Although the idea of exoskeleton robots has been around since 1965, there have been only a few successful exoskeleton robot developments or commercialized products. It is common for exoskeleton robots to lag behind the wearer's motion intent, creating a time lag between the wearer's movement and the robot's response. 

 If the exoskeleton robot cannot accurately reflect the user's motion intentions in time, the interaction will be inconsistent with the user's intentions, which is directly related to safety issues. Therefore, in order to secure the synchronization of motion intention and response, it is necessary to develop a wearer's motion sequence prediction system to prepare the subsequent joint activation in advance according to the motion intention.

 Deep learning and computer vision techniques have been used to improve the accuracy of motion prediction. However, challenges still exist such as handling complex environments and accounting for uncertain factors. The development of accurate motion prediction technology has the potential to revolutionize various industries and improve safety and efficiency.