Exploring collaborative robots in manufacturing automation

Collaborative robots, or cobots, are increasingly becoming a practical tool for manufacturers seeking flexible automation without the complexity of traditional industrial robot cells

The R&D team at Vestas aircoil has been exploring how cobots can support and enhance selected production operations, with the aim of building internal knowledge and reducing future integration effort.

Establishing a cobot platform in R&D

The first step in this journey was the installation of the Universal robot UR10, a medium-duty industrial cobot that offers both a long reach and a high payload capacity, in our R&D laboratory. This provided a controlled environment where engineers could experiment, prototype, and evaluate automation concepts before considering deployment in production.

Cobots are designed to operate safely alongside people, making them particularly attractive for environments with frequent changeovers, mixed production, or limited space. By starting in R&D, we created a foundation for hands-on learning and realistic evaluation of both technical capabilities and practical limitations.

The initial focus was not to automate everything, but to understand which types of tasks are well suited for cobots and where complexity, accuracy, or cycle time could become limiting factors.

Selecting candidate processes

Early evaluations focused on a set of representative production tasks with varying levels of complexity. These included operations such as welding, tube handling, testing, and assembly-related activities. Each candidate was assessed against two key criteria: business impact and ease of implementation.

This approach helped structure discussions and ensured that automation efforts were driven by value rather than technology alone. It also made clear that some processes, while technically feasible, may not be the right starting point due to sensitivity, precision requirements, or throughput constraints.

Welding automation with a cobot

One of the most mature use cases explored was cobot-based welding. The objective was to assess whether a collaborative welding solution could handle a significant share of recurring weld tasks while remaining simple enough for daily operation by production personnel.

The system is the new Migatronic CoWelder 360 for MIG/MAG welding, based on a Universal Robots UR8 Long robot with Polyscope X controller.

The welding setup allows loading and unloading in parallel with welding, supporting efficient use of operator time. Advanced features such as seam tracking, touch sensing, and offset correction were evaluated to understand how well a cobot could handle natural variation in parts and fixtures.

Beyond cost and capacity considerations, the project also highlighted an important human factor: modern automation technologies can make manufacturing roles more attractive, particularly to younger employees who expect digital tools and automated equipment as part of their working environment.

Tube testing: Lessons learned

Another key area of investigation was automated testing using a cobot- mounted tube tester. Earlier manual solutions had demonstrated technical feasibility but also revealed challenges related to cycle time and handling variability.

By letting the cobot provide positioning force and movement, the need for a self-supporting test device was eliminated. Prototype testing showed that the concept works and that faulty joints can be detected reliably under controlled conditions.

However, the trials also highlighted limitations. Small misalignments can significantly affect test results, and maintaining stable force over time proved challenging. These insights were valuable, as they clarified where further development would be required before such a solution could be scaled for production use.

Learning through feasibility Sprints

To accelerate learning, short feasibility sprints were used to explore specific challenges. In one such sprint, the team focused on integrating force control, collision detection, and external control systems. This allowed rapid validation of core concepts without committing to full-scale development.

The outcome was not only technical insight, but also a better understanding of how quickly a cobot can be programmed and adapted to new tasks. This flexibility is one of the strongest arguments for collaborative robots in environments where products and processes evolve over time.

Tube insertion and human-like motion

One of the more challenging experiments involved inserting tubes into a fin stack, a task traditionally considered difficult to automate due to tight tolerances and variable resistance.

Using a simple end effector and force feedback, the cobot was able to perform the task successfully after a short development period. The solution relied on controlled motion, adaptive speed, and small corrective movements that mimic human behavior rather than rigid, pre-programmed paths.

This experiment demonstrated that cobots can handle tasks that benefit from compliance and adaptability, provided that expectations around speed and throughput are realistic.

Building a foundation for future automation

The overarching goal of these activities has been to build internal competence and reduce uncertainty. By gaining hands-on experience in R&D, the organization is better prepared to engage with system integrators, define realistic requirements, and shorten integration time for future automation projects.

Rather than viewing cobots as a universal solution, the work has reinforced the importance of careful process selection, gradual scaling, and close collaboration between engineering, production, and operators.

The journey with collaborative robots is ongoing, but the lessons learned so far provide a strong foundation for making informed, value-driven decisions about automation in the years ahead.