Robots can serve restaurants, do acrobatics and dance, but one of the biggest challenges is forcing them to do activities that require fine motor skills.

That’s why I was surprised Model presented by researchers at the University of Tokyo, where a robot lifts and peels a banana with both arms in three minutes.

Although the two-arm machine is successful only 57% of the time, the rate is quite good considering the difficulties involved. For a robot to perform such a task.

The most interesting thing about this development is not that artificial intelligence can successfully peel a fruit, but that it offers a multitude of possibilities for the future, because such motor skills could help robots perform rigorous tasks. transporting small items from one place to another, picking up and storing delicate objects, etc.

Researchers Heecheol Kim, Yoshiyuki Ohmura, and Yasuo Kuniyoshi They trained the robot using a process. machine learning. In this type of training, several samples are taken to generate data that is then used by the robot to repeat the action.

Kuniyoshi thinks the training method can help AI systems perform all kinds of tasks. It may be simple for humans, but requires a lot of coordination and motor skills. This will support the use of this type of technology in homes, factories and any environment.

In recent years, several developments have emerged that aim to improve the capabilities of robots so that these machines can relieve many repetitive or routine activities.. As in this case, the focus is on training coordination, stability and fine motor skills.

Such is the case with researchers at the University of California, Berkeley. Motion2Vec algorithm, I’m looking to get youA robot can stitch patients with the precision of a human.

With this aim, they used one deep learning system semi-supervised, which the robot learns by watching videos of surgical interventions where stitches are made. With this knowledge, the AI ​​system learns to imitate the movements of healthcare professionals in order to accurately mimic them.

The developers used a Siamese neural network (siamese neural network) consists of using two identical meshes, which takes the two datasets separately and after processing, compares them and displays a final result.

On the one hand, the system takes the video of the doctor doing the stitches, and on the other hand, the application records of the robot. It compares the two clips and learns how to improve the precision of their movements.

The videos used in the training are part of the JIGSAWS database that collects information about surgical activity to model human movement.. The data forming part of JIGSWAS was collected through a collaboration between Johns Hopkins University (JHU) and Intuitive Surgical, Inc. (Sunnyvale, CA. ISI) as part of an IRB-approved study.

It is also compatible with this type of robot butlers. have since models that can pick up objects on the ground and until you order the chaos of the house robot chef to have as an ally in the kitchen. There are options for what you want to imagine, but the truth is that this technology has not yet become a part of everyday life or is still partially spreading because it still needs to mature, optimize some functions and also lower their value, something that will happen. widespread use of these devices.