Skip to main content

Resultados relacionados

Resultados totales (Incluyendo duplicados): 10
Encontrada(s) 1 página(s)

RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

Sustainable optimisation approaches for production planning and control to evolve towards industry 5.0

  • Guerrero-Moreno, Blanca María|||0000-0002-8284-501X
  • Mula, Josefa|||0000-0002-8447-3387
  • Poler, R.|||0000-0003-4475-6371
[EN] Industry 4.0 (I4.0) has led to a very high development potential in the optimisation of production planning and control problems in industrial companies. This study aims to analyse the existing scientific literature on the optimisation of planning, control and management problems, specifically in the production, operations and scheduling areas, to make industry more sustainable in accordance with the I4.0 context; that is, applying new digital technologies and I5.0 by integrating people into this digital transformation. Specifically, 77 research works were identified and analysed. The main findings conclude that the key setting of these optimisation problems lies in manufacturing contexts and highlight the use of modelling and solution techniques through mathematical programming and metaheuristics. These studies, which are mostly empirical, point out different software tools, and underline the use of MATLAB, optimisation with CPLEX, the Python programming language and simulation through AnyLogic. For I4.0 techniques, data analytics and big data, artificial intelligence, the Internet of Things, sensors and simulation and digital twins are the most studied. Proposals addressing worker ergonomics and a three-pillar approach to sustainability are particularly noteworthy. Finally, the main research gaps, challenges and trends for future research are recognised., This work was supported by Generalitat Valenciana: [Grant Number PROMETEO/2021/065]; Horizon 2020 Framework Programme: [Grant Numbers: 101177842 and 101177368]; Ministerio de Ciencia e Innovación: [Grant Number: PDC2022- 133957-I00].
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

AgentBlock: Blockchain-Integrated Multi-Agent Robotic Coordination with Reinforcement Learning for Autonomous Manufacturing

  • Velastegui, Rommel
  • Poler, R.|||0000-0003-4475-6371
  • Díaz-Madroñero Boluda, Francisco Manuel|||0000-0003-1693-2876
[EN] Centralised architectures in contemporary manufacturing systems impose structural constraints on resilience, scalability, and operational transparency that existing approaches have failed to resolve. This work reports the development and empirical validation of AgentBlock, a framework integrating blockchain technology with multi-agent robotic systems to enable decentralised autonomous manufacturing. The architecture operates across three functionally decoupled layers: a React-based decentralised application interface, an Ethereum Sepolia blockchain interaction layer with Solidity 0.8.18 smart contracts following an upgradeable proxy architecture (EIP¿1967) coordinated through an Industrial PoA consensus mechanism, and a physical execution layer comprising two heterogeneous robotic agents (KUKA youBot and UFactory Lite 6) and one edge validation agent on an NVIDIA Orin platform that also hosts Q-Learning optimisation, with inter-agent coordination provided by ROS Noetic Ninjemys under Ubuntu 20.04 LTS. Experimental validation conducted over 15 days across 1500 training episodes in a controlled 5 m × 3 m industrial laboratory reveals a task success rate of 95.58%, sustained throughput of 49.0 tasks per hour, average cycle time of 1.224 min, blockchain transaction latency below 15 s (mean: 11.4 s), and gas costs averaging US $0.000669 per operation. These findings establish that blockchain-enabled autonomous manufacturing is not merely theoretically sound but operationally viable, delivering immutable traceability, decentralised coordination, and transparent verification at performance levels compatible with Industry 4.0 and 5.0 production demands., The research that led to these findings received funding from the Horizon Europe Framework Programme (HORIZON) with Grant Agreement No. 101177368 Agile Manufacturing as a Service through AI Autonomous Agents (MaaSAI).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

Reference architecture for the design and implementation of AI systems in manufacturing in conformity to ISO/IEC 42001

