Keywords: Robotics, Architecture of Automation Systems, Simulation Based Automation Engineering
Abstract: Accurate odometry is fundamental for most mobile robot applications. In this paper, we propose a multimodal sensor fusion architecture for odometry estimation based on an Extended Kalman Filter (EKF), which integrates asynchronous data sources such as wheel odometry, IMU measurements, and LiDAR-based odometry estimates to exploit the strength of each sensor. To correct drift, we employ LiDAR odometry using an algorithm (Kinematic ICP) that refines estimates using point cloud alignment constrained by robot’s kinematics. Its output is fed back into the EKF, forming a closed-loop correction mechanism. We evaluate the system’s performance in a simulated environment with real-world conditions.

Keywords: Robotics, Simulation Based Automation Engineering, Advanced Engineering Methods for Automation Systems
Abstract: This research presents a novel nonlinear control strategy for 3-DOF quadrotor Unmanned Aerial Vehicles (UAVs), combining sliding mode control with a high-gain observer to ensure system stability. The proposed method enables accurate velocity estimation and robust disturbance rejection, including wind effects and model uncertainties. Applications such as traffic monitoring, wildfire detection, and infrastructure inspection highlight the growing role of UAVs in autonomous missions. Emphasis is placed on accurate physical modeling and the integration of miniaturized sensors and control boards. The approach is validated through experimental testing on a lab-scale UAV platform, offering insights into observer-based sliding mode control in underactuated systems.

Keywords: Robotics, Modularization and Flexibility of Production Systems, Information Modelling
Abstract: The automation of flexible production systems using robots is becoming increasingly important in modern manufacturing. This paper addresses the challenge of effectively managing and utilizing functionalities provided by robot modules in automated production systems. To tackle this, we propose the development of an information model, structuring robot module functionalities. By discussing and adapting existing methods, such as the Capability Skill Service (CSS) model, we aim to create a standardized and efficient way to structure and describe the functionalities of robot modules. This approach seeks to reduce engineering costs, improve system integration, and enhance operational efficiency. Ultimately, the development of this information model could lead to more advanced, simplified and flexible solutions for robotic automation.

Keywords: Robotics, Artificial Intelligence and Autonomous Systems, Advanced Engineering Methods for Automation Systems
Abstract: Automation on the construction site requires accurate real-time target pose estimation of robotic tools. This work focuses on a marker-free feedback pipeline guiding a screwing unit to its target pose for timber screwing tasks. The proposed method includes a YOLOv8 segmentation model trained via semi-automatic labeling, orientation estimation post-processing, confidence scoring and an adaptive Kalman filter to detect hole patterns on timber cassettes and estimate a target pose of the screwing unit. The system achieves 94.6% accuracy under ideal conditions, with real-time performance on an embedded Jetson AGX Orin.

Keywords: Advanced Engineering Methods for Automation Systems, Artificial Intelligence and Autonomous Systems, Robotics
Abstract: This paper deals with the stabilization of a folding quadrotor drone undergoing controlled geometric reconfiguration, a process that introduces non-stationary dynamics due to variable inertia, displaced Center of Gravity (CoG) and allocation matrix. The morphing mechanism, activated by articulated arms, couples rigid body dynamics with time-dependent parametric uncertainties. To handle these non-linearities, we develop a stable Lyapunov control law via the Interconnection Passivity and Damping Assignment (IDA-PBC) based control framework. The method enforces passivity by reshaping the energy of the system's Hamiltonian. A non-linear Newton-Euler model, validated by numerical simulations, demonstrates the robustness of the controller, despite inertie and CoG changes. The approach is extensible to systems where dynamic constraints evolve along with actuation, such as variable-geometry spacecraft or adaptive manipulators.

