The average diameter of a human blood vessel is just 3 millimeters, and if the tip of the needle slips against the wall of the vessel or punctures the vessel, it will bring more pain to the patient. In order to guarantee the stability of the puncture to, HIT further technical research and development, so that the recognition of its robotic puncture reached 98%, to protect the medical effect and reduce the human pain.
After repeated polishing and optimization, the medical robot has been prepared for clinical trials at the end of 2021, and the future application of the robot in health care will also be popularized, becoming a good helper for nurses to work.
Viewers at the fair simulated the work of the medical robot
New design: 50 hours condensed to 5 minutes
In order to solve the problem of industrial design intelligence, the Central Research Institute of HAIG Intelligence has also invested great attention to the application of artificial intelligence methods, through the gradient descent method and the deep learning method and other methods, combined with its own experience, a large number of accumulated industrial data, industrial design industry knowledge and deep learning method, and a large number of industrial data, industrial design industry knowledge and deep learning method. The industrial data, industrial design industry knowledge and deep learning algorithm capability, developed an industrial design intelligent product platform, which is important to improve industrial design, simulation and industrial planning, etc., will be the past need more than ten hours of design work to shorten to a few minutes can be completed.
The future results of this artificial intelligence will be widely used in the industrial field of 3C, food, medicine, etc., which further reduces the cost of industrial smart manufacturing, efficiency is important.
New manufacturing: from acquisition to output is not higher than 20ms
At the 2019 China International Industry Fair, HAIG Intelligence released two AI industrial products, including the industrial intelligent OCR vision system.
The system is built on the basis of deep learning, and can realize difficult OCR recognition tasks such as curved surfaces in industrial scenes, character bending, and mutilation, etc. It only takes less than 20ms from image acquisition to result output, and the accuracy rate is able to reach 99.8%, which guarantees stable OCR detection.