balance between accuracy and computational efficiency. This makes MapViT a promising solution for energy- and resource-constrained platforms such as mobile robots. Moreover, the geometry foundation model derived from the self-supervised pre-training stage improves data efficiency and transferability, enabling effective downstream predictions even with limited labeled data. Overall, this work lays the foundation for next-generation digital twin ecosystems, and it paves the way for a new class of ML foundation models driving multi-modal intelligence in future 6G-enabled systems.

The open radio access network architecture (O-RAN) architecture leverages intelligent near-real-time applications, known as xApps, to optimize network performance and services. However, these machine learning (ML)-driven xApps are vulnerable to adversarial attacks that can compromise their functionality and reliability. In this paper, we present a comprehensive study of adversarial threats targeting xApps and explore dynamic defense mechanisms to mitigate these risks. We begin by identifying potential attack vectors that target the near-real-time RAN intelligent controller (Near-RT RIC) and its associated xApps. We then utilize an open-source O-RAN testbed to deploy a key performance indicator (KPI) Monitoring xApp, a detection xApp, and a malicious adversarial xApp. The adversarial xApp carries out sophisticated inference-time attacks, including Carlini & Wagner (C&W) and basic iterative method (BIM), by perturbing key performance metrics in real-time to mislead the ML-based detection xApp. To counter these threats, we develop a defense xApp that integrates sequential anomaly detection techniques, ensemble deep neural network (DNN) inference, and gradient-based heuristics for real-time attack mitigation. Experimental results demonstrate that the C&W attack significantly degrades the baseline detection performance of the target xApp, reducing its accuracy from 92% to just 16% and BIM attack also achieves a comparable impact, lowering the detection accuracy to around 10%. Nevertheless, the proposed defense xApp promptly detects and neutralizes these adversarial manipulations, thereby restoring the effectiveness of the detector achieving up to 84% accuracy under the (C&W) attack and improving BIM (Basic Iterative Method) attack detection accuracy to 93% in the most challenging scenarios. This work presents a closed-loop evaluation of adversarial attacks and corresponding defenses within a real-world O-RAN environment that provides valuable insights into real-world vulnerabilities and mitigation strategies. By introducing a dynamic defense framework, we significantly enhance the security and resilience of ML-driven xApps and maintain reliable O-RAN performance even during attacks.

As next-generation mobile networks increasingly rely on virtualized infrastructure to deliver critical services, it is critical to ensure the efficiency of server farms while ensuring a reliable service. In fact, these server farms must meet stringent reliability guarantees to support time-sensitive applications in emerging 5G and beyond networks. In this paper, we tackle the optimal design of auto-scaling server farms, i.e., the selection of the best server type (and corresponding number of servers), taking into account both the service requirements and the operational and infrastructure costs. To achieve this, we develop an optimization algorithm that integrates (i) a model based on queueing theory to estimate the resources required to support the reliability constraints, and (ii) a general cost model including both capital and operational expenditures, which is minimized. We validate our approach through extensive simulations, comparing it against a benchmark based on classical queueing theory and exhaustive numerical searches. Results show that our method identifies the optimal server deployment in most cases while significantly reducing computation time, making it a practical solution for network planning.

Rapid and accurate localization of individuals during search-and-rescue (SAR) missions is essential for reducing casualties in emergencies. Traditional methods often struggle in disaster scenarios, where obstacles like debris or dense foliage hinder performance, and reliance on user-side actions proves impractical for mission-critical operations. This paper presents a novel approach that leverages a 5G-new radio (NR)-based unmanned aerial vehicle (UAV) localization framework, integrating hybrid techniques with adaptive clustering strategies to localize user equipments (UEs). Unlike traditional methods, our system dynamically adjusts its trajectory to balance latency and accuracy, achieving UEs positioning accuracy within tens of centimeters during simulation tests. By integrating 5G-NR technology into UAV-based localization, our approach provides a robust and scalable solution for mission-critical SAR operations, significantly enhancing the latency and reliability of locating individuals in emergency situations.

The deployment of O-RAN systems on general-purpose computing platforms represents a significant paradigm shift, promising remarkable performance improvements. However, these architectures may potentially increase both the capital and operational expenses of the network. The processor pooling concept is a promising solution to address this problem, consisting of a set of processing units (PUs) in the O-Cloud shared by several virtualized BSs (vBSs). Nevertheless, this strategy requires sophisticated resource assignment mechanisms to provide the expected gains in terms of cost and reliability. This paper proposes a novel online learning framework that assigns computing resources to vBSs in real-time (e.g., every TTI), thus handling the burstiness of real traffic loads. Our algorithm relies on online convex optimization (OCO) theory, extending state-of-the-art approaches in long-term fairness and constrained optimization and allowing discrete decisions. Our method offers an intrinsic closed-form iteration, speeding up the computation process and consequently allowing real-time operation. Moreover, our solution has guarantees in terms of fairness among the vBSs while adhering to long-term energy constraints over the entire operation horizon. We validate our theoretical findings via simulation and evaluate experimentally the algorithms in an O-RAN platform.

