(Manufacturing Autonomy Level 4 – MAL4)
The use case aims to deploy and validate MAL4 cognitive production processes in 4 public-private Pilot Lines (PL) located in:
The sites offer 5000 m2 of cutting-edge I4.0 production systems and bring together over 500 companies.
Evolution of existing modular manufacturing systems needs strategies towards mass customisation for processing a wide range of products. The creation of highly flexible, sustainable (green) modular digital production lines and manufacturing of a new product in a low-volume production (high customisation) requires often implementation of smart rapid response features in connection with self-optimisation, reconfigurations ramp-up, adaptation of the production line and the operations. The fast use and processing of data is of paramount importance to make intelligent automated human-centred augmented and assisted decisions. aerOS new forms of autonomous, flexible edge layer services with intelligent orchestration will introduce the enablers to raise production autonomy to Level 4 (see EFFRA CF2 pathways). aerOS distributed edge-powered modular approach will facilitate IoT edge-cloud continuity, and modular green production orchestration services for the digitalisation of AI/ML workflows support tactile ZDM control with real-time closed-loop processes, active energy monitoring and self-adaptive scheduling.
The 4 Pilot Lines (PL) extend the H2020 DIMOFAC facilities. ZDM Batch PL Spain (EU-DIH)- AIC European Center for the automotive industry: (i) Automotive Smart Factory (ASF) including cutting-edge equipment for stamping, machining, control and in-line verification, AGV and 5G MEC solutions. (ii) Zero Defect Manufacturing (ZDM) Excellence Centre, advanced digital platforms and quality control instrumentations (robot, manual CMM…). Autonomous Intra-Logistics PLs Italy – POLIMI Industry4.0Lab & MADE Competence Centre (EU-DIH). Fully automated assembly and manufacturing line with a robotic cell that simulates an industrial scenario for logistic and lean 4.0 production. Components are transported with AGV controlled by informative systems (orchestrator). Green Batch PL Switzerland – Smart Swish Factory demonstrating innovative Industry 4.0 AI manufacturing concepts built and operated by more than 50 companies. Lot-size-1 Automation PL Germany – Siemens HQ. with a full-fledged digital production line, digital machine, aerospace lab, dedicated AGV area & a dozen of robot arms to demonstrate AGV swarm intelligence & AMRs supported by robots. Represent the living lab for SotA SIEMENS automation portfolio and capacities (highest worldwide autonomy level) for on-the-fly production self-adaptation.
Scenario 1 – Green manufacturing (zero net-energy) and CO2 footprint monitoring product life cycle digital thread and sustainability (CO2 footprint) data models, in connection with the implementation of Digital Product Passport (DPP), enabling a systemic shift towards circular economy, supporting de-manufacturing operations, optimisation of reverse logistics infrastructure and more sustainable product design. It will experiment Gaia-X and aerOS services at PL3 (CH) to implement edge intelligence services (and analytics) to optimise impact and footprint of modular production lines for drones.
Scenario 2 – Automotive Smart Factory Zero Defect Manufacturing. A ZDM approach, ensuring robustness and stability of the process, deploying inline quality control among the manufacturing workflow. Remote tactile human-CMM interaction (where resilience policies and SLAs must be reached); energy efficiency monitoring and real-time machine error compensation extending 5GROWTH set-up to ensure accuracy of product quality dimensional inspection will be showcased at PL1 (ES). The aerOS system will enable the seamless interaction of quality control intelligence engine with a wide range of dimensional instrumentation equipment, hybrid CMMs, arm robots or in machine-tool metrology. Considering that an Industrial IE down in a manufacturing line can have major negative economic impacts, self-recovery, healing and diagnosis will be needed.
Scenario 3 – AGV swarm zero break-down logistics & zero ramp-up safe PLC reconfiguration for lot-size-1 production. The business process incorporates process specific information (MES) with context information about what needs to be done, which is feed to the edge-cloud continuum by means of SIMATIC Industrial edge apps (pub/sub schemas) to feed aerOS. Production flexibility (reconfigurability) is realised by on-the-fly AGVs decision making and robot path calculations. aerOS automatic transport and safe placement of robotic arms is safely realised with safety enabled PLCs that enrich the process with critical device data and communicate with stationary safety devices like light fences via edge. PL2 (IT) will showcase advanced logistic processes (real-time benchmarking); whereas PL4 (DE) will showcase safe and secure automation of production. Diverse IE nodes will co-exist and collaborate, orchestrated by a master node.