Keywords: Robotics
Abstract: This paper presents the development of a ROS-based autonomous navigation system for a tracked mobile robot by designing both the hardware and software parts of the robot. The project starts with the instrumentation and the design of the low-level control architecture for the robot. The high level architecture is based on the Robot Operating System (ROS) and it implements advanced strategies for the robot autonomous navigation including localization, mapping, path planning, obstacle avoidance and goal navigation algorithms. The system's overall performances are analyzed in two scenarios with the T600 tracked robot where its ability to navigate autonomously while overcoming obstacles is assessed. The results demonstrate the robot effective autonomous navigation in achieving the desired goal while avoiding static and dynamic obstacles.

Keywords: Reinforcement Learning, Neural Networks & Deep Learning, Machine Learning
Abstract: Continuous control Deep Reinforcement Learning (RL) approaches are known to suffer from estimation biases, leading to suboptimal policies. This paper introduces innovative methods in RL, focusing on addressing and exploiting estimation biases in Actor-Critic methods for continuous control tasks, using Deep Double Q-Learning. We design a Bias Exploiting (BE) mechanism to dynamically select the most advantageous estimation bias during training of the RL agent. Most State-of-the-art Deep RL algorithms can be equipped with the BE mechanism, without hindering performance or computational complexity. Our extensive experiments across various continuous control tasks demonstrate the effectiveness of our approaches. We show that RL algorithms equipped with this method can match or surpass their counterparts, particularly in environments where estimation biases significantly impact learning. The results underline the importance of bias exploitation in improving policy learning in RL.

Keywords: Architectures for Real-Time Intelligent-Control, Reinforcement Learning, Machine Learning
Abstract: An online Q-learning algorithm is introduced to solve the finite-horizon optimal tracking problem in discrete-time (DT) for linear systems with unknown system dynamics. The finite time-horizon results in time-varying Q-function parameters which are estimated online by approximating them as the product of constant parameters with a time-varying basis function. The reference trajectory is embedded in the tracking problem using a horizon of known future points. Simulation results are included to verify the theoretical claims.

Keywords: Reinforcement Learning, Neural Networks & Deep Learning, Architectures for Real-Time Intelligent-Control
Abstract: Deep Reinforcement Learning (Deep-RL) has emerged as a powerful paradigm for enabling autonomous agents to learn complex behaviors in dynamic environments. Despite its significant advancements and applications in robotics, Deep-RL faces substantial challenges when transitioning from simulation to real-world deployment, due to limited resource availability and the large amount of data required for training. To address these issues, this paper evaluates three state-of-the-art continuous control Deep-RL algorithms in the context of autonomous navigation tasks. A structured experimental methodology is used, progressing from high-fidelity simulations to real-world experiments. This work involves more than 120 hours of real-world experiments and shows evidence of a possible gap between on-paper performance and real-world performance of Deep-RL algorithms due to their different computational requirements and assumptions.

Keywords: Reinforcement Learning, Neural Networks & Deep Learning
Abstract: Resource allocation in complex job shop manufacturing environments presents a challenging and NP-hard optimization problem. This issue is increasingly being addressed with AI-based methods, particularly through Deep Reinforcement Learning (DRL) approaches. This work aims to develop a batching allocation strategy for furnace tools using the Proximal Policy Optimization (PPO) algorithm. The results demonstrate that the derived policy is more efficient in terms of batch efficiency and resource on-time delivery compared to the engineered heuristics commonly adopted in real fabs within the semiconductor manufacturing context.

Keywords: Neural Networks & Deep Learning, Reinforcement Learning
Abstract: This paper introduces MANGO (Multilayer Abstraction for Nested Generation of Options), a novel hierarchical reinforcement learning framework designed to address the challenges of long-term sparse reward environments. MANGO decomposes complex tasks into multiple layers of abstraction, where each layer defines an abstract state space and employs options to modularize trajectories into macro-actions. These options are nested across layers, allowing for efficient reuse of learned movements and improved sample efficiency. The framework introduces intra-layer policies that guide the agent's transitions within the abstract state space, and task actions that integrate task-specific components such as reward functions. Experiments conducted in procedurally-generated grid environments demonstrate substantial improvements in both sample efficiency and generalization capabilities compared to standard RL methods. MANGO also enhances interpretability by making the agent's decision-making process transparent across layers, which is particularly valuable in safety-critical and industrial applications. Future work will explore automated discovery of abstractions and abstract actions, adaptation to continuous or fuzzy environments, and more robust multi-layer training strategies.