  • Mateo-Casalí, Miguel Ángel|||0000-0001-5086-9378
  • Boza, Andres|||0000-0002-5429-0416
  • Fraile Gil, Francisco|||0000-0002-3275-7740
  • Maaßen, M.
  • Heymann, H.
  • Leyendecker, L.
  • Grunert, D.
  • Schmitt, Robert H.
[EN] The increasing adoption of artificial intelligence (AI) and machine learning (ML) in manufacturing is driven by the need for more personalised, efficient and adaptive production systems. However, industrial AI systems must also comply with emerging governance frameworks, particularly ISO/IEC 42001:2023, the AI management system standard. Despite its relevance, manufacturers often struggle to translate the standard¿s high-level requirements into practical design and implementation decisions.This article proposes a reference architecture to support the design, deployment and governance of AI systems in manufacturing in accordance with ISO/IEC 42001:2023. The architecture defines the main system layers, functional components, and governance mechanisms required to ensure lifecycle control, traceability, operational integration, and alignment with compliance. Given the recent publication of the standard, this work provides an early contribution to operationalising AI management system requirements within industrial architectures. The architecture is applied to a real-world metal manufacturing use case focused on predictive maintenance. The implementation demonstrates its feasibility and improves system traceability and operational performance, including overall equipment effectiveness and maintenance-related downtime. Although validated in a specific context, the modular and layered design supports adaptation to other manufacturing environments, considering domain constraints, available infrastructure and legacy system integration., The research that led to these findings received funding from the European Union Horizon Europe research and innovation programme under the Grant Agreement [101177368] Agile Manufacturing as a Service through AI Autonomous Agents (MaaSAI) .
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

A Hybrid Reinforcement Learning-Blockchain Architecture for Sequential Decentralised Coordination of Heterogeneous Robots: Proof-of-Concept Validation and Latency Characterisation

  • Velastegui, Rommel
  • Poler, R.|||0000-0003-4475-6371
  • Díaz-Madroñero Boluda, Francisco Manuel|||0000-0003-1693-2876
[EN] Contemporary manufacturing systems exhibit structural limitations arising from centralised architectures, which constrain their resilience, scalability, and operational transparency. The reliance on central controllers in multi-agent robotic environments creates single points of failure, hinders immutable traceability, and obstructs the independent auditing of critical processes, thereby limiting the feasibility of truly autonomous and self-orchestrating systems demanded by emerging Industry 4.0 and 5.0 paradigms. This study introduces a conceptual framework for blockchain-enabled autonomous manufacturing, integrating decentralised applications, smart contracts, and distributed robotic agents. We propose and experimentally evaluate a three-layer decentralised architecture integrating tabular Q-learning and Ethereum smart contracts for heterogeneous multi-robot coordination in manufacturing environments. The architecture eliminates centralised control servers by routing coordination signals through blockchain-verified events, enabling immutable task logging and sequential verifiable agent synchronisation in a two-agent laboratory configuration. An experimental prototype comprising a KUKA youBot mobile agent-trained via Q-learning over a four-state, four-action navigation policy-and a UFactory Lite 6 industrial manipulator was deployed in a laboratory environment under ROS Noetic. Quantitative evaluation demonstrates that the Q-learning agent converges to a 96% success rate within approximately 600 training episodes, and that blockchain-mediated coordination introduces a mean end-to-end latency of 45.7 s (blockchain overhead: similar to 15 s), acceptable for supervisory task orchestration but incompatible with real-time control loops. These results provide a reproducible proof-of-concept baseline for two-agent sequential blockchain-RL coordination under controlled laboratory conditions, establishing a foundation for future scalability analysis and identify scalability, concurrency, and fault tolerance as critical directions for future work., This work was supported by the Horizon Europe Framework Programme (HORIZON) through "Agile Manufacturing as a Service through AI Autonomous Agents" (MaaSAI) under Grant 101177368.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