Use case will be driven by partners CF and ELECT. It will allow management of containerised edge data centres developed by CF and located directly at energy sources, connected to the smart infrastructure and providing cloud continuity. It will be deployed and validated at renewable energy centres operated by ELECT, in Poland.
Initiatives like European Green Deal and Fit for 55 package, signify growing need for sustainable solutions in edge-cloud industry. Energy intensive nature of traditional datacentres makes it less and less feasible to expand operations by buying, or building, more space, as indicated by DAIRO. Traditionally, resource-demanding services (e.g., AI) have been deployed in the cloud. Thus, approaches based in decentralised edge devices, such aerOS, may improve performance. By leveraging aerOS platform’s capabilities European cloud industry will be able to take advantage of potential benefits for including numerous far and near edge nodes located at renewal energy producing locations, (thus, rapidly scaling operations), while keeping computing/storage capacity due to orchestration capabilities.
The use case deployment will leverage CF’s existing edge-cloud infrastructure (i.e., WAW 2-1 and, potentially, FRA 1-1), monitoring and administration tools, data repositories and processing chains, front end portals and dashboards for processing disposition. CF cloud infrastructure is federated in the framework of the Gaia-X project. The deployment will make use also use the Everything Monitoring and Control System software for energy source and throughput monitoring and reporting tool, already in use in ELECT premises.
Use case will proof applicability of aerOS for managing small, edge nodes located directly at energy producing locations, gathering information and events from the deployed smart devices. The edge nodes will have connectivity to the private cloud infrastructure of CF. The objectives of the use case: (1) aerOS will be used for distributing, monitoring and relaying tasks of stateless processing – allowing execution of batch or FaaS-like activities – among a pool of near and far-edge nodes located at ELECT renewable energy premises (e.g., wind/photovoltaic farms, hydro energy or high capacity batteries); (2) use of heterogeneous information in the orchestration and scheduling model (master DB and registry will be needed along the task collector and distributor), boosting the energy and resource optimisation; (3) proposed application shall also result in lower capital intensity of the system, allowing the operator to abandon redundancy at a node level and assuring it on system wide level, as tasks of a failing node can be transferred to an operational one
Scenario 1.- Green Edge Processing: aims to deploy two federated edge nodes and a private CF cloud located directly at renewable energy premises, and connected to different smart devices and data sources from wind and PhV farms operated by ELECT. Managing the system shall be performed in energy, network and self-conscious manner, measuring the reductions provided (benchmarking of parameters based on real-time own analytics in the IE) in the orchestration of tasks deployed in the edge instead than in the cloud (e.g., AI). With the changes in requirements for computing resources, available energy or network throughput, aerOS will facilitate rapid changes (self-scalability, self-automation) in task distribution though orchestration (managing topology, tasks and services).
Scenario 2 – Secure Federation of edge/cloud: will consider the management of data in the edge nodes, having aerOS edge nodes and the support of ad-hoc cloud components and federation with Gaia-X as support for the applications (considering that there will be external cloud nodes involved, the importance of trust and security is huge). The exchange of frugal AI models, and the evaluation of energy consumption will be performed. Energy premises are considered critical infrastructures, and the cybersecurity component so as privacy and trust (along with policy engine rules) will be evaluated to provide an IoT edge-cloud secure, private and trustable continuum.
to Enable CO2 Neutral Farming (HPCP-F)
Use case will be driven by partners JD and TTC. The smart agriculture HPCP-F use case will be deployed and validated in John Deere European Technology Innovation Center in Kaiserslautern (Germany).
Digital transformation in agriculture is fast progressing. Especially, Precision Farming offers a pathway towards reducing inputs, maximising yields and quality of goods. Digitalisation allows integrated control of machines involved in production. At the same time, farming needs to interact with other production systems and information services in food production and food value chains. Edge computing, in connection with limited / temporary networks (as connectivity in rural areas has strong limitations), will enable the deployment of intelligence without permanent connectivity to the cloud (self-dependability). is a key to synchronise and optimise the tractor work for future productive and sustainable farming. Current systems, e.g., connected and cooperative agricultural mobile machinery, are pushed to resources limits, in tasks like data access and processing, ensure data privacy and security or provide continuity to the cloud. In-vehicle edge nodes (e.g., JD edge), interacting with smart devices, networking components and compute continuum, will get benefited with the support provided by an IoT edge-cloud continuum solution.
