Purpose Developed, in collaboration University of Twente (Netherlands), New Artificial Intelligence (AI) Platform Muscular models And tasks that are named Myo SuitePromises to advance in areas such as Prosthesis development and patient rehabilitation.

MyoSuite allows the use of machine learning on biomechanical control problems, Meta explains in a statement. To do this, he combines motor and nervous intelligence.

Until now, the company notes that the use of machine learning has been hampered by a lack of Physiologically realistic simulation platforms To address biomechanical control issues.

In this sense, MyoSuite is a muscle model More “computationally efficient and large-scale” than existing models, Which allows the simulation of the muscles that make up the human body. And they are 4000 times faster than other simulators When meeting the requirements of modern mechanical algorithms.

In addition, Meta argues that solving some more complex behaviors of motor control, such as a pen or key in hand, or manipulating baoding balls. These models can also support “time-rich muscle dynamics” such as Muscle fatigue and sarcopeniaAnd behavioral tasks appropriate to everyday situations.

The company hopes its results can contribute to real-world problems such as rehabilitation, prosthetics development and ergonomics. This is possible because your models can also highlight activation patterns and muscle groups involved.

For example, wood breakage is cited. First of all, MyoSuite can Transfer of model tendons, A technique commonly used to restore muscle function after this type of injury. In addition, he has the ability Simulation of the result in terms of mobility of the patient after surgery And the impact that rehabilitation will have on the affected parties.

The company has also made a reference to the meta-version and will provide it These models can help developers create avatars For a more expressive and immersive experience with more realistic movement.