5G cellular systems are currently being deployed worldwide delivering the promised unprecedented levels of throughput and latency to hundreds of millions of users. At such scale and reach, security is crucial. Consequently, the 5G standard includes a new series of features to improve the security of its predecessors (i.e., 3G and 4G). In this work, we evaluate the security of currently deployed 5G commercial networks in Europe and North America. Specifically, by collecting 5G signaling traffic in the wild in several cities in Spain, Germany, France, Canada, and the USA, we i) fact-check which 5G security enhancements are implemented in current deployments, ii) provide a rich overview of the implementation status of each 5G security feature in a selection of 5G commercial networks in Europe and North America and compare it with previous results in China, iii) analyze the implications of optional features not being deployed, and iv) discuss on the still remaining 4G-inherited vulnerabilities. Our findings indicate that the rollout of 5G security features in the analyzed commercial networks is still a work in progress. On the one hand, several networks continue to rely on 4G for their core network operations, which hinders the deployment of new security features (e.g., SUCI) and, on the other hand, fully-fledged 5G deployments lack mandatory security measures such as GUTI reallocation after paging. Moreover, we find that some operators fail to provide proper temporary identifier randomization, in both 4G and 5G networks. Some of the obtained results are aligned with results previously reported from China [1] and keep the European and North American studied networks vulnerable to some 4G attacks, during their migration period from 4G to 5G. Conversely, studied networks deployed in North America exhibit stronger adherence to 5G security standards, with near-complete compliance observed, in contrast to deployments in China and Europe, where comparatively lower compliance levels have been observed.

The recent O-RAN specifications promote the evolution of RAN architecture by function disaggregation, adoption of open interfaces, and instantiation of a hierarchical closed-loop control architecture managed by RAN Intelligent Controllers (RICs) entities. This paves the road to novel data-driven network management approaches based on programmable logic. Aided by Artificial Intelligence (AI) and Machine Learning (ML), novel solutions targeting traditionally unsolved RAN management issues can be devised. Nevertheless, the adoption of such smart and autonomous systems is limited by the current inability of human operators to understand the decision process of such AI/ML solutions, affecting their trust in such novel tools. eXplainable AI (XAI) aims at solving this issue, enabling human users to better understand and effectively manage the emerging generation of artificially intelligent schemes, reducing the human-to-machine barrier. In this survey, we provide a summary of the XAI methods and metrics before studying their deployment over the O-RAN Alliance RAN architecture along with its main building blocks. We then present various use cases and discuss the automation of XAI pipelines for O-RAN as well as the underlying security aspects. We also review some projects/standards that tackle this area. Finally, we identify different challenges and research directions that may arise from the heavy adoption of AI/ML decision entities in this context, focusing on how XAI can help to interpret, understand, and improve trust in O-RAN operational networks.

Nanoscale devices with Terahertz (THz) communication capabilities are envisioned to be deployed within human bloodstreams. Such devices will enable fine-grained sensing-based applications for detecting early indications (i.e., biomarkers) of various health conditions, as well as actuation-based ones such as targeted drug delivery. Associating the locations of such events with the events themselves would provide an additional utility for precision diagnostics and treatment. This vision yielded a new class of in-body localization coined under the term "flow-guided nanoscale localization". Such localization can be piggybacked on THz communication for detecting body regions in which biological events were observed based on the duration of one circulation of a nanodevice in the bloodstream. From a decades-long research on objective benchmarking of "traditional" indoor localization, as well as its eventual standardization (e.g., ISO/IEC 18305:2016), we know that in early stages the reported performance results were often incomplete (e.g., targeting a subset of relevant performance metrics), carrying out benchmarking experiments in different evaluation environments and scenarios, and utilizing inconsistent performance indicators. To avoid such a "lock-in" in flow-guided localization, in this paper we propose a workflow for standardized performance evaluation of such localization. The workflow is implemented in the form of an open-source simulation framework that is able to jointly account for the mobility of the nanodevices, in-body THz communication between with on-body anchors, and energy-related and other technological constraints (e.g., pulse-based modulation) at the nanodevice level. Accounting for these constraints, the framework is able to generate the raw data that can be streamlined into different flow-guided localization solutions for generating standardized performance benchmarks.