Keywords: Reinforcement Learning, Evolutionary Algorithms for Control, Neural Networks & Deep Learning
Abstract: Deep Reinforcement Learning (RL) has emerged as a powerful method for addressing complex control problems, particularly those involving underactuated robotic systems. However, in some cases, policies may require refinement to achieve optimal performance and robustness aligned with specific task objectives. In this paper, we propose an approach for fine-tuning Deep RL policies using Evolutionary Strategies (ES) to enhance control performance for underactuated robots. Our method involves initially training an RL agent with Soft-Actor Critic (SAC) using a surrogate reward function designed to approximate complex specific scoring metrics. We subsequently refine this learned policy through a zero-order optimization step employing the Separable Natural Evolution Strategy (SNES), directly targeting the original score. Experimental evaluations conducted in the context of the 2nd AI Olympics with RealAIGym at IROS 2024 demonstrate that our evolutionary fine-tuning significantly improves agent performance while maintaining high robustness. The resulting controllers outperform established baselines, achieving competitive scores for the competition tasks.

Keywords: Internet of Things, Remote Sensor Data Acquisition, Intelligent Homes and Ambient Intelligence
Abstract: In this work, we propose a fully automated framework for classifying coffee capsules based on sensor data acquired during the brewing process. Unlike prior approaches that relied on hand-crafted features defined by domain experts, we demonstrate that end-to-end feature learning architectures achieve comparable or superior classification performance. The system integrates data acquisition, processing, and MLOps into a unified IoT-based pipeline, employing temporal convolutional networks, recurrent neural networks, and structured state space models. Experimental results show strong generalization capabilities and computational efficiency, even in low-data regimes.

Keywords: Traffic Control Systems, Remote Sensor Data Acquisition, Internet of Things
Abstract: This paper proposes a distributed adaptive sensor fusion architecture for an urban smart lighting system. The goal is to solve the problem of estimating the state of multi-agent systems that have three nodes: controllers, sensors, and agents. The system autonomously modulates street lamp brightness to decrease energy consumption based on vehicular traffic on designated road segments. To this end, a traffic model is proposed to estimate the number of vehicles in each segment by resorting to sensor network and state observer methodologies. At each time instant, the available information is adequately exploited by using a multi-agent reputation mechanism in turn based on distributed consensus and Perturb&Observe algorithms. A key feature of this methodology relies on its ability to identify the trustworthy sensors by evaluating an ad-hoc Quality of Service (QoS) metric. Simulations on a structured urban road network validate the effectiveness of the resulting algorithm.

Keywords: Control of Networks, Optimization
Abstract: Formation control allows agents to maintain geometric patterns using local information, but most existing methods assume ideal communication. This paper introduces a goal-oriented framework combining control, cooperative positioning, and communication scheduling for first-order formation tracking. Each agent estimates its position using 6G network-based triangulation, and the scheduling of information updates is governed by Age of Information (AoI) and Value of Information (VoI) metrics. We design three lightweight, signaling-free scheduling policies and assess their impact on formation quality. Simulation results demonstrate the effectiveness of the proposed approach in maintaining accurate formations with no additional communication overhead, showing that worst-case formation adherence increases by 20%.