Integrated Methods in Replenishment Planning Processes

  • Andres, B.|||0000-0002-7920-7711
  • MENGUAL RECUERDA, ANA|||0000-0003-4727-9099
  • Pérez-Molina, Ana Isabel|||0000-0002-1266-7027
  • Pérez Bernabeu, Elena|||0000-0002-9221-7623
[EN] Small and medium-sized enterprises (SMEs) face significant chal-lenges in digitalizing decision-making processes, particularly in demand forecast-ing, replenishment, and production planning. This study investigates the appli-cation of integrated methods in replenishment planning. A set of key hybrid approaches employed in replenishment planning, including deep reinforcement learning, hybrid machine learning forecasting models, and optimization model have been identified. The findings indicate that integrating data-driven models with knowledge-based systems enhances decision-making, optimizes inventory levels, and mitigates procurement risks. Despite the benefits of hybrid approaches, fur-ther research is needed to expand integrated replenishment models, particularly for industries with complex supply chains. Future developments should focus on combining rule-based models with AI-driven predictive analytics to improve replenishment efficiency and adaptability in dynamic markets., The research that led to these findings received funding from the Horizon Europe Framework Programme (HORIZON) with Grant Agreement No. 101177368 Agile Manufacturing as a Service through AI Autonomous Agents (MaaSAI) and with Grant Agreement No. 101177842 Unified Modeling and Automated Scheduling for Manufacturing as a Service (UniMaaS). From Aid for First Research Projects (PAID-06-24) and (PAID-11-24), by the Vicerectorate for Research of the Universitat Politècnica de València (UPV) and from Generalitat Valenciana Grants for doctoral researchers to work in companies in the Valencian Community (CIAEST2024/52).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

Toward lean industry 5.0: a human-centered model for integrating lean and industry 4.0 in an automotive supplier

  • Hines, Peter
  • Magnani, Florian
  • Mula, Josefa|||0000-0002-8447-3387
  • Sanchis, R.|||0000-0002-5495-3339
[EN] This paper proposes a human-centered conceptual model integrating lean and Industry 4.0 based on the literature review and validated it through a case study in the context of an advanced automotive first-tier supplier. Addressing a significant gap in existing research on lean Industry 4.0 implementations, the study provides both theoretical insights and practical findings. It emphasises the importance of a human-centered approach, identifies key enablers and barriers. In the implementation process of the case study, it is considered at group level and model site level through operational, social and technological perspectives in a five-phase multi-method approach. It shows what effective human-centered lean Industry 4.0 implementation look like and how advanced lean tools can be digitised. It highlights 26 positive and 10 negative aspects of the case and their causal relation. With the appropriate internal and external technological know-how and people skills, it shows how successful implementation can benefit the organisation and employees based on the conceptual model that serves as a first step toward lean Industry 5.0., The research that led to these findings received funding from the MCIN/AEI/10.13039/501100011033 and by European Union Next Generation EU/PRTR with grant agreement PDC2022-133957-I00 (CADS4.0-II); the Regional Department of Innovation, Universities, Science and Digital Society of the Generalitat Valenciana entitled "Industrial Production and Logistics Optimization in Industry 4.0" (i4OPT) (Ref. PROM-ETEO/2021/065); and the Horizon Europe Framework Programme (HORIZON) with Grant Agreement No. 101177842 "Unified Modeling and Automated Scheduling for Manufacturing as a Service" (UniMaaS) and Grant Agreement No. 101177368 "Agile Manufacturing as a Service through AI Autonomous Agents" (MaaSAI).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

Revolutionising industrial operations: The synergy of multiagent robotic systems and blockchain technology in operations planning and control

