Revolutionizing Welding and Additive Manufacturing: The ROBIN Journey – Empowering Humans with AI-Driven Innovation

In the heart of Norway’s innovation landscape, where cutting-edge research meets real-world industrial needs, the ROBIN project was born. As partners in this biparty initiative—3D-Components AS (3DC) with its proprietary RobTrack AI system and Mechatronics Innovation Lab (MIL) with its world-class testing facilities—we’re on a mission to reshape robotic welding and additive manufacturing (AM). Launched under the ARISE project, ROBIN isn’t just about automation; it’s about creating intelligent, interactive workspaces that empower operators, minimize waste, and drive sustainability. Drawing from our combined expertise in AI-driven robotics and large-scale industrial validation, we’re building an open-source ecosystem that puts humans at the center while harnessing the power of ROS 2 and FIWARE. Let’s dive into how ROBIN is turning manufacturing challenges into opportunities for smarter, greener production.

The Challenge: Breaking Barriers in Welding and AM

Imagine a welding shop floor where operators wrestle with endless trial-and-error to dial in the right parameters—voltage, current, wire feed rate, robot speed—for each unique job. In robotic welding and AM, static settings and proprietary systems dominate, leading to defects, wasted materials, and hours lost to manual adjustments. Data from sensors and machines piles up, but it’s trapped in brand-specific silos, rarely shared or fully utilized. The result? Inefficient processes, high costs, and environmental strain, with global welding industries contributing to thousands of tons of daily waste and CO2 emissions.

At 3DC, we’ve tackled these issues through our RobTrack prototype, validated at TRL5 for parameter optimization. MIL, with its 2,000 m² state-of-the-art facility, sees these challenges across its 100+ industrial partners, from aerospace to maritime. The core problem is clear: a lack of standardized, data-driven tools that empower operators to make quick, informed decisions without extensive expertise. ROBIN was born to address this, offering a brand-agnostic platform that streamlines parameter selection, enhances quality, and keeps humans at the heart of the process.

The Solution: A Smart, Collaborative Ecosystem

ROBIN is a game-changer, integrating 3DC’s RobTrack with open-source ROS 2 and FIWARE middleware to create a seamless, intelligent system. At its core, RobTrack uses laser profilometry sensors mounted on welding torches, capturing real-time data like bead geometry and process metrics. Paired with GPU-accelerated AI, it delivers precise parameter recommendations, ensuring stable, high-quality welds. ROS 2 powers intuitive human-robot interaction, allowing operators to teach toolpaths and adjust settings via user-friendly interfaces. FIWARE’s Orion-LD Context Broker aggregates multi-source data, and triggers instant alerts for deviations—think catching a weld imperfection before it becomes a costly flaw.

This closed-loop system transforms raw data into actionable insights. Operators receive AI-driven suggestions, visualized on dashboards, to optimize parameters in real-time, reducing trial-and-error. Unlike proprietary systems, ROBIN’s open-source framework fosters interoperability across brands, enabling knowledge sharing while safeguarding IP. The result is a flexible, scalable platform that delivers consistent quality and efficiency, validated at MIL’s industrial-grade facilities for real-world reliability.

Human-Centricity: Operators as Strategic Innovators

ROBIN flips the script on automation’s reputation as a job replacer. Instead of sidelining workers, it positions them as strategic decision-makers. Operators interact with robots through intuitive ROS 2 interfaces, approving AI recommendations without needing coding skills. This reduces cognitive load and physical strain, as repetitive parameter tuning is automated, freeing workers for creative problem-solving. At MIL, where academic research meets industry, we’re refining these interfaces with operator feedback, ensuring they’re ergonomic and inclusive—multilingual and adaptable for diverse skill levels and abilities.

By empowering operators, ROBIN boosts job satisfaction and safety. It cuts training time, making advanced manufacturing accessible to novices and experts alike. Our approach aligns with Industry 5.0’s vision: technology that amplifies human expertise, creating 97 million new collaborative roles by 2025, per industry forecasts. From shop floor to boardroom, ROBIN fosters a culture where humans and AI work in harmony, driving productivity and engagement.

Contribution to ARISE: Pioneering Open Innovation

ROBIN is a cornerstone of ARISE’s mission to advance standardized, human-centric automation. By releasing our integrated ROS 2 and FIWARE application on GitHub, we’re fostering an open-source ecosystem that encourages community contributions while protecting proprietary AI algorithms. This open HRI module—covering data ingestion, real-time monitoring, and parameter optimization—sets a new standard for interoperability. It allows manufacturers to integrate ROBIN with existing systems, breaking free from vendor lock-ins.