Keywords: Intelligent Homes and Ambient Intelligence, Industry 4.0/5.0, Internet of Things
Abstract: This study investigates the potential of multi-disciplinary collaboration and hackathon-based approaches in addressing smart city challenges, with a particular focus on Kochi, India. It examines how interdisciplinary teams and participatory innovation models contribute to smart city development while identifying the key benefits and challenges of this approach. Employing a case study methodology, the research uses a hackathon as an experimental setting where participants from diverse backgrounds worked collaboratively to develop solutions for urban challenges in Kochi. Data was gathered through observations, stakeholder interactions, and collaborative analysis, with findings structured around key smart city attributes. The results indicate that multi-disciplinary collaboration enhances problem-solving depth but requires structured facilitation to align diverse disciplinary perspectives effectively. While the hackathon format fosters rapid innovation, its time constraints limit deep stakeholder engagement and iterative refinement. Additionally, digital collaboration tools improve efficiency, yet face-to-face interactions remain crucial for collective synthesis and solution integration. The study underscores the importance of contextual understanding, bottom-up approaches and bridging the smartness through technology to people-centric initiatives. As the findings are based on a single case study in Kochi, broader generalisability is limited. Future research should explore extended engagement strategies and comparative analyses across multiple smart city initiatives. By offering practical recommendations for improving hackathon-driven urban solutions, this study contributes to the discourse on people-centric, context-aware city and community development that focus on sustainability and quality of life.

Keywords: Internet of Things, Remote Sensor Data Acquisition, Industry 4.0/5.0
Abstract: Visual Anomaly Detection (VAD) is a key task in industrial settings, where minimizing operational costs is essential. Deploying deep learning models within Internet of Things (IoT) environments introduces specific challenges due to limited computational power and bandwidth of edge devices. This study investigates how to perform VAD effectively under such constraints by leveraging compact, efficient processing strategies. We evaluate several data compression techniques, examining the tradeoff between system latency and detection accuracy. Experiments on the MVTec AD benchmark demonstrate that significant compression can be achieved with minimal loss in anomaly detection performance compared to uncompressed data. Current results show up to 80% reduction in end-to-end inference time, including edge processing, transmission, and server computation.

Keywords: Internet of Things, Machine to Machine Communications, Mobile Sensor Networks with Low Energy
Abstract: This paper presents a compact ultra-wideband (UWB) microstrip patch antenna with triple-band rejection, printed on FR-4 substrate, desirable for various applications in IoT, communications and RF energy harvesting. The small antenna of dimensions 18 × 24 mm², achieves a broad impedance bandwidth from 3 GHz to 10.6 GHz, while effectively suppressing interferences from WiMAX (3.5 GHz), WLAN (5.5 GHz), and X-band satellite downlink (7.5 GHz). The band-notched characteristic is achieved by integrating accurately tuned thin rectangular slots into specific locations in the radiating patch. The antenna maintains a stable gain across the UWB range, with significant decrease at the notched frequencies. Simulation and measurement demonstrate strong agreement, validating then the proposed approach as low-cost, compact, and efficient solution for UWB applications and selective sensing, requiring interference mitigation. Furthermore, a use-case of RF energy harvesting was successfully conducted, with satisfactory output DC voltage, needed in wireless powering sensors with low energy.

Keywords: Information Modelling, Digital Twins, Asset Administration Shell
Abstract: To increase the effectiveness of maintenance of production machines, assistance systems are being developed to guide personnel more quickly and safely. The knowledge partly originates from technical documentation (TD), which today is mostly document-based. With the Model Based Systems Engineering (MBSE) approach, a shift towards a model-based TD emerges. This paper analyses how an MBSE model can be used for assistance systems and what is necessary to facilitate it. A theoretical structure model for TD is described which shows the interdependencies between maintenance tasks and artifacts of a TD. The modelling is done by using Systems Modelling Language and the Asset Administration Shell.

Keywords: Simulation Based Automation Engineering, Digital Twins, Advanced Engineering Methods for Automation Systems
Abstract: This paper presents a Model Predictive Control (MPC) strategy to optimize container and press operations in a waste sorting plant. To overcome the inherent complexity of the original integer programming problem, a simplified optimization method was developed. By analyzing the structure of the control action, the problem was reduced to optimizing only the press waiting time, dramatically decreasing optimization complexity. Validation across ten realistic scenarios showed a 13.2% efficiency improvement over the standard Optimal Analytic Procedure (OAP) and complete prevention of overflow. This work demonstrates how structural simplification in MPC can enhance operational reliability and efficiency in waste sorting systems.