Indoor localization systems are poised to revolutionize mobile applications, enabling precise navigation, immersive augmented reality, and context-aware pervasive computing. Emerging 5G/6G Joint Communication and Sensing (JCAS) architectures offer a promising pathway for achieving accurate and ubiquitous indoor positioning driven by rapid technological advancements and standardization efforts (e.g., 3GPP). This paper presents Echoes, a novel system within the Open RAN (O-RAN) architecture that leverages smart surfaces for precise user positioning. Echoes improves current smart surfaces sensing systems inferring the received signal phase solely from power measurements of reference signals used to measure the uplink channel. Echoes achieves this by exploiting an inherent property of smart surfaces: steering configurations are not singular, and introducing a constant phase offset to all elements achieves the same intended reflection direction but with a different reflected phase. Echoes combines the sensed direction of a user with the angle of arrival estimation to estimate the position of a user in a sensing area. Our prototype, evaluated in a real-world indoor setting, exhibits a mean absolute user localization error (MAE) of 0.9 meters. These results highlight the potential of Echoes to provide robust and accurate indoor positioning for next-generation mobile applications.

Reconfigurable Intelligent Surfaces (RISs) have emerged as a promising technology for next-generation wireless communications, offering energy-efficient control of electromagnetic (EM) waves. While conventional RIS models based on phase shifts and amplitude adjustments have been widely studied, they overlook complex EM phenomena such as mutual coupling, which are crucial for advanced wave manipulations. Recent efforts in EM-consistent modelling have provided more accurate representations of RIS behavior, highlighting challenges like structural scattering-an unwanted signal reflection that can lead to interference. In this paper, we analyze the impact of structural scattering in RIS architectures and compare traditional and EM-consistent models through full-wave simulations, thus providing practical insights on the realistic performance of current RIS designs. Our findings reveal the limitations of current modelling approaches in mitigating this issue, underscoring the need for new optimization strategies.

Miniaturized Uncrewed Aerial Vehicles (UAVs) can access indoor and hard-to-reach spaces, but severe constraints on payload and autonomy have limited their use in demanding tasks such as high-quality 3D reconstruction. We introduce a novel system architecture that enables autonomous, high-fidelity 3D scanning of static objects with sub-100 gram UAVs. Our core innovation lies in a closed-loop active viewpoint selection framework specifically tailored for ultra-constrained micro-platforms, advancing beyond standard static or offline active reconstruction methods. The framework establishes a dual-reconstruction pipeline that creates a real-time (RT) feedback loop between data capture and flight control. A near-RT process uses Structure-from-Motion (SfM) to generate an instantaneous point-cloud of the object. A systematic trajectory adaptation algorithm analyzes the model quality on the fly and dynamically adapts the UAV's trajectory based on parameterized spatial partitioning to intelligently capture new images of poorly covered areas, ensuring comprehensive acquisition. For the final, high-fidelity output, a non-RT pipeline employs a Neural Radiance Fields (NeRF)-based Neural 3D Reconstruction (N3DR) approach, fusing SfM-derived camera poses with precise external location data, evaluated across both radio-based Ultra Wideband (UWB) and visual motion-capture setups, to correct sensor noise and achieve superior accuracy. We implemented and validated this architecture using Crazyflie 2.1 UAVs. Our experiments, conducted in both single- and multi-UAV configurations show that algorithmic dynamic trajectory adaptation consistently improves reconstruction quality over static flight paths. This work demonstrates a scalable and autonomous solution that unlocks the potential of miniaturized UAVs for fine-grained 3D reconstruction, a capability previously reserved for much larger platforms.

The Open Radio Access Network (O-RAN)-compliant solutions often lack crucial details for implementing effective control loops at various time scales. To overcome this, we introduce MAREA, an O-RAN-compliant mathematical framework designed for the allocation of radio resources to multiple ultra-Reliable Low Latency Communication (uRLLC) services. In the near-real-time (RT) control loop, MAREA employs a novel Martingales-based model to determine the guaranteed radio resources for each uRLLC service. Unlike traditional queueing theory approaches, this model ensures that the probability of packet transmission delays exceeding a predefined threshold -- the violation probability -- remains below a target tolerance. Additionally, MAREA uses a real-time control loop to monitor transmission queues and dynamically adjust guaranteed radio resources in response to traffic anomalies. To the best of our knowledge, MAREA is the first O-RAN-compliant solution that leverages Martingales for both near-RT and RT control loops. Simulations demonstrate that MAREA significantly outperforms reference solutions, achieving an average violation probability that is x10 lower.

