Robots learn without human involvement: A new breakthrough in robotics

Scientists from the University of Surrey and the University of Hamburg have developed a revolutionary method that enables social robots to learn and be tested without human participation. This technology accelerates research into human-robot interaction and opens prospects for applications in education, healthcare, and customer service.

How Does the New Technology Work?

Social robots interact with humans through speech, gestures, and facial expressions, but their training has traditionally required human volunteers. The new method eliminates this need through a gaze prediction model known as the scanpath model. This model allows robots to “guess” where a human would look in a given situation, mimicking human eye movements.

Researchers used two sets of open-source data to train robots to replicate natural responses. The model compares the robot’s behavior to how humans allocate attention in similar scenarios, assessing how accurately the machine “looks” at key objects.

Advantages of the Method

“Our approach verifies whether a robot focuses on the right elements as a human would, all without real-time human involvement,” said Dr. Di Fu, co-author of the study and a lecturer in cognitive neuroscience. He emphasized that the system remains reliable even in noisy and unpredictable environments, making it suitable for real-world scenarios.

This approach reduces the cost and time of experiments that previously required volunteers. Robots can now be tested and trained in virtual environments, simplifying development.

The Future of Social Robots

According to Dr. Fu, the new technology aims to create robots with heightened social sensitivity. Such machines will better navigate complex communication situations by anticipating human behavior. This is particularly crucial for sensitive fields like elderly care or child education.

Scientists believe their method is a step toward creating smarter, more empathetic robots capable of not just mimicking human actions but predicting intentions. This could lead to machines that intuitively understand how to interact with humans across diverse contexts.

Conclusion

The development of a method for autonomous learning in social robots marks significant progress in robotics. This technology, enabling machines to learn without human involvement, not only accelerates research but also brings us closer to creating robots that seamlessly integrate into daily life. In the future, such machines could become indispensable assistants, understanding humans at a level previously unattainable by technology.