Keywords: Industrial Control as a Service, Architecture of Automation Systems, Cloud-Based Control Systems
Abstract: The biopharmaceutical industry faces challenges in time-to-market, complexity, and scalability due to rigid traditional automation architectures. Industrial Control as a Service (ICaaS) offers a flexible, server-based alternative, virtualizing control functions to enhance the modularity and the pre-qualification of automation systems. Using the Design Research Methodology, this study investigates the impact of ICaaS on the biopharmaceutical industry, demonstrating a reduced time-to-market, an improved top-line growth and an overall reduced investment risk. ICaaS modernizes biopharmaceutical automation by enabling the usage of pre-qualified software modules and by lowering the effort required to deploy advanced technologies like AI or model predictive control. Additionally, it reduces investment risk by shifting CAPEX costs to OPEX costs. Nevertheless, future research should explore the standardization and the business models of ICaaS in more detail to allow for a broader adoption of this approach.

Keywords: Cyber-Physical Production Systems, Modularization and Flexibility of Production Systems
Abstract: Control software collects signals from sensors, processes them, and uses the results to drive actuators. This tight interaction means that developers or tools cannot directly infer the overall behavior of control software without also knowing the physical system. This particularly affects developers of IEC 61499-based software due to the introduced event-driven execution model. It allows them to precisely define the execution order of components, but can hinder understanding of the control flow, which may depend on data values. For maintaining or testing control software, developers can therefore benefit from visualizations of the control software behavior. This paper categorizes the structural and behavioral models of IEC 61499, illustrates their usage based on a demonstration example, and outlines how the application structure affects the presentation of the behavior. Distributed design patterns facilitate reuse and adaptability, but the implicit interactions between components may hinder understanding of the control software. Future work should therefore focus on compensating the disadvantage, for instance, by visualizing these interactions.

Keywords: Simulation Based Automation Engineering, Artificial Intelligence and Autonomous Systems, Advanced Engineering Methods for Automation Systems
Abstract: This paper presents a state estimation framework for water distribution systems (WDS) using a High-Gain Observer (HGO) built upon the recently proposed Elastic Water Column Model (EWCM). Unlike traditional methods that rely on algebraic relationships between head and flow or Partial Differential Equations (PDEs), the EWCM captures both hydraulic and elastic dynamics using ordinary differential equations (ODEs), facilitating a state-space representation of WDS. The proposed method employs the HGO to estimate unmeasured states by leveraging sparse sensor measurements. A 51-pipe network is used to demonstrate the performance of the proposed approach. Sensor nodes and prediction nodes are selected, and Gaussian noise is injected into the measurements. The results show that the HGO effectively estimates the pressure head across the network, showcasing its potential in real-time monitoring and control of WDS.

Keywords: Modularization and Flexibility of Production Systems, Asset Administration Shell, Architecture of Automation Systems
Abstract: Modular Plants (MPs), designed for rapid deployment and adaptability, offer faster time-to-market. Their energy-efficient operation can significantly lower the carbon footprint of industrial processes. This work presents an Energy Management System (EnMS) Plug-&-Monitor architecture that enables rapid EnMS deployment and immediate insights into MP performance and energy consumption. An Asset Administration Shell ensures energy transparency through Energy Key Performance Indicators (eKPIs), aggregated across modules. It facilitates the structured and standardized representation and exchange of energy data in cooperation with the Process Orchestration Layer. Proposed submodels, including eKPI, energy baseline, and energy data exchange, provide valuable datasets for the Digital Product Passport.