Our collaboration leverages MIL’s testing prowess and 3DC’s AI innovation, creating a reusable framework for welding and AM. This supports ARISE’s goal of scalable, distributed robotics solutions, applicable not just to welding but to adjacent fields like aerospace component fabrication. By sharing knowledge through open standards, we’re accelerating industry-wide adoption of smart manufacturing.

Market Impact: Driving Sustainability and Competitiveness

ROBIN taps into the $4.76 billion arc welding equipment market (6.1% CAGR, reaching $8.11 billion by 2032), addressing a critical need for efficient, sustainable production. By optimizing parameters, it slashes material waste and energy consumption, aligning with EU Green Deal goals to reduce environmental impact. We estimate significant reductions in global welding waste and CO2 emissions, supporting circular economy principles.

For manufacturers, ROBIN delivers measurable value: improved first-time-right quality, reduced setup times, and enhanced flexibility for variable workloads in sectors like automotive, energy, and maritime. For operators, it means safer, more engaging roles with less manual drudgery. MIL’s “Test before Invest” methodology ensures these benefits are validated, building trust among SMEs and large firms. By fostering inclusivity and upskilling, ROBIN strengthens Europe’s industrial sovereignty and workforce resilience.

The Road Ahead: A Blueprint for the Future

ROBIN is more than a project; it’s a vision for manufacturing where technology empowers people and protects the planet. We’re starting with pilots at MIL, refining the system for real-world impact. Looking ahead, we’ll expand into diverse industries, leveraging partnerships with academic and industrial leaders to scale innovation. Our gratitude goes to ARISE for enabling this journey.

Join us in shaping the future of welding and AM. Explore our open-source app on GitHub, request a demo at MIL, or connect to collaborate on smarter, greener manufacturing. ROBIN is proof that humans and AI can build a better, more sustainable industry together.

3D-Components AS
Mechatronics Innovation Lab AS
ARISE – ROBIN

Related Articles

On the Ground Insights Informing the SSH Framework

March 24, 2026

In September of 2024, three of our researchers from Demos Helsinki arrived in Barcelona at the headquarters of PAL Robotics– a robotics company specializing in biped humanoid robotics. Unlike many of the visitors that PAL receives, we didn’t have much to offer in terms of technical or business expertise. Instead, we were there as ethnographers– documenting, observing, and engaging with robotics developers to understand how robots get built. In visiting the spaces where robotics development takes place, we were hoping to understand how we might develop a framework for human-centric robotics innovation that aligns with the day to day realities of HRI work.

PAL was only the first of several destinations. Over the next 2 weeks, we would listen in on team meetings, interview workers individually, host workshops, and collect over 12 hours of recordings of interactions we shared with robotics developers across 4 organizations. The HRI workers we met were generous with their time and candidness– they spoke openly about a range of topics from project management tools to internal team dynamics to the role that science fiction played in shaping their robot designs. 

From this body of qualitative data, we took note of insights that might help us develop an SSH framework for human-centricity that’s tailored for HRI contexts. After a preliminary round of analysis, we grouped our learnings into 4 overarching themes: 

  1. Structured and hierarchical decision-making: HRI work typically takes place in highly structured environments with a clear division of responsibility, agreed upon deadlines, and consistent workflows. Bringing in novel considerations, such as human centricity and ethics, into this highly structured work environment can be seen as disruptive rather than productive if not carefully introduced. Higher-level decision making may also not be accessible to everyone working on the project, nor at every phase of the project, leading to challenges in ownership and accountability when it comes to addressing emerging ethical concerns. 
  2. Limited capacity for interdisciplinarity: Due to the highly technical and complex nature of robotics development, HRI workers are often expected to utilize their specialized backgrounds to contribute to a niche part of the robotic solution, limiting their interactions with other teams and experts to what’s strictly necessary. HRI workers might encounter difficulties thinking in interdisciplinary terms because of the lack of shared language around human centricity in HRI development, the limited contextual scope of their contribution relative to the entire solution, and a general lack of time, energy, and resources for building new capacities.
  3. Context specific needs and challenges: Different HRI contexts pose different needs and challenges when it comes to ethics and human centricity. The type of organization the HRI solution is being developed in, target stakeholders, deployment contexts, cultural norms– all of these play a role in shaping how a team is able to conceptualize and mitigate ethical concerns. Developing a SSH framework that can be operationalized in a variety of contexts likely requires a carefully balanced, adaptable approach of both specificity and generalizability. 
  4. Leveraging existing norms, trends, and practices: While there are a number of reasons why HRI contexts might be resistant or unable to incorporate SSH perspectives, our team also came across existing norms and practices that might be leveraged to support human-centric robotics development. For instance, HRI developers typically work in iteration, incorporating learnings and failures in cycles of experiments. These cycles might be opportunities to inject small, manageable actions to continuously embed SSH perspectives and support.

