CASANDRA completes its first year with the detailed definition of use cases and application scenarios, laying the foundations and initiating the development of the different technologies that will be applied to them

CASANDRA
(EXP – 00169779 / MIG-20242031)

Europe leads the manufacturing of complex and large-scale components. In Spain, the metal-mechanical sector is also booming, with significant growth (>8%) in recent years. This privileged competitive position is supported by the quality of manufactured products, their characteristics, high level of customization, efficiency, and lifecycle-related services. However, compliance with increasingly demanding regulations, high quality requirements, and product complexity are pushing existing technologies to their limits, threatening the sector’s competitiveness and its current privileged position. CASANDRA was created to address these challenges.

The main objective of the CASANDRA project is to develop a closed-loop digital manufacturing methodology that enables the concepts of a connected, intelligent, and autonomous factory, allowing the evolution of the digital twin of the component to be recorded throughout its lifecycle. This requires distributed monitoring and control tools, following a distributed data space model.

At the factory level, CASANDRA proposes the development of flexible, ubiquitous, and connected digital tools (collaborative robotics, AR/VR, and exoskeletons), designed to extend operators’ capabilities and increase efficiency, quality, and precision in the execution of required manufacturing processes.

Monitoring and control tools, together with transversal digital tools, open the door to the implementation of data analysis strategies that allow the collected data to be exploited. This enables CASANDRA to implement an autonomous factory concept through a decision-support system in which, based on collected live data, strategies are developed to adjust processes and continuously adapt to changing conditions and degradation of physical models. This involves working on edge intelligence, federated learning, and frugal AI by combining active learning and XAI.

After the completion of the first year of the project, the following conclusions can be established:

• Detailed requirements and specifications for the use cases have been defined

Figure 1. Respective visits to the facilities of IDESA and NODOSA to define the application scenarios for the technologies to be developed within CASANDRA

• A service-oriented digital architecture has been proposed to exchange data smoothly, interoperably, and reliably throughout the lifecycle, evolving towards a cloud-edge orchestration model capable of supporting a continuous Digital Thread between the plant and the cloud.
• After analyzing the functional requirements of the defined use cases, it was determined that the data security and integrity technologies currently implemented robustly meet the levels of trust required by the platform. The modular architecture of CASANDRA already guarantees traceability and immutability of records through highly efficient encryption and internal audit protocols, so the use of Blockchain infrastructure is not considered strictly necessary at this operational stage.
• The foundations have been laid for the generation of the product digital thread. Progress has been made in analyzing the practical relationship between CAD enriched with PMI, QIF for the definition of inspection features and tolerances, and their integration as the single reference product model. Likewise, the evaluation of AutomationML as a neutral exchange format and its linkage with AAS for structuring the product digital twin has begun. In parallel, the integration of OPC UA and AAS has been addressed as a key interoperability line.
• In order to develop tools and mathematical and physical models that allow the evaluation and prediction of the future behavior of manufacturing processes, finite element models for simple bending have been defined using ANSYS, and for the cladding process, initially in LS-Dyna and finally in ANSYS.
• User scenarios and worker task sequences have been defined in order to develop user-centered tools. In this regard, IDESA aims to redesign tasks adapted to real working conditions, considering both physical and cognitive ergonomics. NODOSA, on the other hand, aims to monitor confined spaces and exoskeletons, where operator fatigue can be monitored.
• In relation to the above, different types of sensors to be used to assess operator effort, stress, and fatigue levels, as well as the ergonomics of the activities performed, have been evaluated. In this sense, the use of different types of sensors has been proposed (IMUs, EMG, heart rate sensors, depth cameras, or textile sensors).
• For the specific plasma cutting scenario and the development of the user experience, a module for the quality control application has been designed and developed, including an adaptive graphical interface developed for sending images from a mobile device.
• An initial approach to solutions based on AR/VR technologies has been addressed to leverage operator monitoring information to improve safety in confined spaces through alert systems in critical situations, and to reinforce occupational risk prevention training through immersive experiences aimed at correcting ergonomic postures.
• Application scenarios for robotic technologies in intralogistics and collaborative robotics have been defined.

Figure 2. Interaction between a mobile platform and an industrial robot for intralogistics tasks, particularly the transport of naval sector parts between factory points while tasks such as GMAW welding are performed on them.

• Application scenarios for solutions based on augmented reality tools have been defined (projection system for positioning external elements and monitoring critical operator situations and generating ergonomic support alerts) to assist operators in manufacturing, assembly, and repair operations.

Figure 3. Projection system for positioning external elements on a bulkhead for the naval sector

• Different designs of passive upper-body and lumbar-support exoskeletons have been proposed to fit the activities performed by operators in the use cases defined by IDESA and NODOSA. The development and integration of a first prototype of a sensorized exoskeleton for data acquisition has begun.
• The development of the Asset Administration Shell (AAS) has started, with the aim of representing and managing the assets involved in the manufacturing of large components throughout their entire lifecycle. This is necessary for generating digital twins of products and processes.
• A benchmarking of Edge devices for distributed factory control has been carried out, selecting the NVIDIA Jetson AGX Orin Development Kit (64GB) platform as the most suitable for the CASANDRA project.