  • Velastegui, Rommel
  • Poler, R.|||0000-0003-4475-6371
  • Díaz-Madroñero Boluda, Francisco Manuel|||0000-0003-1693-2876
[EN] This article presents a meticulous literature review that focuses on the integration of blockchain technology (BCT) into multiagent robotic systems (MARS) for operations planning and control (OPC) in industrial settings. By employing a systematic approach involving research question formulation, document exploration, article screening and data analysis, we rigorously examined 276 articles from scientific databases, including WOS, IEEEXplore, and Scopus. By the PRISMA, CIMO, and Cochrane methodologies, we developed a refined selection process and established a clear taxonomy that elucidates the intricate relations among MARS, BCT and OPC based on their objectives, findings, solution methodologies, industry sectors, limitations and benefits. We additionally conducted an in-depth analysis of the current state of the art, by critically evaluating existing research. With this comprehensive review, we answered pertinent research questions by unveiling the profound benefits of applying BCT to MARS in production planning and operations. Notably, this integration enhances data security, ensures the traceability of MARS activities, and significantly improves transparency, efficiency and overall operational robustness in OPC in industrial contexts to provide valuable insights for companies to consider adopting this technology in their industrial operations., The research that led to these findings received funding from the MCIN/AEI/10.13039/501100011033 with Ref. PDC2022-133957-I00 "Validation of transferable results of optimisation of zero-defect enabling production technologies for Supply Chain 4.0" (CADS4.0-II) and the Horizon Europe Framework Programme (HORIZON) with Grant Agreement No. 101177368 "Agile Manufacturing as a Service through AI Autonomous Agents" (MaaSAI) and Grant Agreement No. 101058589 "AI Powered human-centred Robot Interactions for Smart Manufacturing" (AI-PRIMS).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

Analysis of Deployment Patterns for Edge Hosted Multimodal AI in Cloud Edge Robotic Systems Utilizing ROS 2

  • Mateo-Casalí, Miguel Ángel|||0000-0001-5086-9378
  • González-El Yachouti, Daniel|||0009-0005-5163-0509
  • Boza, Andres|||0000-0002-5429-0416
  • Fraile Gil, Francisco|||0000-0002-3275-7740
[EN] Cloud-edge robotics combines on-board computing with cloud services to enhance the intelligence and scalability of robotic systems. However, professionals still lack guidance on how to package and deploy AI components. This article presents a comparative evaluation of four ROS 2 deployment patterns for cloud-edge robotic AI: monolithic containers, microservices, dynamic module loading and overlay workspaces. All AI inference components run on the edge node in every pattern, so the comparison characterises the packaging and lifecycle mechanics of an edge-hosted, orchestrated, delivered, and observed multimodal workload. The four patterns are deployed on a K3s cluster combining an NVIDIA Jetson AGX Orin edge node with two amd64 cloud nodes, and run an identical workload combining perception (YOLOv8-nano), visual reasoning (LLaVA-1.5-7B) and voice interaction (Voxtral-Mini-3B), with a deterministic LLaVA trigger ensuring identical contention across patterns. Deployment cost is measured under three explicit regimes, warm restart, image cache cold start and fully clean deployment, with means and 95 per cent confidence intervals over replicated campaigns, and runtime behaviour is characterised with queue wait and frame drop instrumentation together with a sensitivity sweep of the reasoning trigger period from 30 to 300 seconds and of the camera input rate. The results reveal a clear trade-off. The dynamic pattern hot swaps a module in the live process in 0.78 to 31.10 seconds depending on the module, which is about two orders of magnitude faster than the rebuild and redeploy cycle for the lightweight nodes but only about one order of magnitude faster for the two model heavy nodes, and it collapses the real-time perception path when several AI nodes share a single Python interpreter, a degradation that a controlled hybrid experiment and a GPU utilisation control attribute to single interpreter serialisation rather than to the reasoning schedule or to accelerator contention. The overlay pattern restarts an operating point on par with the monolithic baseline yet is the slowest to provision from scratch or to update because of its coarse layer granularity, while the process isolated monolithic and microservices patterns trade update agility for runtime robustness. The results show that no pattern dominates. The overlay pattern offers an operating point restart on par with the monolithic warm baseline and, once its per layer carrier is activated, the smallest recurring update payload, but the slowest fully clean deployment of the four patterns, and the runtime collapse of dynamic loading is specific to the evaluated all Python single process composition and persists across the whole range of trigger periods, so it reflects the composition rather than the workload intensity. Fleet-scale benefits discussed for the overlay pattern are extrapolations from a single-edge testbed rather than measured results., This work was supported in part by the European Union Horizon Europe Research and Innovation Program Agile Manufacturing as a Service through Artificial Intelligence (AI) Autonomous Agents (MaaSAI) under Grant 101177368; and in part by the CLOTHIA Project (Agéncia Valenciana de la Innovació, IVACE+i Innovació, co-financed by the European Union through the European Regional Development Fund (ERDF) Program of the Comunitat Valenciana), from 2021 to 2027 (INNEST/2025/608). Funding for open access charge: CRUE-Universitat Politècnica de València.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

