Wolf Warrior Small Conference Room!

Ye Xiuwen, Luohan, Young Eagle, 'Xanadowa', and Shi Dangdang, sat down in priority and prepared for a meeting.

It stands to reason that this kind of high-level meeting, the Eagle and the 'Xanadowa', do not have the right to participate.

But this time, it was an exception, Shi Kangdang held an enlarged meeting and invited two people over.

At this time, Shi Dangdang took Ye Xiuwen's report, looked at it and said, "I really didn't expect that it would take half a month for you to leave, and so many things have happened." ”

"Yes, but there is still some gain, at least we have found a base for the Skeleton Legion.

Moreover, this base looks not small, maybe it's their headquarters, or not. ”

Ye Xiuwen said with a smile, and at this time, Shi Dangdang nodded and asked Xiang Ying, "Have you scouted this base?" ”

"Well, when the deputy team sent me the location, I used the reconnaissance satellite to lock on to the target.

But so far, nothing suspicious has been found. But this is a base, it can be confirmed. Within two hours, three cars were in and out of the area, which should still be in use. ”

The eaglet made a brief report.

"Yes, as long as they're still using it!" Shi dared to respond, and then said to 'Xanadova': "The little wolf brought back those robots, have you studied them?" So what the hell is that? ”

"I did some systematic research and found that this is a new generation of robots made in the shape of the sci-fi TV Iron Man,......."

"How well does it perform?" Ye Xiuwen interrupted 'Xanadova'.

"In terms of performance, I can only say that it is average. Although it is advanced. "'Xanadova' said.

"How?" Ye Xiuwen asked.

At this time, 'Xanadova' took out a mannequin and said, "Look, what I have in my hand is a mock-up of the human body.

I won't go into detail about the specific functions of each part.

I'll just give you an example. For example, when humans grasp, they don't need to think through their brains, it is a natural response plus muscle memory, and this reaction and muscle memory have actually been deeply imprinted in our motor nerves in our infancy through a lot of grasping exercises.

Now, human beings are looking back at this problem, in fact, they want to find a suitable way to control the manipulator to carry out humanoid grasping actions, and the following will show you the more famous manipulators in the industry."

As he spoke, 'Xanadova', he took out a number of pictures of the robotic arm for everyone to see.

In these pictures, there are full-drive robotic arms and under-driven robotic arms.

When Ye Xiuwen and others had some preliminary understanding of the robotic arm, 'Xanadova', then continued: "For the robotic arm, I think everyone already has a certain understanding.

Next, let's explain what the principle of grasping objects with human hands is.

From a biological point of view, the IKI defines the human hand as follows: the hand is part of the front end of the arm of a human or other primate, consisting of five fingers and the palm of the hand, which is mainly used to grasp and hold things, and the two hands are symmetrical and mirror images of each other.

The human hand is actually an extremely dexterous and complex human organ. The trunk is mainly composed of phalanges, and the skin that surrounds the trunk is a sensor with a large number of sensitive tactile nerves, and when the human hand needs to move concretely, the tendons act as a medium to transmit power, and the muscles and skin distributed in each joint play a good role as a compliant structure.

Among the five fingers, the thumb is the most special, first of all, it is composed of only two knots (Phalanges), and it is the most flexible of all fingers, and the finger with the widest range of end-effector workspace (ork space), specifically, its Metacarpals Joint is a ball joint.

The above is, what I am talking about, the "hardware" composition of the human hand, let's talk about the specific control.

In fact, when we use the current robotics to analyze this process, it is actually like this: when we want to reach out and grasp an object (AI makes a decision), we will roughly look at the position of the object (positioning through Vision), and when we go to grasp, We are mainly based on Trajectory Planning of the position of our fingertips, and we use our arms to move the palms of our hands to approach objects (Manipulation Progress).

In the process of approaching, our vision has been giving us feedback (Feedback), and when the object is in the operating space of the palm, the brain will control our specific fingers to grasp, and how to grasp is based on our large life experience, to automatically generate the optimal grasping scheme (Machine Learning Concept) suitable for the contour of the object.

In the operation of specific fingers, we still control the movement of each finger joint based on the trajectory planning of the fingertips (Under-Actuated or Fully Actuated ?). Eventually, the tactile nerve on the skin will give us a feedback on whether we have grasped something, and use the contact force transmitted by the tendon to contact the static friction generated by the object to operate the object, and the tactile nerve will always give us feedback on the shape, hardness and even temperature of the object, with the help of a large number of practical grasping experience, automatically generate the most optimized, most stable and most labor-saving grasping scheme.

And the basic description of the above large section of control, normal humans can complete the best within 1-2s (for example, now you go to get the mobile phone on the table, you don't need to think about it at all, you just reach out and take it).

Therefore, if you look at such a "body and system" as a human being from the perspective of a robot at the current level, it is simply terrifying.

Finally, let's talk about the specific functions of the human hand. There is no doubt that the specific grasping and manipulation of objects is the main function of the human hand, so it is natural that such a function as gesture can also be thought of.

As for the former's features, its performance can be described in two ways: dexterous and robustness.

Okay, now back to the second half of the question: how did the manipulator come to be?

The overall idea of designing the manipulator should return to the first half of the problem: what is the principle of human hand grasping objects? Specifically, it is necessary to design the hardware that can achieve the corresponding function according to the principle of grasping objects by human hand - that is, the manipulator.

Speaking of which, I think it's already clear to everyone. Even if a manipulator imitates a human grasping action, it is so complex, which shows how advanced the robot we are facing will be.

Let's look at it again, the mechanical arm in my hand is the ,....... brought back by the deputy team."

'Xanadova' said, took out a mechanical arm, and it was still able to move, waving to everyone present! ......

(End of chapter)