The use case will profit current infrastructure in JD premises in Germany, agricultural machine technologies and related services, including precision farming, test fields, prototype construction machines, and wired as well as battery-supported tractors and swarm-based field device technology. A wide range of commercial devices and sensors including Safety Electronic Control Units; I/O modules; cloud, edge and IoT gateways; highly robust Human Machine Interface displays, and networking architectures (including classic communication schemas like publish-subscribe) will be leveraged for the use case.
The main objective of the HPCP-F use case is to integrate, test and validate High-Performance Computing Platform for connected and cooperative agricultural mobile machinery. The proposed robust and flexible solutions need to provide M2M connectivity from everywhere for large-scale agricultural production system; real-time performance with low latency networking; federated frugal AI capabilities to improve performance in the edge and data management including autonomy (self-automation), ownership (trust for origin needed), storage and interoperability. This use case will contribute to enabling sustainable farming solutions for energy optimisation and noise reduction.
Scenario 1 – Cooperative large-scale harvesting: The use case will optimise a large-scale harvesting system. Data from sensors (e.g., cameras, LIDAR, Radar) as well as operating instructions from cloud will be safely and securely processed to feed a grid-connected electric swarm. An IoT edge-cloud continuum approach will be adopted for the orchestration of services (AI/ML-based assignment), optimisation of data autonomy (with semantics) of a swarm of vehicles, each equipped with own far edge node executing aerOS (IE registry will play a key role here) and connected to the smart devices and sensors of the vehicle. Developed solution will be capable to e.g., perform computational tasks in support of demonstrating fully electric swarm of vehicles safely and securely operating in platooning (up to TRL5), governed by a robust policy engine, which will be integrated and validated in the JD’s functional prototype vehicle.
Scenario 2 – CO2 neutral intelligent farming: deploying of tasks in the edge, it is possible to reduce the latency and reaction time, by using low-latency networks like 5G or TSN. The benefit is also related with the energy consumption when transferring AI and real-time embedded analytics, what may reduce the CO2 impact. The scenario will measure in a collaborative swarm of vehicles the energy consumption reduction due to the use of aerOS and different federation topologies, considering how frugal AI may affect performance.
The logistics pilot will be driven by the Industrial partner EUROGATE and the scenarios will be deployed and validated in the container terminal located in the Port of Limassol (EGCTL), which is the largest port in Cyprus. EGCTL amalgamates the activities of container handling, reefer services and industrial storage and it is a critical node of the European sea-logistics supply chain. Since 2017 EUROGATE has invested more than 30 M€ in the CHE of the terminal improving the QoS while reducing both arrival/departure idle times and cargo delivery time.
Cargo Operations in EGCTL are currently handled by several Quay and Yard cranes. These cranes are heavy machinery that have several subsystems running internally, managed by different PLC controllers. The PLCs take the input directly from the cranes, execute a particular logic programmed, and generate outputs to enable terminal staff to manipulate the crane. PLCs are, therefore, the most accurate representation of the status of a port crane. However, current Big Data, AI/ML, and IoT are based on remote servers and/or cloud technologies. Thus, there is a gap between the place where the data is most precise, and where the KPIs are analysed for carrying out estimations and predictions. This gap has implications with respect to observability in real-time at the upper layers of the terminal system due to higher latencies between the central server and the source where the data is originated. These constrains limit the terminal efficiency as it could not prevent possible unforeseen issues such as outages.
Beyond cranes and PLCs, the TOS and the Computerised Maintenance System (CMS) of EGCTL will be accessible and integrated to aerOS While the TOS enhances the operational efficiency of the terminal assets, the CMS facilitates the processes of scheduling the availability of the equipment required for maintenance.
Since physical expansion of terminals is difficult or impossible the only way to improve the operational performance is by following the Industry 4.0 (I4.0) digitalisation paradigm, which will allow to take better decisions by improving the availability of the information and the way it is presented to the staff. Whilst better computer vision and accurate predictive maintenance services directly deployed in the edge are not possible through the first generation of IoT architectures, aerOS will allow to orchestrate smart services in the edge, allowing maritime companies to react faster without the need of a high-performance processing in the cloud. The expected benefits of this pilot are: (1) ensure that the data generated in the sources of the information are manipulated at the edge; (2) enough computing performance in the edge elements in compliance with the smart orchestration approach of aerOS, and (3) new AI and abstraction cloud methodologies application that will allow sophisticated cognitive services validation.
