The first phase of the project focused on assessing and identifying processes to define use cases and the technologies to be implemented. This resulted in the creation of a portfolio of user-centric tools with the potential to enhance key performance indicators (KPIs) such as Safety, Time, Process Control, Quality, Ergonomics, and Cost. As well as three testbenches installed at AIMEN facilities to support the incremental validation methodology.
During the second phase, the portfolio of technologies was finalised and validated at DF level by conducting three tests’ sprints on an iterative design and development procedure. Internal training sessions were also defined and carried out.
In the final phase of the project, the developed technologies were successfully demonstrated at the end users' facilities, NODOSA and BRODOSPLIT Shipyards.
Main outcomes are detailed below for the specific solutions:
Aerial surveillance technologies for shipyard. Small drones were successfully tested, despite stability challenges, for environmental measurement operations in confined spaces. Their performance was verified in terms of setup efficiency, measurement accuracy, and impact on other operations.
For production planning, a web-based platform was established for logistics planning tasks and reverse engineering. Tests were conducted using LIDAR-equipped drones. Interoperability of user-centric tools was ensured by creating a methodology for exchanging 3D data in STEP format and suggesting the OPC-UA protocol to connect robot information to digital infrastructure.
Robotic validations, Mobile manipulator showed significant improvements in functionality and robustness, especially in the robot's perception, grasping performance, and ability to select and use different grippers in real-time. The demonstrations highlighted the robot's ability to efficiently assist human operators in executing intralogistics tasks, thereby enhancing the ergonomics and appeal of such workstations.
Collaborative robotics, successfully implemented for welding and cutting applications in shipbuilding. The portability of the solution, its localisation capability and the automatic detection and execution of cuts within structures were demonstrated. Simultaneously, real-time data integration into the digital thread achieved vertical integration. Time savings and improved results are observed compared to manual methods. WelderBot technology proved not only viable but also transformative, significantly enhancing safety and efficiency in complex, high-risk shipyard tasks. Its adoption is expected to boost productivity and enhance worker well.
The testing of High payload robots in shared spaces with human achieved significant outcomes, delivering a functional demonstrator validated by project KPIs. Key achievements include a demonstrator for a ship subassembly, reducing cycle times and programming for welding paths, while improving ergonomics for operators. The hand guidance solution also met its objectives, achieving time savings and reducing the need for external elevation systems.
Augmented Reality and Mixed Reality solutions were developed using Tablet devices, headsets and projectors to help the workers in different manufacturing stages.
Projection systems provide an immersive Human-Machine Interface to assist human operators in performing marking and cutting tasks during ship outfitting, allowing a faster and more accurate performance, reducing mistakes and use of printed materials.
Handheld devices, offer a web-based application for progress authoring and monitoring, and a mobile application for supervision support. The results show the feasibility for current shipyard structures, providing a significant reduction of time on supervision tasks, and the immediate availability of the information in digital format.
Headset technology was developed to meet the need for educational tools adaptable to different jobs. AR could be easily adapted, enable skills transfer and knowledge retention. Tests of the device show positive results, robust, shock and dust proof, with noise cancellation features.
Exoskeletons deployed at end-user facilities to check technical aspects before entering the fabrication stage of the two novel independent solutions created in the Mari4_YARD initiative, developing a lightweight, semi-active spring-loaded exoskeleton for shoulder and lumbar flexion support, creating an AI-based algorithm for user effort recognition, and developing process perception technology using AI techniques. As a result of this development and subsequent validations two US provisional patents were filed by IUVO in November 2024.
Transversal tasks were completed with good results in terms of sustainability, communication and operational activities. A comprehensive set of training materials can be found on the project website, and the DF Network will continue its activity after the end of the project.