Data-Driven and Knowledge-Based Methods for Strategic Decision-Making in Supply Chain Management

  • Mojica, Ana
  • Andres, B.|||0000-0002-7920-7711
  • Torre-Martínez, María Rocío de la|||0000-0002-8662-8901
[EN] The COVID-19 pandemic underscored the importance of resilient supply chains (SC), leading to research on Artificial Intelligence (AI) in disruption detection and contingency planning. AI enhances data analysis, demand forecasting, and resource optimization, improving SC efficiency, resilience, and sustainability. This study presents a literature review aimed at identifying the current trends in the use of AI data-driven methods, knowledge-based methods, and hybrid methods that combine both, to support strategic decision-making within the supply chain management, such as, supplier selection, network design, capacity definition, or risk management. The objective is to define a conceptual framework that supports determining which methods are most used to support each of the strategic decisions, so that can be applied to enable robust and informed decision-making in supply chain management., The research leading to these results received funding from the Vicerrectorado de Investigación of the Universitat Politècnica de València (PAID-11-24), from the Horizon Europe Framework Programme (HORIZON) with Grant Agreement No. 101177368 Agile Manufacturing as a Service through AI Autonomous Agents (MaaSAI) and also with Grant Agreement No. 101147855 Intelligent and Sustainable Building Management powered by Cross-Sectoral Lifecycle (DATAWiSE).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Artículo científico (JournalArticle)

Towards a Reference Architecture for Machine Learning Operations

  • Mateo-Casalí, Miguel Ángel|||0000-0001-5086-9378
  • Boza, Andres|||0000-0002-5429-0416
  • Fraile Gil, Francisco|||0000-0002-3275-7740
[EN] Industrial organisations increasingly rely on machine learning (ML) to improve quality, maintenance, and planning in Industry 4.0/5.0 ecosystems. However, turning experimental models into reliable services on the production floor remains complex due to the heterogeneity of operational technologies (OTs) and information technologies (ITs), including implementation constraints, latency in edge-fog-cloud scenarios, governance requirements, and continuous performance degradation caused by data drift. Although Machine Learning Operations (MLOps) provides lifecycle practices for deployment, monitoring, and retraining, the evidence is fragmented across tool-centric descriptions, case-specific pipelines, and conceptual architectures, offering limited guidance on which industrial constraints should inform architectural decisions and how to evaluate solutions. This work addresses that gap through a PRISMA-guided systematic review of 49 studies on industrial MLOps (with the search and screening primarily targeting Industry 4.0/IIoT operationalisation contexts, as reflected in the search strategy and corpus) and an evidence-based synthesis of principles, challenges, lifecycle practices, and enabling technologies. From this synthesis, industrial requirements are derived that encompass OT/IT integration, edge-fog-cloud orchestration, security and traceability, and observability-based lifecycle control. On this basis, a reference architecture is proposed that maps these requirements to functional layers, data and control flows, and verifiable responsibilities. To support reproducibility and practical inspectability, the article also presents an open-source architectural instantiation aligned with the proposed decomposition. Finally, the evaluation is illustrated through a predictive maintenance use case (tool breakage) in a single CNC machining cell, where the objective is to demonstrate end-to-end feasibility under realistic operational constraints rather than cross-scenario superiority or broad industrial generalisability., The research that led to these findings received funding from the European Union Horizon Europe research and innovation programme under the Grant Agreement 101177368 Agile Manufacturing as a Service through AI Autonomous Agents (MaaSAI) .