The maturity and commercial roll-out of 5G networks and its deployment for private networks makes 5G a key enabler for various vertical industries and applications, including robotics. Providing ultra-low latency, high data rates, and ubiquitous coverage and wireless connectivity, 5G fully unlocks the potential of robot autonomy and boosts emerging robotic applications, particularly in the domain of autonomous mobile robots. Ensuring seamless, efficient, and reliable navigation and operation of robots within a 5G network requires a clear understanding of the expected network quality in the deployment environment. However, obtaining real-time insights into network conditions, particularly in highly dynamic environments, presents a significant and practical challenge. In this paper, we present a novel framework for building a Network Digital Twin (NDT) using real-time data collected by robots. This framework provides a comprehensive solution for monitoring, controlling, and optimizing robotic operations in dynamic network environments. We develop a pipeline integrating robotic data into the NDT, demonstrating its evolution with real-world robotic traces. We evaluate its performances in radio-aware navigation use case, highlighting its potential to enhance energy efficiency and reliability for 5Genabled robotic operations.

5G mobile networks introduce a new dimension for connecting and operating mobile robots in outdoor environments, leveraging cloud-native and offloading features of 5G networks to enable fully flexible and collaborative cloud robot operations. However, the limited battery life of robots remains a significant obstacle to their effective adoption in real-world exploration scenarios. This paper explores, via field experiments, the potential energy-saving gains of OROS, a joint orchestration of 5G and Robot Operating System (ROS) that coordinates multiple 5G-connected robots both in terms of navigation and sensing, as well as optimizes their cloud-native service resource utilization while minimizing total resource and energy consumption on the robots based on real-time feedback. We designed, implemented and evaluated our proposed OROS in an experimental testbed composed of commercial off-the-shelf robots and a local 5G infrastructure deployed on a campus. The experimental results demonstrated that OROS significantly outperforms state-of-the-art approaches in terms of energy savings by offloading demanding computational tasks to the 5G edge infrastructure and dynamic energy management of on-board sensors (e.g., switching them off when they are not needed). This strategy achieves approximately similar to 15% energy savings on the robots, thereby extending battery life, which in turn allows for longer operating times and better resource utilization.

The virtualization of Radio Access Networks (vRAN) is rapidly becoming a reality, driven by the increasing need for flexible, scalable, and cost-effective mobile network solutions. To mitigate energy efficiency concerns in vRAN deployments, two approaches are gaining attention: ($i$i) sharing computing infrastructure among multiple virtualized base stations (vBSs); and ($ii$ii) relying upon general-purpose, low-cost CPUs. However, effectively realizing these approaches poses several challenges. In this paper, we first conduct a comprehensive experimental campaign on a vRAN platform to characterize the impact of computing and radio resource allocation on energy consumption and performance across various network contexts. This analysis reveals several key issues. First, determining the optimal allocation of computing resources is difficult because it depends on the context of each vBS (e.g., traffic load, channel quality) in a non-trivial and non-linear manner. Second, suboptimal resource assignment can lead to increased energy consumption or, even worse, degradation of users' Quality of Service. Third, the high dimensionality of the solution space hinders the effectiveness of traditional optimization or learning methods. To tackle these challenges, we propose AegisRAN, a framework for optimizing computing resource allocation in vRAN. AegisRAN addresses the dual objective of minimizing energy consumption while maintaining high system reliability. Moreover, when computing resources are overbooked, our solution ensures a fair resource partition based on vBS performance. AegisRAN leverages a discrete soft actor-critic algorithm combined with several techniques, including multi-step decision-making, action masking, digital twin-based training, and a tailored reward signal that mitigates feedback sparsity. Our evaluations demonstrate that AegisRAN achieves near-optimal performance and offers high flexibility across diverse network contexts and varying numbers of vBSs, with up to 25% improvement in energy savings compared to baseline solutions in medium-scale scenarios.

In the era of Industry 4.0, precise indoor localization is vital for automation and efficiency in smart factories. Reconfigurable Intelligent Surfaces (RIS) are emerging as key enablers in 6G networks for joint sensing and communication. However, RIS faces significant challenges in Non-Line-of-Sight (NLOS) and multipath propagation, particularly in localization scenarios, where detecting NLOS conditions is crucial for ensuring not only reliable results and increased connectivity but also smart factory personnel's safety. This study introduces an AI-assisted framework employing a Convolutional Neural Network (CNN) customized for accurate Line-of-Sight (LOS) and NLOS classification to enhance RIS-based localization using measured, synthetic, mixed-measured, and mixed-synthetic experimental data, that is, original, augmented, slightly noisy, and highly noisy data, respectively. Validated through such data from three different environments, the proposed customized-CNN (cCNN) model achieves 95.0%-99.0% accuracy, outperforming standard pre-trained models like Visual Geometry Group 16 (VGG-16) with an accuracy of 85.5%-88.0%. By addressing RIS limitations in NLOS scenarios, this framework offers scalable and high-precision localization solutions for 6G-enabled smart factories.