CASANDRA is made up of two consortia: a business consortium (CDTI grouping), and a research centers consortium (AEI grouping) formed by the following members

DGH – Spanish company located in the Boecillo Technology Park (Valladolid) specialized in industrial maintenance engineering. With more than 35 years of experience, DGH stands out in the design, construction, and commissioning of automated solutions for various industrial sectors. It is a benchmark in the automotive industry and has expanded its activity to sectors such as aeronautics, construction, and food.

IDESA – Founded in 1993 and headquartered in Avilés, it has become one of the most recognized and respected companies worldwide in the design, manufacture, and supply of static and modular equipment such as coke drums, vacuum columns, fractionation columns, reactors, FCC and FCK units, as well as all types of vessels and drums.

ONS – An SME founded in 2010 and based in Madrid, specializing in cloud infrastructure management and virtualization solutions. The company develops and supports OpenNebula, an open-source platform that enables the deployment and management of cloud computing environments and virtualized data centers.

NODOSA – NODOSA SHIPYARD is a leading Spanish company in the shipbuilding sector, dedicated to the construction, repair, and maintenance of ships. With more than four decades of experience, it has established itself as a benchmark both nationally and internationally, standing out for its innovation capacity and commitment to excellence.

ARSOFT – Spanish technology company focused on the development of Augmented and Virtual Reality (AR&VR) systems for the industrial sector, transforming traditional production processes into intelligent production processes.

ARSOFT has developed its own authoring platform (Eyeflow) for the digitalization of industrial processes using mixed reality (XR) technologies, reaching agreements with major leading companies in their sectors.

TRIMEK – Company founded in 1993 in Altube (Álava), focused on dimensional metrology engineering, developing measurement systems such as Coordinate Measuring Machines (CMMs), 3D scanning systems, and metrology management software platforms for different industrial sectors such as automotive, aeronautics, naval, energy, transport, and machine tools.

GOGOA – SME established in 2015 and based in Abadiño (Bizkaia), dedicated to the design, manufacture, and commercialization of exoskeletons to improve people’s physical capabilities. With one of the widest product ranges on the market and a solid intellectual property portfolio resulting from its strong commitment to R&D, it offers solutions for the clinical, industrial, military, and sports markets.

NVISION – Technology company specialized in bringing innovative solutions based on the Internet of Things (IoT) to market, combining technologies such as Artificial Intelligence (AI), data analysis (Big Data, data analytics), and data security and privacy (cybersecurity), to provide services oriented to monitoring and advanced system management, including decision-support tools and digital twins applicable to different sectors.

Research Centers Consortium (AEI grouping):

AIMEN – Located in Galicia, AIMEN is a private non-profit research association, a multidisciplinary center specialized in Materials Science, Manufacturing Process Engineering, Digitalization, Artificial Intelligence, and Sustainability. Created in 1967, its main objective is to increase technological competitiveness in the manufacturing industry through participation in R&D projects, providing advanced technological services to industry and technology transfer.

In recent years, AIMEN’s R&D activity has stood out in research on how digital technologies enable the application of new, more flexible, reconfigurable, sustainable, circular, and high-quality manufacturing strategies and methodologies, with the capacity to validate these developments both at laboratory scale (TRL4) and in representative environments (TRL5/6) using AIMEN’s industrial equipment.

IRI-CSIC – The Institut de Robòtica i Informàtica Industrial is a university institute jointly owned by the Spanish National Research Council (CSIC) and the Polytechnic University of Catalonia (UPC), founded in 1995. It is one of the leaders in robotics and computer vision research in Spain, given the volume and quality of its scientific output and its strong presence in European R&D projects with international partners. The center offers excellence and technological capabilities in robotics, computer vision, artificial intelligence, and automatic control, among others.

CDEI-UPC – The Universitat Politècnica de Catalunya (UPC) is an internationally recognized higher education institution focused on engineering, architecture, applied sciences, and technology. It stands out for its leadership in research and technology transfer, positioning it as a benchmark in both academic and business fields.

CDEI-UPC, the Industrial Equipment Design Center, specializes in the development of industrial products and equipment, dynamic analysis of complex systems, and mobile robotics, working closely with national and international companies to innovate in the design and optimization of industrial equipment.

CITIC-UDC – The Research Center in Information and Communication Technologies (CITIC) is a research center created in 2008 that promotes progress and excellence in applied ICT R&D. Driven by the University of A Coruña, CITIC is a meeting point between the University and industry where R&D departments from ICT companies and university researchers converge.

CITIC’s research is divided into four major areas: Artificial Intelligence, Data Science and Engineering, High-Performance Computing, and Intelligent Services and Networks; and a transversal line across all four: Cybersecurity.

INNOVALIA – Innovalia Association is a business R&D unit, member of the Basque Network of Science and Technology and a Technology Center founded in 2002 by a group of technology SMEs with international presence. This private research laboratory was created by the INNOVALIA group to build a solid base to achieve its long-term strategic research objectives.

INNOVALIA identifies and develops talent in partner companies, forming multidisciplinary teams to achieve long-term research objectives and meet the demand for advanced technological capabilities and transnational cooperation.

FUNDING: