Training humanoid robots for the real-world workplace – automation, digitalization and AI

Humanoid robots open up new possibilities for industry, crafts, horticulture and agriculture – especially in areas where traditional automation reaches its limits.

Fraunhofer IFAM in Stade is establishing a training and competence center that, for the first time, systematically combines humanoid robotics with practical training, application-oriented research and direct technology transfer. In a realistic environment, not only do people learn how to interact with humanoid robots, but, in particular, the robots also learn from the application- and practice-oriented experience of humans.

Training and competence center for humanoid robots – a place for learning, testing and transfer

Humanoid robotics is on the verge of transitioning from a research topic to practical application. At the same time, many companies and organizations still lack the experience, confidence and suitable testing environments needed to evaluate this technology in a realistic setting.

This is exactly where the Fraunhofer IFAM training and competence center for humanoid robots in Stade starts:

• The center provides an open, practice-oriented space where humanoid robot systems can be tested outside of normal operations.
• Here, companies as well as businesses, skilled workers, trainees and researchers work together on real-world tasks for humanoid robots – from initial feasibility studies to demonstrators up to qualification and integration concepts.

The focus is not on technology alone, but on its meaningful use in conjunction with people.

Learn from each other: Humans train robots – robots assist humans

A central unique selling point of the training and competence center is its bidirectional training approach: While people learn how to interact with humanoid robots safely, efficiently, and usefully, the robots are simultaneously trained for their future work environments.

Experienced professionals – such as journeymen and journeywomen, master craftsmen and master craftswomen as well as technical experts – contribute their technical and implicit knowledge: manual skills, movement sequences, decision-making logic and useful tricks based on practical experience. This experiential knowledge is systematically captured through demonstrations, sensory recordings as well as AI-supported learning methods, and then translated into digital models.

This gives rise to a new form of “hands-on training” for humanoid robots – similar to the traditional apprenticeship system for trainees in the workplace – but here transferred to adaptive algorithms and motion models.

From empirical knowledge to a digital model

The training and competence center establishes the technical and methodological foundation for making human expert knowledge permanently accessible. Demonstrations, movements, force trajectories and decision-making processes are analyzed, abstracted and translated into artificial intelligence (AI) models.

These AI models are not tied to a single robot system. Rather, they are developed in such a way that they can be transferred to and reused on individual systems regardless of the manufacturer or design. The result is a scalable knowledge pool that makes it possible to efficiently adapt proven capabilities to new systems, tasks or industries.

For companies and organizations, this means: less development effort, faster integration and greater acceptance in the work environment.

Cross-industry test environment for production – from manufacturing to agriculture

The training and competence center for humanoid robots in Stade deliberately positions itself as a safe experimental space – “playground” – for humanoid robotics. Here, interested parties can bring their own processes, components or activities and, together with the experts in Automation and Production Technology at Fraunhofer IFAM, investigate whether and how humanoid robots can be usefully applied in their specific practical contexts.

This is not about short-term automation, but about making well-founded decisions:

• What is technically feasible?
• Where does the economic value lie?
• How will work change for employees?
• What skills will be required?
• How will it be integrated into practice?

This cross-industry approach forms the basis for a wide range of specialized applications – from industrial production and assembly via logistics up to agriculture and horticulture.

Training, qualification and acceptance

In addition to technological development, qualification plays a central role. The training and competence center offers practice-oriented training programs in which specialists learn about the safe handling of humanoid robots, evaluate application scenarios and gain their own experience.

Acceptance is built through direct interaction with the technology – not through theory, but through individual experience. At the same time, user feedback is immediately incorporated into the further development of the systems. This creates a continuous learning cycle between humans, machines and AI.

 

Fraunhofer IFAM in Stade – the strategic location with perspective

With the training and competence center for humanoid robotics at its Stade location, Fraunhofer IFAM is establishing a hub for application-oriented robotics research and development, qualification and transfer. The close connection between research, training and practice creates an environment in which innovation does not arise in isolation, but is developed collaboratively.

From here, branch- and application-specific development projects and demonstrators can be established in the future – for example, for commercial horticulture including fruit growing, skilled trades, industrial manufacturing and other fields of work that involve a high degree of manual labor.

Fraunhofer IFAM brings many years of expertise in research and development in the fields of automation, robotics, human-machine interaction and transfer into practice to the training and competence center. The goal is to shape humanoid robotics not as a vision of the future, but as a practicable technology for real, efficient work environments in the context of Industry 5.0.