Keywords: Real-Time Aspects of Intelligent Control, Fault Management, Knowledge Processing and Representation
Abstract: The development of global manufacturing centers has shifted industry focus towards efficient and reliable supply chain networks composed of decentralized processing layers along the manufacturing pipeline. To benefit from this transition, it is important to deliberately account for dynamic demands, unpredictable disruptions in the capabilities of the processing layers, and also accommodate the diverse customer requirements expressed as high-level intents. This paper introduces an intent-based Model Predictive Control framework to optimize the scheduling of admissible flows between the Supply Chain Network, considering also fluctuating customer demands and the production rates of each processing node. The framework is compared against other baseline approaches, and the evaluation results indicate a significant reduction in the overall cost by achieving at least 39% reduction.

Keywords: Machine Learning, Computational Intelligence Methods, Intelligent Systems and Instrumentation
Abstract: Motivated by Smart Manufacturing and Industry 4.0, we introduce a framework for synthesizing Abstraction-Based Controller Design (ABCD) for reach-avoid problems from Natural Language (NL) specifications using Large Language Models (LLMs). A Code Agent interprets an NL description of the control problem and translates it into a formal language interpretable by state-of-the-art symbolic control software, while a Checker Agent verifies the correctness of the generated code and enhances safety by identifying specification mismatches. Evaluations show that the proposed approach increases the success rate of solving reach-avoid problems from 20%, when using LLMs directly, to 80%, while also enhancing robustness to linguistic variability across test cases. The proposed approach lowers the barrier to formal control synthesis by enabling intuitive, NL-based task definition while maintaining safety guarantees through automated validation.

Keywords: Fault Management, Knowledge Processing and Representation, Search Methods and Decision-Making, Architectures for Real-Time Intelligent-Control
Abstract: The adoption of Digital Twins paves the way for transforming the traditional operations of the manufacturing domain into the new era of digitalization. The aviation industry is adopting cutting-edge technologies to reform the daily business operations of aircraft maintenance schedules, shifting to a new service-oriented paradigm. This paper introduces a Digital Twin based on Coloured Petri Nets to evaluate the performance and provide dynamic adaptation of the maintenance fleet scheduling strategies by incorporating possible disruptions in the maintenance schedule. Simulation results showcase the framework's efficacy by providing a 22.3% reduction in cost compared to static maintenance strategies.

Keywords: Design Methodologies, Intelligent Systems and Instrumentation, Real-Time Aspects of Intelligent Control
Abstract: The increasing digitization and networking of manufacturing processes have led to the emergence of Manufacturing-as-a-Service (MaaS) platforms. However, the success of such platforms critically depends on the early and systematic elicitation of stakeholder-specific requirements. Therefore, the objective of the paper is to present a phase-based framework for requirements elicitation, which supports the identification, structuring and analysis of platform requirements by integrating diverse stakeholders. It is validated through its application to four industrial pilot use cases. The proposed approach comprises a multi-method elicitation process, combining process modeling, structured interviews, workshops and standardized documentation templates. As a result, the framework fosters a systematic development process for MaaS platforms, ensuring stakeholder alignment and technical feasibility.

Keywords: Data-Driven Applications, Real-Time Aspects of Intelligent Control, Neural Networks & Deep Learning
Abstract: This paper proposes a kernel-based spatio-temporal wind forecasting model that updates incrementally in real time onboard a quadcopter. We present algorithms for updating, downdating, and discounting model parameters, and importantly, an adaptive method to update and split cluster centres using a directional bimodality coefficient (BC) test. To predict wind at new locations, we use gaussian process (GP) interpolation, enabling the quadcopter to map wind conditions. Real-world evaluation shows improved forecasting accuracy.

Keywords: Diagnosis and Fault-Tolerant Control, Data-Driven Applications, Machine Learning
Abstract: Alarm floods remain a significant challenge in industrial processes. However, benchmark datasets for evaluating alarm flood analysis methods remain scarce. This paper introduces a novel dataset derived from a fluid catalytic cracking unit simulation model. We adapt the original simulation by incorporating an alarm system and defining realistic disturbances. The dataset's utility is demonstrated through an evaluation of three relevant alarm flood classification methods from the literature.