Scenario 1 – Predictive maintenance of Container Handling Equipment: Predictive maintenance can help identify maintenance issues ahead of time, allowing terminal staff to perform cost-effective duties and to extend the lifespan of the industrial assets with minimal cost. Proprietary software solutions like the TOS and CMMS will be able to exchange data in a secure, trusted environment, allowing the maintenance team to take better decisions faster. Supported by the aerOS self-*, analytics and AI tools, tailored to distributed autonomously-orchestrated continuum.
Scenario 2 – Risk prevention via computer vision in the edge: Efficient computer vision algorithms at the edge will allow the terminal to automatically: (1) identify containers with damages, and (2) check for the existence of container seals without the need of human action. Intelligent orchestration of distributed applications will permit analysing video streams at the edge, lowering human mistakes and safety risks while empowering Port Communities industrial ecosystems, achieving a secure, trustable and self-orchestrated IoT edge-cloud continuum via leveraging open standards, recommendations and security-related tools provided by aerOS.
Use case will be driven by partners COSM, NCSRD, FOGUS, INF and UPV. It will demonstrate gains of the aerOS architecture in an edge deployment for energy efficient, sustainable, flexible and health-safe smart buildings. The use case will be will be deployed and validated in an office enterprise building of COSMOTE (Athens, Greece).
Under the constraints shaped by the COVID-19 pandemic, fitting as many people as possible into a workspace, to maximise utilisation of commodities is not acceptable. Proper employees’ placement, social distancing, energy efficiency, along with business and personal preferences becomes a complex dynamic task. Real-time processing of data and decision making close to events, supporting distribution through aerOS capabilities, can offer autonomous solution for safe and sustainable workplaces.
Implementation will leverage the IoT-edge solution prototyped by COSMOTE R&D and already installed in residences and COSMOTE’s sites. Architectural layout of the solution consists of: (a) Wide range of commercial (and custom) end-devices/sensors, (b) Multi-purpose IoT gateways to support wide range of use cases, (c) Cloud infrastructure, enabling gateway/device management, data storage, data processing, (real-time and historical) data visualisation, cloud-based dashboards. Solution uses open source software, commercial hardware, and open APIs.
Use case aims to demonstrate the aerOS architecture: (1) applied in Smart Buildings market to optimise efficiency and safety of enterprises based on process and data autonomy and self-orchestrated IoT ecosystems, (2) energy efficiency of the large buildings using real-time processing and (frugal) AI, (3) use 5G and smart network components of the infrastructure like NFV and NetApps to extend aerOS capabilities. Expected benefits will be derived from aerOS nodes intelligence, addressing distinctive infrastructure characteristics of buildings, through autonomous and decentralised decision-making at the edge. Moreover, the aerOS unique abstraction approach will offer an adaptable solution that can bridge heterogeneity (data and platforms), so that sensors, systems, and analytics could be orchestrated in the IoT edge-cloud continuum, and new IE or federating with new elements added.
Scenario 1 – Intelligent Occupational Safety and Health – aims to integrate aerOS in the IoT ecosystem of a COSM corporate building in Athens to measure energy, luminosity, CO2, humidity, temperature, motion detection, and desk occupancy. Collected data (real-time and historical ones) and relevant algorithmic criteria (via embedded analytics) shall determine the proposed clustering and seating of employees. Actuation on ventilation, heating/cooling systems, or luminosity powered with edge intelligence among aerOS IEs (supported by frugal AI) shall allow secure, trusted decentralised management of each room, so that the office becomes self-organised in terms of health and efficiency, providing low latency reaction if condition change.
Scenario 2 – Cybersecurity and data privacy in building automation – COSM will deploy authentication and authorisation mechanisms, so as the aerOS cybersecurity component in order to secure the whole IoT edge-cloud continuum. Data governance mechanisms in order to store, ownership and sharing of the data will be deployed using the corresponding aerOS components. Use of 5G capabilities, VNF and potentially NetApps to execute certain security and privacy functions will be evaluated. As video identification will be put in place for safety and security, anonymisation mechanisms and secure storage will be used in the scenario.