 

Relying on the foundations of these four themes, we wrote the updated release of the ARISE SSH Framework (D3.2) in December of 2026**. Recognizing the needs and challenges of HRI work from our ethnographic visits, this second release focuses on operationalizing high level principles with concrete recommendations for frameworks, tools, and processes that can be adapted in HRI workflows.

Entering the final year of the ARISE project, our work continues to build on the belief that an informed and actionable ethics framework requires on-the-ground engagement. Throughout the open call process, our team is collaborating directly with beneficiary projects, engaging them in workshops and one on one discussions about their project’s ethics management approach.

The more we progress in this project the more it becomes apparent that the future of human-centric robotics starts with a willingness to open dialogue across disciplinary boundaries, whether that be in the digital spaces of a video conference or the break room of a robotics lab, surrounded by bi-pedal robots.

**This framework isn’t available for public viewing yet, but will be linked when it becomes available. Both deliverables already published are available in our resources page.

ARISE and FIWARE in the ADRA Book: Building Open Digital Infrastructures for Robotics

The recently published open-access book “Artificial Intelligence, Data and Robotics: Foundations, Transformations and Future Directions” provides a comprehensive overview of the technologies and initiatives shaping Europe’s digital future.

Published within the framework of the AI, Data and Robotics Association (ADRA), the book reflects the collective work of the European AI-Data-Robotics ecosystem, bringing together researchers, innovation projects, and industry stakeholders working to accelerate the development and deployment of these technologies across Europe.

Among the initiatives featured in the book is ARISE, together with its sister projects FORTIS and JARVIS, which are highlighted in the chapter “Advancing Industrial Collaboration: The Next Generation of Human-Robot Interaction.”

The book reflects the broader strategic vision behind the European AI, Data and Robotics Partnership, which aims to strengthen Europe’s leadership in these technologies while ensuring that innovation remains human-centric, trustworthy and aligned with European values.

Artificial intelligence, robotics and data technologies are increasingly converging to create new possibilities across sectors such as manufacturing, mobility, environmental monitoring and healthcare. This convergence is explored in the chapter “Data, AI, Robotics: Transformative Power in Industry 5.0” where the authors (among them partners in the ARISE project Polimi, Engineering and Cartif) explain how Industry 5.0 builds on the foundations of Industry 4.0 by integrating human-centric, resilient and sustainable manufacturing practices. This chapter reviews a series of European initiatives and industrial pilot cases demonstrating how AI, robotics and data technologies can support circular manufacturing processes, data sharing across value chains and new decision-support capabilities through technologies such as digital twins and real-time industrial data infrastructures.

However, unlocking the full potential of these technologies requires more than individual innovations. It requires digital infrastructures that allow heterogeneous systems to interoperate and share data seamlessly.

This is where open digital infrastructures become essential. Across the European ecosystem, researchers and innovators are increasingly focusing on building interoperable platforms and architectures that allow AI systems, robotics technologies and data infrastructures to work together across domains.

ARISE: contributing to the next generation of human-robot collaboration

In our chapter, we explore (together with our sister projects JARVIS and FORTIS) the evolution of human-robot interaction (HRI) and the technological foundations required to support collaborative robotics in industrial environments.

In particular, we present the ARISE framework, designed to accelerate the deployment of collaborative robotics systems that integrate human expertise with advanced automation technologies. Through experimentation in Testing and Experimentation Facilities (TEFs) and collaboration with SMEs, ARISE aims to bridge the gap between research and real industrial deployment.

Our work focuses on enabling more intuitive interaction between humans and robots, safer collaboration in industrial settings, scalable robotics deployments, and the integration of artificial intelligence into industrial decision-making processes.

FIWARE and ROS2: an open middleware for industrial robotics

A key innovation highlighted in the book is the architectural approach used in ARISE.

The project integrates FIWARE technologies with ROS2 robotics frameworks, creating an open middleware that enables interoperability between robotic systems, industrial applications and AI services.

This architecture allows data to flow seamlessly between robots, digital platforms and industrial systems, enabling:

  • real-time monitoring of robotic operations

  • integration of AI-based services

  • coordination between robots and industrial systems

  • scalable deployment of robotics applications.

By combining ROS2 robotics capabilities with FIWARE context-management technologies, ARISE demonstrates how open-source platforms can create a flexible and interoperable digital infrastructure for robotics.

This interoperability layer is essential for enabling complex industrial environments where robots, machines, sensors and digital platforms must work together.

Interoperable data infrastructures for AI and robotics

The role of FIWARE technologies is not limited to robotics architectures.

In another chapter of the book, dedicated to the CyclOps project, “CyclOps: Leveraging Semantic Technologies for AI and Data Life Cycle Management and Governance”, FIWARE technologies appear again through the use of NGSI-LD context brokers, which support interoperable data exchange and semantic data management across distributed environments.

