Poor recognition accuracy means that the amplitude of bioelectric fluctuation is small, and it also means that the comparison error of two bioelectric waves is small.

But here, you can't just look at it as a percentage.

High precision, although the percentage is large, but the data sample she collects is richer, and it can cross from microvolts to millivolts, which is equivalent to two comparison schemes. It will only have a higher recognition rate than Ottobock's inductive patches.

Ye Qing repeated the action several times, and the error of each biological wave changed.

After a total of five times, Ye Qing asked the electrocrystal, if the program is allowed to intelligently identify, then the program can find the key and judgmentable hand retraction signal from these biological radio waves?

"Boss, in these five comparisons, the instrument has recorded four simultaneous bioelectric fluctuations from 32 microvolts to 66 microvolts. Standing in front of the wavesetter, the electrocrystal kept pinching its pointed chin, "If a complete biopotential database is established, I think it can be used as a basis for judgment through fuzzy comparison." ”

"But I can't guarantee that this will happen to everyone when they collect data. In addition, the number of sample collection is too small, and I think that at least tens of thousands of data comparisons can be more certain. ”

"It's not difficult to collect, there are a lot of people in the company. ”

It seems that there is no progress, but in fact, Ye Qing feels that there has been a lot of progress. As long as the signal of the palm can be accurately judged, it is equivalent to adding a 2 on the basis of 1. Because all non-1 signals can be classified as 2.

"Let's test the biopotential signal of the waving hand. ”

Ye Qing, who calmed down, began to wave his left and right hands simultaneously.

This action was repeated ten times.

Through comparison, Ye Qing was surprised to find that among these ten actions, ten of them were collected, which could be used to determine the same bioelectric fluctuation fragment for his own use.

After analysis, Ye Qing and Dianjing agreed that this judgment came from the bioelectric signal released when the wrist swinged.

It is also a muscle bioelectric signal released by the wrist muscles during movement. Because it only moves in the microvolt range and is highly consistent.

"There's a head. After reading the comparison chart, Dianjing looked at Ye Qing again.

"It's really funny. Ye Qing nodded, with an incredulous expression of "Is it that simple".

The signal of the palm opening released by the nerve is 1, and the signal of non-this action is 2.

The wrist or arm muscles, when doing a specific movement, can be listed as 3.

Of course, muscles can also produce bioelectrical activity during movement, but it is much smaller than the bioelectric signals released by nerves. Ottobock's inductive patch can't be recognized, and the electrocrystal manufacturing is completely okay.

1, 2, 3, so easy to find the idea?

"Boss, it's all too simple, isn't it? Don't you talk about Ottobork, the technology in this area is the first in the industry, I think this is the first, the water is a mess. ”

"I think it'...... too"

Ye Qing thought for a moment and said, "In the final analysis, it should be the gap in industrial technical capabilities. Perhaps they don't have the ability to improve the recognition accuracy of the sensor patch by an order of magnitude. They can detect changes in microvolts of biological voltage, but that device would be cumbersome and could be the same as an ECG recorder in a hospital. ”

"In addition, after detecting these subtle changes, it is another question whether they can be applied to the mechanical movement. ”

"This difficulty is higher than that of the induction patch, and without our bionic hand movement technology, no matter how good the induction chip is, it is useless. ”

"Finally, there's the programming difficulty. ”

"How to program these signals into a set of simple and effective control commands is also a problem. ”

Electrocrystal patted his chest and said leave everything to it.

Easy to understand?

This is easy to do, the electrocrystal can build a set of special control programs for the bionic hand. Then enable the super server cluster of the company and the factory, so that the program with a certain intelligence can separate all the actions of the human hand, and each action is simulated tens of billions of times, looking for the most concise set of commands.

Sample collection for bioelectricity?

The factory began mass production of high-precision induction patches that night, and by the afternoon of the next day, 500 sets of bioelectric collection devices were delivered to the company.

Each employee of the company spends two hours a day sitting on a chair and putting a patch on his arm, and making fixed two-handed movements over and over again according to the video.

Whether it's shaking hands, or clapping, or holding up a teacup, or giving a gesture of admiration or contempt.

These data sources are continuously gathered into the super service cluster, and every person, every gesture, will be compared in a centralized manner.

Data collection has not been without its challenges, and there have been many unexpected problems.

For example, when the server compares the abnormal data and submits it to the electrocrystal for review. By looking through the surveillance video, Dianjing found that the employee was holding his mobile phone in one hand to read the message during the recording of data.

Through repeated tests, the electrocrystal found that when the mobile phone is close to the induction patch, the electronic components working in the mobile phone will release electromagnetic interference, which will seriously affect the recognition accuracy of the patch.

In the packaging process of the induction chip on the production line, it has been treated with anti-interference, and now it seems that it is far from enough.

The electrocrystalline quickly came up with a solution, and these inductive patches are currently exposed during data acquisition. When it comes to the product, it is actually installed on the bionic hand, where it connects with the user's residual limb. Simply adding an electromagnetic shielding barrier to the outside of the bionic hand can solve this problem perfectly.

There are anomalies where data overflows, and there are anomalies where data is greatly weakened.

After the investigation, it was found that the data came from a young female employee who had rubbed body lotion and sunscreen on her arm.

The titanium dioxide and zinc oxide in the sunscreen shield severely block the transmission of bioelectric signals.

Electrocrystal asked the product department to add a note on the instruction manual that before use, please clean the limb connection and do not apply cosmetics.

There are also people who neither play with their mobile phones nor wear makeup, but the data has been abnormal.

The electrocrystalline turned over and over again and did not find out the cause of the problem, and later it was the cleaning department employee who thought of the problem. She said she had a mild epilepsy and had an EEG at the hospital last year, and the doctor said she had abnormal nerve discharges.

There are many more similar problems.

But every time, the electrocrystal always comes up with a quick solution.

In addition, the hardware changes in the hands of the second-generation Bionic are also significant.

Inside the second-generation Bionic Hand, Electrocrystal has added a set of processor system similar to Phantom mobile phones.

Users can connect through their mobile phones to customize and modify operation commands. You can also upgrade the system of the bionic hand anytime and anywhere through your mobile phone.

The connection to the residual limb has also been doubled so that the bionic hand has a stronger connection with the user.

The electrocrystals also collect the nerve signals fed back by the palms of the employees when they come into contact with various objects. It tries to feed these signals back to the end of the user's residual limb through another nerve feedback patch of the bionic hand, so that users can experience the touch brought by the real palm.

At present, this system is not perfect, and can only simulate 30% of the touch, which is more like the feeling of numbness.

However, Dianjing thinks that it can give users a collection chip and let him stick it to other parts of his body to collect tactile signals, and then tailor it to everyone's exclusive touch.

In the midst of the busyness, news also came from Mead Hospital, and they had a plan for the timing of the promotion of the bionic hand.