AI training: how mice make us see with the robotics!

AI training: how mice make us see with the robotics!

Researchers at the Technical University of Munich (TUM) have made significant progress in the field of artificial intelligence (AI). You have found that artificial neuronal networks, which are trained with biological data from the early development of the sense of vision, develop an improved ability to predict movements. This is not only important for humans and animals, but also for applications in robotics and driving autonomous. The research results show that the training of the neural networks can be optimized by a preventive program that is carried out in the retina before birth.

The preventive training that occurs in vertebrates before opening your eyes consists of spontaneous activity patterns. This activity spreads wavy on the nerve tissue of the eye and coordinates the early switching of the retina with the visual system of the brain. As Julyjana Gjorgjieva, Professor of Computational Neurosciences at TUM, explains, this natural training concept can have led to artificial neuronal networks work more efficiently because they are prepared for one scenario from the start.

The connection of AI and robotics

The results of this research are particularly relevant in the context of the robotics, where AI is increasingly ensuring more flexibility and adaptability. In the past, robots were rigid machines that only performed preprogrammed tasks. Thanks to the progress in the AI, robots can now see, feel and react to changes in their environment. This has far -reaching effects on different areas of application of robotics, such as in industrial production and service robotics.

In industrial robotics, AI-based image processing systems are used, which enable quality control and can recognize errors. Predictive maintenance also helps to identify potential failures at an early stage. The movement planning and avoidance of collision in real time is also crucial, especially for human-robot collaboration. This leads to a significant improvement in efficiency and security in production.

variety of possible uses

The applications of AI in service robotics are diverse. This includes household robots such as vacuum cleaners and robotic lawnmowers, which are increasingly entrusted with new tasks. Robot take on service tasks in the hospitality industry, while in the medicine there is robot support for diagnoses and care. Autonomous robots are also used in agriculture to efficiently edit fields.

The future of robotics is also shaped by developments such as the Advanced Machine Intelligence (AMI). Experts like Yann Lecun von Meta emphasize the need to develop world models in order to enable machines a deeper understanding of their surroundings. Here is both great potential and a number of challenges, such as clarifying the roles of AI and human intelligence that have to be covered.

Overall, the current developments show that AI offers great opportunities for robotics, but also raises resource consumption and ethical questions. It is clear to the TUM researchers: Innovative training for neural networks could be the key to further promote the integration of artificial intelligence into new technologies.

For more information on this exciting topic, please visit the articles from tum and Robotikverband .