Keywords: Cyber Physical Systems
Abstract: This paper presents a federated learning (FL) framework for fault diagnosis of railway point machines (RPMs) in distributed environments. The proposed method integrates a personalized layer for client-specific adaptation and an Adaptive Feedback Kalman Filter (AFKF) mechanism to dynamically adjust aggregation weights based on local accuracy. Experimental results demonstrate enhanced convergence speed and diagnostic accuracy, especially under data heterogeneity and communication delays.

Keywords: Cyber Physical Systems, Control Through Networks, Intelligent Homes and Ambient Intelligence
Abstract: This paper explores the application of dynamic Periodic Event-Triggered Control (PETC) in HVAC systems for thermal management in buildings, considering external disturbances. A co-design approach is proposed for developing event-based control strategies aimed at optimizing energy consumption while ensuring thermal comfort. The optimization problem is formulated with Linear Matrix Inequality (LMI) constraints, focusing on minimizing the number of events and enhancing energy efficiency. The proposed method is evaluated through computational simulations to assess system performance under various operational conditions, highlighting the potential of PETC to improve the energy efficiency and robustness of HVAC systems in buildings.

Keywords: Mobile Sensor Networks with Low Energy, Traffic Control Systems, Smart and Connected Cars
Abstract: Recently, Vehicular Ad-Hoc Networks (VANETs) have emerged as a key component of intelligent transportation systems, facilitating communication between vehicles and roadside infrastructure. However, VANETS has a major challenge in ensuring secure and reliable authentication of violation reports, especially in environments that are prone to malicious attacks and mobility. This work proposed a signature-based authentication scheme and leverages the Rivest-Shamir-Adleman (RSA) algorithm to verify violation reports that help protect the integrity and privacy of communication between vehicles and road infrastructure. To evaluate the proposed scheme, we used OMNeT++, Veins, and Simulation of Urban Mobility (SUMO) tools. In addition, we performed an all-round analysis for key performance metrics, i.e., validation and violation report numbers, communication overhead, computational overhead, and signature verification time. The results achieved proved the effectiveness of the proposed scheme in ensuring the efficient provision of robust signature-based authentication messages. We consider this work a stepping stone for further investigation into sophisticated security models that will be developed for still- and moving-vehicle environments.

Keywords: Robotic Networks, Remote Sensor Data Acquisition, Industry 4.0/5.0
Abstract: This paper proposes a local path planning method for quadruped robots based on the fusion of 3D LiDAR point clouds and 2D visual semantic data, aiming to address the limitations of single LiDAR systems in capturing ground semantic information. The method optimizes path planning using extracted critical semantic features to ensure operational safety. Specifically, it employs semantic segmentation technology to extract key scene features and maps them into 3D point cloud space, constructing a spatial information representation integrated with visual semantics. Additionally, a spatial point cloud filtering strategy is designed to eliminate non-ground point clouds during local path planning, thereby reducing interference caused by semantic model misclassification. Experimental results demonstrate that the proposed method significantly enhances the safety and adaptability of robot path planning in complex scenarios.

Keywords: Intelligent Homes and Ambient Intelligence, Internet of Things, Remote Sensor Data Acquisition
Abstract: In recent years, the integration of digital technologies into urban infrastructure has driven the evolution of smart campuses, serving as testbeds for smart urban development and key platforms for innovation in sustainability and energy management. While the smart campus concept continues to gain traction, its practical implementation still faces considerable challenges, particularly in terms of energy efficiency, systems interoperability, and the complexity of managing diverse infrastructural components. This paper proposes a simulation model aimed at analyzing and optimizing energy usage within a smart campus environment. The model enables the evaluation of various energy management strategies and supports decision-making by modeling the building infrastructure and simulating scenarios related to space exploitation and workload, heat loads in lecture halls, cooling demands, and overall energy consumption. This simulation model will not only evaluate the current building energy consumption and provide recommendations for its improvement, but will also serve as a practical testbed for scalable energy solutions. By deploying this testbed in a smart campus environment, it will demonstrate how integrated smart technologies can improve energy efficiency in buildings and urban spaces, supporting broader sustainability goals.