FIWARE-based interoperability approaches are also referenced in discussions around emerging data ecosystems, such as in the chapter “Toward the Irish Mobility Data Space: Challenges, Opportunities, and Requirements”, which explores the technological foundations needed to enable cross-domain mobility data spaces.

Context brokers such as Orion-LD, Stellio and Scorpio, available within the FIWARE ecosystem, enable the storage, retrieval and sharing of contextual information across domains such as smart cities, Industry 4.0 and IoT applications.

These technologies illustrate how standardized data infrastructures can support the integration of AI and robotics systems within broader digital ecosystems.

Toward open digital infrastructures for robotics

Taken together, the projects and technologies presented in the book highlight an important transformation in the way Europe approaches technological innovation.

Rather than focusing solely on isolated technological components, the European AI-Data-Robotics ecosystem is increasingly building open digital infrastructures that allow different technologies to work together across sectors and domains.

Within this vision:

  • ARISE contributes to the development of human-centric robotics systems

  • FIWARE provides the interoperable data infrastructure that connects digital services, AI applications and industrial systems

This combination of open platforms, interoperable data architectures and collaborative robotics represents a key step toward the realization of Industry 5.0.

The inclusion of ARISE in this ADRA publication highlights the project’s contribution to Europe’s rapidly evolving AI, Data and Robotics ecosystem. By combining open-source technologies, interoperable architectures and human-centric design, ARISE demonstrates how collaborative robotics can be deployed in ways that enhance human capabilities while supporting industrial innovation. At the same time, the broader presence of FIWARE technologies across different chapters of the book reflects the growing importance of open standards and interoperable data infrastructures as foundations for the next generation of digital systems.

Together, initiatives such as ARISE and FIWARE are helping to build the open digital infrastructures that will support Europe’s future in AI, data and robotics.

VITAWELD: Vision, Intelligence and Human-Robot Teaming for the Future of Welding

VITAWELD tackles one of the most persistent challenges in industrial manufacturing: welding large metal components without compromising quality, efficiency, or accessibility. Traditional processes struggle with heat-induced distortions, late detection of defects, and manual parameter adjustments that slow production and introduce variability. Programming industrial robots for such tasks requires advanced technical expertise, creating a barrier for companies with limited resources.

The solution brings a fundamentally different approach to welding automation

VITAWELD blends advanced computer vision, AI, and natural user interfaces to keep the operator in control while enhancing their capabilities. The system enables robot programming through natural language and real-time visualization in augmented reality, making adjustments intuitive and immediate. AI-powered vision algorithms detect potential defects as they happen, while automated parameter optimization ensures repeatability and minimizes waste. Built on a modular ROS 2 and FIWARE architecture, the platform can adapt to different industrial contexts and scale with changing needs.

Within the ARISE project, VITAWELD exemplifies how collaborative robotics can democratize access to advanced automation. FIWARE integration ensures interoperability and supports a standards-based approach, making the solution replicable across industries. Most importantly, it reflects a vision where technology empowers rather than replaces, aligning closely with the principles of Industry 5.0.

The impact is measurable

Early projections indicate a reduction of more than 50% in robot programming time, over 30% improvement in first-time-right weld quality, and the ability to train new operators in less than two days. Energy consumption could be reduced by 15%, with significant savings in material waste. These advances not only improve productivity but also make welding safer and less physically demanding, encouraging greater diversity in the workforce.

Market potential is equally strong

The global industrial human-robot interaction market is expected to reach $8.6 billion by 2025, with the large-component welding segment valued at $1.2 billion. By focusing on amplifying human capabilities rather than replacing them, VITAWELD differentiates itself from conventional automation. Its modular design and human-centered philosophy allow rapid adaptation to other processes such as TIG, MIG, and additive manufacturing, delivering a return on investment in just 18 to 24 months.

Inclusivity is embedded in the design

Interfaces are multilingual and configurable, with universal iconography and color-blind-friendly visualizations. Ergonomic considerations accommodate operators of different heights, handedness, and working styles. The system can adapt to local cultural norms in units of measurement, alerts, and training materials, ensuring usability across global manufacturing environments.

VITAWELD represents more than a technical solution—it redefines the relationship between people and machines in industrial welding. By combining precision, adaptability, and inclusivity, it sets a new standard for human-robot teaming, one where innovation strengthens human expertise, fosters sustainable manufacturing, and builds a safer, more productive future.

Rovimática S.L.
IDONIAL Centro Tecnológico
ARISE – VITAWELD

All News
Powered by  GDPR Cookie Compliance
Privacy Overview

This website uses cookies so that we can provide you with the best user experience possible. Cookie information is stored in your browser and performs functions such as recognising you when you return to our website and helping our team to understand which sections of the website you find most interesting and useful.

More information in  Privacy Policy.